New Customer Acquisition for Multichannel Retailers

Successful New Customer Acquisition: A Perspective  

A brand has only two fundamental sources of future revenue: purchases by customers who have bought from the brand in the past and purchases by those who have not bought previously. Repeat buyers drive current profitability. New buyers drive growth and future profitability. Both aspects of this success equation are critical.  

This article deals with acquiring new customers. Years ago, legendary direct mail genius Ed Mayer first articulated the now well-known axiom that "Success in direct mail is 40 percent lists, 40 percent offer, and 20 percent everything else." Whether or not one agrees with Ed's specific percentages, no reasonable marketer disputes the fact that reaching the right audience with a strong offer is the foundation of success.  

In today's world it remains true in the direct mail channel that selecting the right lists to make up a prospecting audience is absolutely essential. The principle applies just as much in the digital channels. To be successful, digital display, video, mobile, and social network advertising must reach a productive audience.  

Methods for developing audiences have changed markedly over the years. This has been true in the direct mail channel for decades, dating back to early use of computers (see Multichannel Retail, Today & Tomorrow for a historical view). Change has been even more rapid over the past 20 years in the digital world. The 1990s seem like a very long time ago.  

More change is coming. Costs will continue to rise and technology will change. Costs will rise, for example, because of higher postage in the direct mail channel and higher media cost in digital channels as more and more marketers bid for the same eyeballs at the same time. And the pace of change in technology will at the very least continue and it may accelerate. At Wiland we want to help our clients be successful despite higher costs and rapid change. That's what we are all about: Delivering the Future. As it relates to new customer acquisition this means we strive to deliver a future in which our clients waste less marking budget (offsetting rising costs) and are more successful. We use our unparalleled data and technology resources to create better, more productive audiences that are affordable yet produce superior long-term ROI for our clients. 

Every Multichannel Retail brand should utilize our technology and the consumer intelligence we provide to help them compete more effectively and be more successful despite rising costs and massive change. One way we do this is by helping our clients focus on acquiring the right new customers, those that are likely to produce good Long-Term Value (LTV). In the years ahead, Multichannel Retailers need to increase their focus on strategic new customer acquisition, not just on an acceptable cost of initial customer acquisition but also on the contribution new customers make to overhead and profit over the long run. 

Successful customer acquisition is critical to marketing success and long-term organizational viability. It is the lifeblood of most Multichannel Retail companies. Every brand should set aside time to think strategically about how it acquires new customers, whether some of them are undesirable, and how to improve the overall value to the brand that new customers provide. This is what Wiland does. We help our clients stop investing to acquire customers that never buy again or buy so little that they are never profitable; and, we help our clients identify and target desirable new customers that drive brand success. 

Successful New Customer Acquisition: Replacing Lost Customers with Better Customers and More of Them

Every company experiences some degree of customer attrition each year, ranging from 10% to 20% on the low end to 80% or higher for other merchants. In all cases, lost customers must be replaced with new or "reactivated" buyers in order to preserve or grow revenue and maintain a healthy, vibrant business. For Multichannel Retailers seeking significant growth, the importance of successful prospecting is even greater.  

But what exactly is "successful prospecting?" It has been said that the main objective of customer acquisition is to do the least amount of work, and spend the least amount of money, to get as many customers as possible into the fold. Following this logic, a marketer would be expected to promote all prospect audiences for which estimated cost per customer acquired falls below the acceptable threshold. On the surface, this seems like sound marketing strategy, as it aims to maximize the number of new customers acquired while limiting investment to an acceptable level. Indeed, many Multichannel Retailers have historically followed this methodology, marketing to all prospect lists and other non-customer audiences whose performance meets or exceeds a stated short-term cost objective. This approach can produce good results, but sometimes it produces bad results. Even if the result of using this "tried and true" approach is good, that does not necessarily mean it is the best method. Something else could be better, maybe a lot better.  

At Wiland, we believe there is a much better approach, one that produces superior long-term brand success. We have proven it repeatedly in our years of experience with clients. While we acknowledge the importance of achieving quantity objectives in customer acquisition, we strongly believe that the quality of customers matters just as much; maybe even more. In order to achieve sustained success and profit growth, Multichannel Retailers must focus on adding customers who produce strong Long-Term Value (LTV). There really is no value in investing to acquire new customers who will never buy again, or who will buy so seldom that the initial investment in them is never recovered. In fact, doing so is detrimental. Thus, our view of successful prospecting is that lost customers must be replaced with a larger number of new customers that, on average, are superior to those lost to attrition; and superior means that they produce higher LTV, leading to revenue growth and a better bottom line.  

Delivering Value: High-LTV Prospects

Company founder Phil Wiland and his team have been helping Multichannel Retailers acquire new customers since the 1970s, well before anyone had fully articulated the concept of a "multichannel retailer." No other company can match the depth and breadth of Wiland's capabilities and experience with new customer acquisition. Leveraging our massive data resources, advanced technology, proprietary analytical platform, and talented, seasoned staff, Wiland creates and delivers offline and online prospect audiences that not only perform well initially but generate new customers who buy again at a high rate, delivering strong LTV.  

We help Multichannel Retailers improve the productivity of acquisition marketing in all promotional channels. We enable clients to optimize their promotional budget and bolster their financial performance by substantially reducing marketing budget waste and spending the saved budget more wisely, driving both bottom line and top line growth. Many of our clients have discovered that new customers acquired using Wiland prospect audiences produce new customers who maintain an ongoing relationship with them for a longer time than is the case with other prospect sources. And they spend more money in total, adding margin points to the brand's bottom line. Many others are enjoying the benefit of superior LTV customers acquired via Wiland but don't realize it because they don't track LTV by new customer source. Sadly, some are not experiencing superior results because they or their agency or broker insist that we focus on initial response rate without regard for whether the new customers so acquired ever produce any profit. A key part of our strategy is to encourage our clients to focus not just on acquiring new customers but also on whether the new customers they are acquiring are worth the investment.  

Wiland's unparalleled expertise in delivering desirable new customers is evidenced in our high growth rate in the multichannel retail market. Everyone knows that this has not been a high growth market in the past decade. Yet, for the past ten years, Wiland has grown 15% to 50% every year, without exception, right through the recession and continuing this year. We are not capturing market share because we sell more aggressively. Quite the opposite: We are growing rapidly because we focus on doing right by our clients, finding them a large volume of high potential prospects that become high LTV customers, driving client growth and success. 

Key Advantages: Data Depth and Superior Strategy

Wiland offers Multichannel Retailers a tremendous advantage: a massive database with a diversity and depth of transactional data that surpasses others in the industry. We track not only what, when, and from whom consumers buy, but also the channel utilized for each purchase. With over 2,500 participating clients, the Wiland database includes transactions from large, well-known retail brands, huge internet retailers, small specialty merchants, and everyone in between. Collectively, our client base offers consumers virtually every product category imaginable.  

In addition to merchandise transactions, Wiland has unsurpassed publishing and nonprofit data. Our database includes the magazine and newsletter subscription activities of about 100 million consumers, as well as their donation activity across hundreds of charities. These non-merchandise behaviors are often strong predictors of merchandise purchases. Our database also includes demographic and psychographic data on virtually every consumer in the United States. The result is a robust, 360-degree view of buyer behavior that enables the creation of new customer acquisition audiences that precisely target the right prospects for a given brand and reach those audiences across the full range of channels used in today's multimedia environment.  

Another key advantage is our dedication to good marketing strategy. Some providers of prospect audiences measure their success by how many prospects they provide. We want to provide more prospects, of course, but that isn't how we measure success. We want to provide more prospects that deliver good long-term value for our clients. This difference in strategic approach makes us a superior new customer acquisition partner. Clients that work closely with us to acquire desirable new customers ultimately improve their brand as a result. We strategize. We improve. We refine. And ultimately, our clients win the prize: more new customers that actually produce good LTV.  

It is important to point out that our clients don't have to choose between quantity and quality. We want to deliver both. We keep the focus on quality and encourage more testing and refinement. Clients who work with us in an iterative fashion of test and refinement ultimately obtain more new customers than before. These new customers are of better quality, so over time the single-buyer portion of their customer base declines, repeat orders come sooner and are more frequent, and our client grows, with both revenue and profit. It isn't easy. You have to stop and think about it. You have to change the way you view customer acquisition. But isn't that what good business people do? They question current practices and attempt to beat them. Do you want better results tomorrow than you have today?  

We're Ready to Help

Whether the main objective is rapid growth, profit maximization, or a balanced combination of both, Wiland stands ready to help Multichannel Retailers realize a better future—a future of superior new customer acquisition campaigns that improve profitability and brand strength.  

In the sections that follow, we will delve further into successful customer acquisition strategies and discuss how Wiland helps clients execute them to meet their objectives.  


Customer Acquisition Strategies

Establishing the Correct Marketing Objectives

A brand's customer acquisition activities occur in the broader context of the firm's overall marketing efforts. The marketing staff at a Multichannel Retail company should attempt to achieve Three Prime Marketing Objectives, depicted in the graphic below. The essence of good overall marketing strategy is constantly striving to achieve each of those objectives.


 

While Wiland assists Multichannel Retailers in achieving all of the Three Prime Marketing Objectives, the focus of this section is Objective #1: Affordably Acquire Desirable New Customers that Produce Good Long-Term Value (LTV).  

The goal should be to acquire the right customers in sufficient quantity to sustain, strengthen, and grow the brand.

This objective is fundamental. Without customers that produce good LTV, a brand cannot prosper. The majority of a merchant's profits come from marketing to established customers, but today's established, profitable, loyal customers were once prospects in whom an initial investment was likely made. Good brand marketing begins with good customer acquisition strategy that produces profitable new customers. Most brands make two mistakes in their thinking about new customer acquisition, potentially dangerous mistakes that diminish brand profitability:  

  • The first mistake is operating as if the job were simply to acquire new customers in sufficient quantity. Quantity is crucial, but it is not enough. Quality matters too. It is actually harmful to invest in acquiring a new customer who never buys again, or buys so little that the initial investment in them is never recovered.  
  • The second mistake is assuming that LTV is consistent across sources, and as a result, not tracking and studying LTV sufficiently. LTV is not consistent across sources. Some sources are simply better than others from an LTV viewpoint.  

The goal should be to acquire the right customers in sufficient quantity to sustain, strengthen, and grow the brand. Large investments are often made in acquiring new customers, many of which will never buy a second time. In many cases, the brand will promote these "single buyers" many times before finally discontinuing marketing to them, creating a severely negative impact on the bottom line. With such a substantial investment being made in acquiring and converting new customers it is essential that the marketing team focus on acquiring the right new customers to drive the brand—those that will produce adequate long-term value. 

Finding the Right New Customers

The goal of acquiring the right new customers in sufficient quantity to drive brand success requires that "right" be defined. Wiland helps clients with this definition by using a carefully devised methodology that is focused not just on initial response but also on long-term value. Whatever method is used, there should be clarity on who the right customers are. A general definition such as "people who will buy again" is not sufficient. Every brand should have a very specific policy statement concerning its new customer acquisition efforts, such as:

"The focus of our prospecting strategy is to acquire desirable new customers whose ongoing purchase activity results in the recovery of initial acquisition investment plus contribution of at least XX% to overhead and profit within NN months. We will also focus on eliminating acquisition spending that produces undesirable new customers."

At the tactical level, we help clients pursue such an objective in several ways:

  • We want to understand client marketing objectives. Most clients share strategy with us. Some don't. We produce our best results when our team understands the client's goals, problems, and opportunities. We design our solutions to maximize our clients' goals.  
  • Our models predict demand or 'surrogate' LTV, not just response rate. Too many prospect sources provide a large volume of recent buyers that produce decent response but fail to deliver good LTV. These sources are optimizing response rate over demand thus treating all responses equally. Our approach identifies more of the 'right' customers and puts money on our client's bottom line rather than feeding them a lot of unprofitable customers.  
  • Our models leverage the breadth and depth of our database. The models that we develop take thousands of variables into account. They consider historical spending levels, intervals between purchases and other factors that increase the likelihood of producing good LTV. Our approach is also sensitive to data anomalies. Let's say, for example, that when viewed collectively, consumers who bought recently in a particular product category produce good response. Now, assume that buried in such a "good cell" are people who have been buyers for years but had never purchased in the product category prior to their recent transaction, which was a $9.00 gift for someone at their office who is interested in a product category in which the buyer has absolutely no personal interest. This person almost certainly is not a desirable prospective new customer. Yet, a pure recency select would include them. Our models systematically work to remove these undesirable consumers. 

Wiland Client Service Representatives and Analysts do their best work when they know a lot about the client's brand and its needs. Thus, the more open clients are with us about strategic objectives and problems, the better job we can do to increase the LTV of new customers acquired from our prospect audiences. 

Defining Long-Term Value

Throughout this discussion, we've used the term "Long-Term Value," rather than the often used term "Lifetime Value." We are not critical of those who speak of lifetime value and promote that concept. It is a good concept. But we believe Long-Term Value is a better way to think about the subject and establish policy for several reasons:  

  • Most who discuss "Lifetime Value" really do so in a Long-Term Value context anyway. They don't propose that marketers invest in people age 20 and try to make a lifetime profit on them by the time they are 90. There is virtually always a timeframe involved in lifetime value calculations.  
  • It is a very rare brand (maybe non-existent) that can afford to wait a lifetime to achieve profits on a new customer. Even long periods that are less than a lifetime likely are "too long". For example, while it may be theoretically satisfactory to invest in new customers if their repeat purchases over the next, say, ten years justify the initial investment, there is a short-term component to financial success that likely makes ten years too long a period over which to recoup investment in new customers.  
  • Things change over time. The longer the timeframe, the more they change. Even if good math indicates that a new customer will generate sufficient ROI within ten years, it probably isn't wise to pursue a ten-year return strategy. Tastes change. Cultures change. Attitudes change. The competitive environment changes. Everything changes. The economic and competitive environment that produced an accurate ten-year lifetime value analysis may no longer exist ten years from now.  

Thus, we promote the use of a Long-Term Value strategy rather than lifetime value. But, how long is long-term? The answer is not the same for every brand, but there are several guidelines that we believe apply to most situations:  

  • "Immediate breakeven" is almost always the wrong strategy. Some companies seek immediate breakeven from new customer acquisition, ignoring LTV altogether. While this approach may be necessary for a brand that is undercapitalized, it is almost always a mistake for companies seeking growth and longer term viability. Would a merchant lose $10 to obtain a new customer on which it will earn a $40 profit within six months? The answer should be "yes."  
  • Five years is the outside limit. Large, well-capitalized sellers of non-trendy merchandise may be able to operate successfully with a strategy to recover initial investment and have respectable margin within five years of acquiring a new customer, but in most cases five years is too long. Longer becomes very risky.  
  • The right answer is usually one to three years. Within this window, risk tolerance should likely be the guide. If capital to sustain a brand through a possible downturn is sufficient then the right answer may be three years. If the brand is adequately capitalized but might have severe cash issues in a downturn, then 1 year may be the right answer. If the brand has cash issues now, something shorter than 1 year may be necessary, possibly even the extreme of immediate breakeven.

The key is to decide on a payback period and then track long-term results accordingly. Then, having established the payback period, the brand should make two more LTV related decisions. Only then will a complete LTV strategy be in place.  

The key is to decide on a payback period and then track long-term results accordingly.

First, the marketer must determine what must have occurred by the end of the chosen payback period. Must the full investment in new customers have been recovered plus at least one penny? Or, must the full investment have been recovered plus a respectable, clearly defined ROI? Either can work, but a brand must decide on one approach or the other. A key consideration is this: If a brand chooses a short payback period and requires a high ROI by the end of that short payback period it has effectively put a cap on its growth potential. On the other hand, if a brand chooses a long payback period and requires only breakeven by the end of that long payback period it has effectively chosen to emphasize growth at the expense of short-term profit margin.  

Second, what exactly does "good LTV" mean? This has just one correct answer. It means that a set of newly acquired customers must, as a set, produce XX% ROI or higher within a specified period of time. XX% may vary from one brand to another, but should never be less than the brand's overhead percentage plus something. New customers that never pay their fair share of overhead plus something for the bottom line should never have been acquired. They are undesirable new customers.  

This is what the concept of LTV is all about: acquiring new customers that will, within a reasonable timeframe, produce enough marketing profit to pay a fair share of overhead and contribute to the bottom line.

The goal should be to acquire desirable customers, not undesirable ones, and acquire them in sufficient quantity to maintain a strong brand. This is what the concept of LTV is all about: acquiring new customers that will, within a reasonable timeframe, produce enough profit to pay a fair share of overhead and contribute to the bottom line. Marketers should focus as much attention on avoiding undesirable new customers as they do finding the desirable ones.  

All too often, a brand pays little or no attention to the quality of new customers being acquired. Rather, the brand assumes that a new customer is desirable if it was produced by a customer acquisition campaign or segment of a campaign that produced acceptable overall results initially. This approach may never have been wise, but in times past it may have been a reasonable point of view because costs were lower and segmentation tools were less advanced. Today, the "any new customer is desirable so long as it came from an acceptable segment" philosophy is not only outdated, it is harmful.  

Marketers would be well advised to look under the hood. They must, before making the marketing investment, identify and eliminate prospects within new customer acquisition audiences who are either unlikely to buy in the first place or, if they do buy once, are unlikely to produce good LTV. Wiland offers tools to identify such prospects and cull them from customer acquisition campaigns before promotional budget funds are wasted on them. These tools—which include Marketing Budget Optimization, Outside List Optimization, and Cross-Title Optimization—are discussed in a separate section. The key point is this: whatever the source of prospects, optimize them to eliminate waste, pushing the quality of your customer base up by reducing the portion of it comprised of consumers who will never produce good LTV.  

Enhanced Customer Lifecycle

Many Multichannel Retailers pursue the Three Prime Marketing Objectives with relative effectiveness. Nevertheless, many brands fail to optimize, and do not give enough consideration to how the objectives interrelate. Many new customers are desirable but some are not. They either cost too much to acquire or have little chance of producing good LTV, or both. A typical approach to new customer acquisition is illustrated in the Typical New Customer Lifecycle chart that follows.

 


The Typical New Customer Lifecycle chart may seem to depict the right way to do things, but in fact it depicts one of the most difficult dilemmas that marketers face: all too often marketers invest in new customers who never buy again or buy so seldom in such small amounts that they never produce good LTV. Losses incurred to acquire undesirable new customers plus additional losses resulting from marketing to them unsuccessfully as one-time buyers erode the bottom line significantly.    

Marketers often think about the "single buyer" dilemma in too narrow a context. They think about how to motivate single buyers to buy again. This is good. Marketers should never stop thinking about motivating single buyers to buy again. But they should also employ an enhanced approach to customer acquisition that reduces the incidence of acquiring expensive single buyers who will never buy again. The approach that we favor is illustrated in the Enhanced Customer Lifecycle chart that follows.



To fully understand Enhanced Customer Lifecycle as depicted in the chart, it is important to understand the ways in which losses are incurred in the marketing cycle and how to minimize them.  

New customers are typically acquired at some cost. The justification for losing money to obtain the first order is that subsequent orders will be profitable and thus compensate for the initial loss. Let's say, for example, that a brand receives a first order of $108 from a new customer. Marketing cost associated with the order is $71, while product and fulfillment (P&F) costs are $57. Thus, the brand "bought" the new customer for $20. If the new customer quickly orders again and spends $123, with a marketing cost of $3 and P&F cost of $65, the brand now has a $35 cumulative profit on the new customer—and an outlook for more orders to come. Such a scenario leads to highly satisfactory ROI.  

The future will be brighter when strategy and tactics reduce prospecting waste and a higher percentage of new customers produce good LTV.

But what if that new customer, acquired at a $20 cost, never buys again? In such a scenario, the brand may spend $15 or more on ongoing marketing to them before giving up. Now the loss on that new customer is $35, likely never to be recovered. There will always be some such cases because no one can predict future outcome perfectly. But the blend needs to improve: every brand should strive diligently to acquire more new customers who buy again and produce profits while acquiring fewer new customers who never buy again despite additional marketing investment.  

That's the point of the Enhanced Customer Lifecycle™ strategy: Acquire more desirable new customers that produce good LTV, while acquiring fewer new customers at a loss who never or seldom buy again. To the extent a merchant can actually accomplish this, it can add net margin points to its bottom line. It is worth taking time to understand the Enhanced Customer Lifecycle strategy and make changes to implement it. 

 Such changes may not be easy to implement, but savvy Multichannel Retailers should focus on this topic, and never stop focusing on it. The future will be brighter when strategy and tactics reduce prospecting waste and a higher percentage of new customers produce good LTV. Achieving this is the essence of effective new customer acquisition strategy. Multichannel Retailers must consider how to jump the hurdles and improve results rather than seeing the hurdles as insurmountable. And service providers need to get on board, determining how to achieve their own objectives in the context of producing optimized results for the client.  

One fundamental process that is critical for Multichannel Retailers seeking desirable new customers that produce good LTV is Marketing Budget Optimization. Enhanced Customer Lifecycle™ requires that Marketing Budget Optimization become standard. This often requires extra time in the merge/purge production cycle, between the time a net file is produced and when it is presorted. At Wiland, we typically perform the optimization in one business day, sometimes two. But what a difference a day makes. One new job step and results improve, sometimes dramatically.  

Wiland clients have seen significant improvements in overall customer acquisition results, both in the short and longer terms, as a result of utilizing our optimization services – even when the rental cost of dropped records from outside audiences is considered. In an ideal world, agencies and list brokers would be compensated differently than they are today, and acquisition sources, loath to admit that the audiences they provide contain poor prospects, would accept being paid only for names utilized. If and when this happens, Multichannel Retailers will see an even greater boost to customer LTV. In the meantime, Multichannel Retailers still benefit from the significant cost savings associated with not marketing to low-value prospects. We optimize audiences provided by others; and we are accepting of our audiences being optimized by others so long as we are part of the optimization mix. At the end of the day, whether a prospect audience originates from us or from a competent competitor, it ought to be optimized for the client's benefit.

Optimizing Acquisition Audiences

There are three primary sources of acquisition audiences:    

1.  Owned Prospect Audiences: The first primary acquisition source is owned audiences, sometimes referred to as "internal prospects" or "non-buyers" or "sister titles." Many Multichannel Retailers own multiple brands or possess various types of consumer audiences that have never purchased from a particular brand. These consumers may be buyers from another owned brand, catalog requestors, inquirers, gift recipients, website registrants, or some other type of non-buyer. There is a tendency for marketers to prefer owned prospects because they can be utilized "for free." This is good strategy if the owned prospect audiences produce desirable new customers. But some internal prospects do not produce desirable new customers. The graphic that follows illustrates a Typical Owned Prospect Audience Utilization Strategy™. 

 

 
While utilization of owned prospect audiences is prudent in many cases, it is possible to overuse them, putting a brand into what we refer to as a Weak Customers Failure Loop™ without the merchant even realizing it has happened. 

 

 


Just as it is true that owned audiences contain many good prospects, it is also true that they contain many prospects that, even if they do buy once, will either never buy again or will buy so seldom or in such small amounts that they harm the bottom line. To make matters worse, it is easy to produce large numbers of undesirable new customers from owned audiences because the cost to market to owned prospects is lower than that of outside prospects due to the absence of rental cost. Because there is no audience cost, marketers understandably utilize a lower initial performance requirement. This can make owned audiences look good initially even if they are producing undesirable new customers.  

Whether a prospect audience is owned or rented it is harmful to acquire undesirable new customers from it.

Even so, we encourage use of owned audiences. But whether a prospect audience is owned, rented, or exchanged, it is harmful to acquire undesirable new customers from it. 

Many marketers think of optimization as a tool for improving performance with rented audiences only. We recommend optimizing all acquisition sources, including rented audiences, owned audiences, and exchange audiences. An undesirable new customer is exactly that, undesirable, regardless of its source.  

Every brand should institute a carefully devised Owned Prospect Audience Optimization Strategy such as the one depicted in the graphic that follows.

 

 


In almost all cases, the impact of implementing Wiland's Owned Prospect Audience Optimization Strategy will be positive, with the following results being typical:  

  • The "Selected by the Brand and by Wiland" group will perform very well, and the resultant new customers will produce excellent LTV.  
  • High scoring "Selected Only by Wiland" segments will perform well. This actually increases the available universe of marketable owned prospects and delivers good LTV. Lower scoring segments of this group may perform poorly, and in such cases they should not continue to be utilized.  
  • High scoring "Selected Only by the Brand" segments may perform well too. Low scoring segments will perform poorly, and should be eliminated.  

The combination of these effects will be (1) lower marketing cost as a percent of sales and (2) higher quality new customers. Often there is a 3rd benefit: more total utilization of owned prospect audiences.  

For clients utilizing Wiland as its primary optimization provider, we will credit poor scoring segments at 100%, as long as the comparison is done by an independent service bureau not associated with a cooperative company. Thus, all of the savings from not marketing to weak segments is accretive to the brand's bottom line.  

Every brand should move into the future that Owned Prospect Optimization enables—eliminating wasted budget and increasing the long-term value of the new customers.    

2. Exchange Prospects: The second primary source of prospects for new customer acquisition campaigns is exchange prospects. In the direct mail channel, list exchange is a longstanding practice. In some markets, it is utilized extensively. Basically, two list owners agree to "trade lists for one-time use." Each provides an audience to the other for a particular marketing campaign, and in most cases each utilizes the entire audience provided. As long as performance is acceptable they continue the arrangement. Such arrangements are usually negotiated by a list brokerage firm that also keeps track of "exchange balances," which is simply the number of prospects each party owes the other.

The effect, except for whatever fee the marketer pays to the broker for placing the order, is that exchanges are "free," just like owned prospects. No payment is made for the audience. Thus, cost is lower and the performance requirement is lower.

As is the case with owned audiences, exchange audiences include both good prospects and bad prospects. Rarely are they optimized. The result is that marketers waste significant marketing budget marketing to poor prospects from exchange sources. Often the waste with exchanges is even higher than with owned prospects. We recently conducted an Exchange Performance Case Study in which we observed the performance of exchange audiences for 50 Wiland Multichannel Retail clients over the course of a year. The study revealed that an alarming 34% of all exchange audiences marketed by the 50 brands performed miserably. The marketers would have been much better off had they eliminated poor prospects in advance and either saved the marketing budget or re-deployed it to more productive audiences.

Marketers should not overlook the opportunity to improve their bottom line by optimizing exchange audiences. In addition to recovering marketing budget that can be redeployed elsewhere or kept as incremental profit, Exchange Audience Optimization will improve the quality of new customers, increasing repeat purchase rates and boosting ROI.

3.   Rented Prospect Audiences: Rented prospect audiences come from many sources. Audiences rented from companies that have a cooperative database represent the vast majority of prospecting volume for most Multichannel Retailers because they usually cost less and typically perform better. Nevertheless, some marketers also use rented "vertical" audiences, which typically have affinity to the marketer's brand or customers. Some brands also use compiled lists. 

Whatever the source, rented audiences should be optimized using theWiland Net File Optimization™ service to remove poor prospects that are buried in a hotline select or a model segment. For example, when a marketer rents a 90-day "hotline" selection from a vertical source, everyone that meets the recency criteria is provided. Some of the consumers provided are just right for the renting marketer, while others are poor prospects, such as people who just made their one and only direct purchase of the year. As our recent Rental Performance Case Study reveals, there is a significant amount of waste in most customer acquisition programs. This is largely attributable to the rental audience owner's lack of visibility into its customers' broader transactional behaviors that would enable it to provide more refined selections. With transactional data from over 2,500 brands, Wiland possesses such visibility, enabling us to effectively optimize any rental audience. 

Modeling for the Right Goals

Underlying all of the preceding topics relative to acquiring new customers is a very basic, fundamental question: When a predictive model is built for a multichannel retail brand, what should the modeler be attempting to predict? The answer seems so obvious to many industry professionals that the matter is seldom discussed. "We're trying to predict who will buy, right?" Well, yes, but that isn't the whole story, and leaving it at the simplistic "who will buy" level is a mistake. 

The best strategy is to market to people who will spend enough money to produce profit within a reasonable timeframe and who will remain profitable over the life of their relationship with the brand.

Responsiveness is an important attribute. Certainly a Multichannel Retailer needs to market to consumers who will buy. But merely marketing to people who will buy really isn't the best strategy. The best strategy is to market to people who will spend enough money to produce profit within a reasonable timeframe and who will remain profitable over the life of their relationship with the brand. Now, that's a lot more words than "who will buy," but it changes the game dramatically, and very favorably.  

Consider two scenarios for a catalog prospecting campaign:

A.  2.5% response rate and a $40 average order value  

B.  2.0% response rate and $60 average order value

Which is better? Some would choose Scenario A, arguing that a higher response rate and more new customers are better. Many have built good brands with this philosophy. But it isn't optimum. Scenario B, which produces a lower response rate but higher total revenue, is actually better for two reasons. First, higher revenue means lower marketing cost as percentage of sales and better short-term profitability. Second, the higher average order value means higher quality customers that will likely produce more future profit than the more numerous customers generated in Scenario A. While Scenario B is clearly superior (it produces more revenue and higher quality new customers), a surprising number of Multichannel Retailers base audience performance mostly or entirely on response rate.  

Here's another way to view the same subject. Many brands pay insufficient attention to the fact that there is usually an inverse relationship between response rate and order size. There are exceptions to every rule, but it is nonetheless true that high response rate is usually associated with low dollar amounts while high dollar amounts are usually associated with lower response rates. Thus, a marketing program focused on achieving high response rates may unintentionally be focused on finding low-value customers at the expense of finding high value customers.  

For a merchant selling a single item that is priced the same for everyone, there is no difference in modeling for response rate and modeling for dollars. Similarly, a low price-point brand that is selling items that almost anyone can afford may do very well by focusing on response rate. But even value oriented brands could benefit by shifting their modeling objective. Wouldn't a low price point brand rather have new customers who will buy two items at the same time rather than one, or three items rather than two, or once a year rather than once every two years?  

A lot of brands are underperforming today, and one of the reasons is that their entire marketing approach is predicated on the pursuit of the wrong goal: They are set on the notion that initial response rate is the most important measure and that maintaining or increasing the size of their 12-month buyer file is absolutely crucial. A growing number of recent customers is absolutely a good goal, but not if a large portion of the new customers acquired will never produce a profit.  

This problem goes far deeper than one may first imagine. Let's say, for example, that a marketing campaign brings in 40,000 new customers, 30,000 of which will forever be single buyers and 10,000 of which will buy again. All 40,000, the majority of which are weak, are added to the customer base. These 30,000 weak customers are promoted as hotline buyers, not only by the brand from which they purchased but also by other brands, consuming a huge amount of marketing budget. Perhaps worst of all, the brand's own merchandise analysis may be based more on the spending preferences of the unprofitable new-to-file customers than on the spending preferences of the ones who will produce good LTV. Wow, it's easy to really mess things up and not even know it happened!  

The solution is to model for the right goal, which isn't to merely "find people who will buy." Rather, to restate our opening point, it is to find people who will spend enough money to produce profit within a reasonable timeframe and remain profitable over the life of their relationship with the brand. The way to do this is to build models that predict customer spending, which is a combination of response rate and amount spent.   

A Perspective on Marketing Channels

Most of today's retail brands pursue customer acquisition in a multichannel environment, with determination of the appropriate mix of marketing channels being a key consideration. Many relatively new brands were born in the digital era and first acquired customers via viral marketing, search, affiliate marketing, or other digital channels. Multichannel retailers that grew up in the catalog world have historically relied on direct mail as the dominant new customer acquisition channel, and it still garners the majority of media expenditure for many brands. Regardless of when and how a brand was born, a diverse multichannel marketing mix is clearly emerging at many companies.

It is widely understood and acknowledged that digital channels are growing rapidly. What is not widely understood is the fact that direct mail remains substantial. The United States Postal Service classification "Advertising Mail" reached peak volume in 2007, at 103.5 billion pieces. During the recession that followed, Advertising Mail fell to 81.8 billion pieces by 2009. Since then, Advertising Mail volume has ticked up slightly. 

While direct mail remains huge, it is wise to view the recent stability of Advertising Mail volume from a historical perspective: 40 years ago, direct mail received almost the entire marketing budget at most multichannel retail brands.

While there certainly are exceptions, virtually every brand should be utilizing multiple marketing channels today. For those with a gigantic portion of their marketing budget dedicated to just one channel, testing in multiple channels likely should be a priority. Traditional direct mail marketers should be testing and learning the digital channels. Newer, digital driven companies should be learning direct mail. Wise marketers have abandoned the notion that some channels are new, cool, and tomorrow while others are old, boring, and yesterday. Instead, they are finding the right role for each channel in their marketing mix. Tomorrow's profits likely demand that Multichannel Retailers become proficient at marketing in both online and offline channels.

At Wiland, our goal is to enhance the productivity of the customer acquisition process, regardless of channel. We identify highly responsive offline and online audiences and enable clients to connect with them via carefully targeted direct mail and through the Wiland Digital Network™, featuring real-time bidding for display, video, mobile, and social media ads across dozens of networks. Each of our solutions is discussed in further detail in the section that follows. 

Prospect Audiences for Every Marketing Channel

Each of the Wiland Acquisition Audiences described below can be utilized to identify the right audience, regardless of the marketing channel in which the audience will be addressed:  

  • Digital Campaigns (display, video, mobile, social): Wiland provides the audience and manages the digital campaign. The Wiland Digital Network™ connects client offers to the target audience wherever a consumer may be online. The Wiland Digital Solutions™ team deploys the audience in an anonymized manner via our proprietary Wiland Tag™, recognizes the consumer, and bids in real time. Different consumers may be of different value. Many options are available for refining results, including: finding the target audience on specific web sites, avoiding specified types of web sites, varying bids by time of day, varying cadence (number of times an ad is served to the same consumer per minute, hour, day, or week), etc. A key performance metric is agreed upon and the Wiland team manages to that metric. Campaigns that are not tracking toward achieving the metric are slowed or stopped while refinement strategy is developed. All digital campaigns are measured carefully and accurately, employing a holdout methodology that proves the value of the digital campaign.  
  • Direct Mail Campaigns: Wiland provides the audience and, typically, ships it to a trusted service bureau that receives audiences from all sources before running a merge/purge to consolidate duplicates. The entire file of unduplicated consumers to be mailed should then come to us for optimization, as discussed above, to remove undesirable prospects. This process, discussed more fully in Marketing Budget Optimization™ for Multichannel Retailers, often not only reduces marketing cost as a percent of revenue but also increases total revenue. 

Often, digital campaigns and direct mail campaigns are coordinated. Testing has proven, for example, that a digital campaign directed to a direct mail audience (often beginning just prior to a scheduled in-home mail date) can not only pay for itself but increase response to the direct mail. Whatever the channel, the key is this: Wiland Acquisition Audiences are carefully developed and deployed with the objective of producing a large volume of desirable new customers who will produce good LTV.

Wiland Acquisition Audiences

The Wiland database houses billions of transactions from thousands of brands spanning many markets and industries, giving us unprecedented visibility into consumer spending and interests. Leveraging this information, we identify highly responsive prospect audiences and enable clients to connect with them in multiple channels. Designed to produce excellent initial response as well as high long-term value, our predictive models are built using multiple techniques, including a proprietary, pre-validated response modeling method developed and enhanced over the past 20 years.  

Wiland's Multichannel Retail clients utilize a variety of acquisition model types that fall into two primary categories: response models and profile models. Response models predict which consumers are likely to respond to a promotion, while profile models find audiences that "look like" the client's customers. Each major model type is described in further detail, followed by recommended guidelines for testing. 

Wiland Prospect Response Models

Wiland's prospect response models identify consumers in the Wiland database that resemble responders from the client's past acquisition campaigns. We analyze responders and non-responders from the client's promotion files to ascertain the important characteristics that differentiated them at the time of the promotion. We then build a model that predicts which consumers are likely to buy from a similar offer or promotion today. We build response models for both offline campaigns (e.g., direct mail) and digital campaigns (e.g., online display advertising). We typically study three distinct campaigns to create a single model.  

Response models frequently outperform profile models because they specifically identify prospect audiences that resemble past prospects who made a purchase from the merchant when promoted with a similar offer in the same marketing channel. Wiland clients utilize multiple response models that serve different purposes, including:  

Comprehensive Response Model™

Wiland's Comprehensive Response Model™ utilizes a proprietary algorithm and hundreds of industry, market, and brand-level variables to identify audiences with buying behaviors and characteristics that are similar to the client's direct responders. Validation across multiple promotions results in a stable model that identifies "pockets of success" often missed with traditional statistical techniques. This model requires analysis of three past prospect promotion files.  

Seasonal Comprehensive Response Model™

This is a special version of the Comprehensive Response Model™ that identifies consumers who resemble responders to season-oriented campaigns. This model is built using past promotions from a particular season. It is an excellent choice for marketers whose demand or product mix fluctuates from season to season.  

Core Response Model

Like the Comprehensive Response Model™, this model identifies prospect audiences that resemble direct responders, but through a different analytic technique that uncovers large, supplemental universes of highly responsive prospect names. The technique requires just one promotion file, making it a great choice for marketers with limited promotional history.  

Acquisition Response Model

Developed using a special technique designed to maximize response rate, this model identifies prospects on the Wiland Direct database that resemble past responders using response rate as the key driver, with per-contact revenue remaining an important but subordinate consideration. This model is recommended for acquisitive marketers for which high response rate and file growth are vital.  

Long-Term Value (LTV) Response Model

This model identifies prospects that most closely resemble the client's highest-value buyers by analyzing responders and non-responders from past promotions to their own customers. This model delivers a substantial prospect audience that will respond well initially and possess a strong tendency to buy again.  

Product Response Model

This model identifies prospect audiences that resemble responders from past prospecting efforts who purchased a specific product or group of products from the client. This model is recommended for offers focused on a particular product or core product line.  

Repeat Buyers Response Model

This model focuses on individuals who bought as a result of a new customer acquisition campaign and made at least one subsequent purchase in the ensuing year. It tends to identify prospect audiences that produce customers who quickly produce good long-term value.  

Wiland Prospect Profile Models

Wiland's prospect profile models find responsive consumer audiences that resemble desirable segments of the client's customer file. "Desirable" may be defined differently for different clients, but as an example, customers might be considered desirable for profile modeling purposes if they have these three characteristics: purchased at least once in the past 12 months; purchased at least twice in the past 24 months; and, spent a total in the past 12 months that is greater than an average order. However they are defined, we analyze the key traits of desirable customers; find the characteristics that best differentiate them from other consumers in the Wiland database, and build models to identify prospects with similar characteristics and buying behaviors. The concept is that prospects who "look like" the client's desirable buyers will be highly responsive and produce good LTV. Repeated past success proves that our profile models work. Among the profile models being successfully used by Wiland clients are:  

Best Customer Model

This model draws upon the vast data in the Wiland database to find consumers that most closely resemble the client's best customers. This model utilizes numerous industry, market, and cluster-level variables to pinpoint prospects that share transactional, demographic and lifestyle characteristics with the client's core customer, making them excellent candidates to respond.  

Brand Affinity Model

This model utilizes advanced techniques to identify the best prospects from highly correlated brands, which are chosen based on merchandise offering, creative presentation, and statistically observed synergy with the client's customer base. The model employs data from both synergistic brands as well as the overall database to pinpoint prospects most likely to buy.  

Comprehensive Correlation Model™

This model is built on the same proprietary modeling platform as the Comprehensive Response Model™ using recent new-to-file customers as input. Customers are compared to the overall database to identify prospects whose characteristics resemble those of the client's newest buyers. The model uses hundreds of powerful cross-industry variables and is validated across multiple data sets, resulting in excellent performance and stability.   

Enhanced Prospect Solutions

In addition to our response and profile models, Wiland offers other highly productive customer acquisition solutions. The Wiland Integrated Performance Series™ (WIPS) utilizes custom algorithms that assign a composite score to each consumer in our database based on ranking across multiple models, revealing an integrated universe of the most responsive names for the campaign. This methodology offers three proven approaches to new customer acquisition.  

Universal Performance

Perfect for higher-volume marketers, Wiland builds a series of powerful models, which are brought together to identify a single universe of the very best prospect names for client acquisition efforts. Models are weighted by predicted performance, and customers are given a single rank based on scoring across all models—simplifying the ordering process and resulting in large, high-performing segments that are easy to test and read.  

Incremental Depth

For marketers already taking large volumes from multiple Wiland models, we will rank customers based on scoring across all models, enabling clients to target incremental names that are likely to perform well. Ideal as a "best of the rest" approach to supplement already successful Wiland models, this solution requires integration of at least three models beyond those being used extensively.  

Dynamic Interaction

With this approach, we score consumers in the database using two models and make selections based on the intersection of model segments. This popular selection methodology produces excellent results by enabling Multichannel Retailers to target their ideal audience based on two separate dependencies, such as product affinity and likelihood to respond.

Acquisition Model Testing Strategies

The most successful users of Wiland prospect solutions test multiple model types and follow several best practices:  

Best Practice 1: Provide the best possible research files. Typically, this means at least three audience files from past marketing. The more research files we have, and the better we understand the files and the offers associated with them the better our models should perform.  

The most successful users of Wiland prospect solutions test multiple response model types, and follow several best practices.

Best Practice 2: Test several models types. This includes both response models and profile models. If test budget is limited then test at least two.  

Best Practice 3: Test more than Segment 1. Test at least Segments 1 and 2, and ideally Segment 3 and beyond. Clients often succeed well beyond Segment 1. If test budget is limited, test just Segment 1, but when it works, test deeper.  

Best Practice 4: Don't over-suppress. Best practices concerning Prior Order Omission Strategy and Customer Suppression are thoroughly discussed in separate articles. In general, the goals are:  

  • No prior prospect suppression in digital campaigns.
  • In direct mail campaigns, suppress prior prospects only from campaigns that will be less than 80% complete on the in-home date of the planned campaign.
  • Suppress only current customers that will actually be marketed to in the planned campaign.  

Best Practice 5: Conduct a depth test. The depth test is meant to determine how deep into a model universe the brand can market successfully. How deep to test? We recommend that clients utilize Nth Cross Sections across model segments, testing deep enough that if the depth test is successful the universe could be 50% to 100% of total customer acquisition volume. Thus, if a brand typically markets to 1,000,000 prospects, it should test to a depth of between 500,000 and 1,000,000. The purpose of depth testing is not for Wiland to become the client's sole prospecting source. Rather, there are two purposes. First, a large depth test will lead to determination of the depth at which the client can successfully market, enabling the brand to maximize audience size and increase overall success in obtaining desirable new customers. Second, after the depth test has been conducted there will be a large universe of buyers produced by the depth test that can be used for research to refine the model, leading to better results in subsequent campaigns. A brand seeking excellent performance on large volumes would conduct a very large depth test, anticipating that the first such test may fall short of key performance metrics for success.  

A second depth test is then conducted using the refined model. Iterative depth testing until a model is fully optimized almost always leads to success at a higher volume than would otherwise be possible. This "test deep and refine" methodology, for brands who stick with it over several promotion cycles, is what leads to the largest successful volume, fueling brand growth and an improved bottom line.  

One way a brand seeking excellent performance on large volumes could proceed is to conduct a depth test derived from a Wiland Universal Performance (UP) model. In such a scenario, the UP model will be constructed from 6 to 12 underlying models. After initial depth testing, each underlying model is back-tested to determine its merit in the overall solution. Under performing models are replaced with new models to solidify the overall UP solution.  

The "test deep and refine" methodology, for brands who stick with it over several promotion cycles, almost always leads to the largest successful volume and thus fuels brand growth and an improved bottom line.    

Best Practice 6: Never stop testing. The Wiland database is huge. Virtually every adult in America is in it. So it can't get much bigger. But…it gets deeper constantly. We add multiple new clients every week. We post new transactions from over 2,500 clients regularly. We develop new techniques, and new variables, and add overlay data sources. As a result, unless the client's competitive position or offer weakens we are almost always able to produce improved results as time passes. Never stop testing.  

Best Practice 7: Test Multichannel. Direct mail. Web site retargeting. Digital new customer acquisition. Direct mail co-targeting. Digital customer campaigns. Measured Brand Advertising™. It all works. Not every time, but usually. Test, and learn. Find breakthroughs.  

Wiland Prospect Optimization Solutions

The methodologies and models described thus far are utilized by our clients to create productive customer acquisition audiences. But our success is more about data than about modeling techniques. The vast Wiland cooperative database, which includes virtually every direct responsive person in the United States, is unsurpassed in breadth and depth. As a result, the audiences we create are sufficient to meet the entirety of a brand's prospecting needs. But even if Wiland can provide 100% of a client's needs, wise marketers will reach beyond, seeking additional audiences. No marketer should put all of its eggs in one basket. If a brand relies solely on Wiland (or anyone) for prospect audiences it loses the ability to compare. How would such a brand ever know whether they are using the best sources? Multiple providers should always be used, even if one is the vast majority. Brands need comparisons, and should take great care to assure that the comparisons are valid, rather than prioritized in a manner that essentially pre-determines the winner. 

Thus, because it is what is best for our clients, we encourage the use of supplemental prospect audiences, from sources in addition to Wiland. Often they contain excellent prospects. Test, and find the best sources. But, whatever the source, prospect audiences almost always intermingle good prospects and poor prospects. To achieve success with these audiences, Wiland provides several optimization services that separate the good from the bad, enabling clients to market only to strong prospects that will deliver good long-term value. 

Outside Audience Optimization

Multichannel Retailers sometimes utilize an audience source that performs marginally or inconsistently. It may be a rented or exchanged outside list. It may be a list of online inquirers. It may be a sister brand, owned by the same company. Or it may be a compiled list. Whatever the source, Wiland can optimize it by identifying responsive segments and, equally important, informing the client which portions of the list would likely represent a poor investment from a marketing ROI perspective. You will find additional information about Outside Audience Optimization here.

Net File Optimization

Don't waste marketing budget on bad prospects just because they happen to survive the merge/purge process.

Most Multichannel Retailers procure lists or audiences from multiple sources and run them through a merge/purge process to standardize records and eliminate duplicates. The resulting "net file" is sent to a printer if it is a direct mail campaign or to some other service provider with digital advertising, email, or telemarketing campaigns. Net files tend to have one thing in common: they contain individuals who do not merit the marketing investment required to promote to them. These poor prospects are mixed in with the good ones, blocking the marketer's view of them. Our studies indicate that the portion of a net file that is not worth promoting varies widely, from a low of 5% to a high of 50% or more with 10% to 25% being typical. 

To improve overall campaign performance, Wiland can optimize the merge/purge net file using a custom-built, client-specific model that leverages all the data in the Wiland database. We will separate the good from the bad, so that marketing dollars that would have been spent on people who won't respond sufficiently can be kept as incremental profit or redeployed in a more productive manner via Wiland Digital Solutions™ or some other marketing program. Once we have ranked the names on the net file, the client has several options:

1.     Drop them and replace with higher-scoring Wiland names

2.     Drop them and keep the savings as profit

3.     Redeploy the saved budget to a Wiland-managed digital campaign that reaches an even larger audience

4.     Redeploy the budget elsewhere 

Whatever combination of optimization options is chosen, the key is this: don't waste marketing budget on bad prospects just because they happen to survive the merge/purge process. 

Balance Modeling

Sometimes, particularly in the direct mail channel, a Multichannel Retailer needs to achieve a specific audience volume. Let's say, for example, that a merchant purchased two million envelopes and printed two million letters and other insertion materials. The merchant chose prospect audiences and customer segments for the mailing that, based on experience, it anticipated would net out to two million names. But it doesn't turn out as expected. The merge/purge net file is significantly larger or smaller than two million. What to do?  

If a client falls short of its target mail quantity upon completion of a merge/purge, Wiland can supplement the net file with modeled "balance" names likely to perform well—and leave campaign volume perfectly matched with available materials. If there are too many names on the merge/purge net file Wiland can optimize it to the specific desired quantity (see Marketing Budget Optimization™ for Multichannel Retailers).  

Other Key Prospecting Considerations

The foregoing discussions of audiences and optimization are crucial, but they are not the end of the new customer acquisition subject. Clearly "picking the right audiences and optimizing them" is fundamental. But after that is done there remain other key considerations. 

Suppression of Previously Promoted Audiences

Should prior audiences be suppressed in the current marketing campaigns? Should owned databases be treated as suppression files in acquisition campaigns? If so, what are the precise suppression criteria? Should every audience ever used be suppressed or just recent audiences, and what does "recent" mean? Should all owned names be suppressed, or only those that have high affinity with the brand being promoted…or maybe just those scheduled for use in an upcoming campaign? These are important questions, and all too often they are answered quickly, without enough thought.  

On the surface the answers may seem obvious, even simple. There is no point in marketing to people who have failed to respond in the past, so suppress past audiences; and there is no point in paying for a rental name that is already owned by the company, so suppress all owned audiences as well. Such points of view are understandable. They may seem wise on the surface. But in fact, they are seldom correct.  

Let's say there is an upcoming campaign for which the audience is about 2 million consumers. Just a short time ago there was a campaign that addressed 2.5 million consumers. Should the recent 2.5 million prospect audience be suppressed from the current campaign? In practice, marketers deal with this subject in a variety of ways. For example, one marketer maintains a database of everyone it has ever contacted with a prospect offer, tracking number of contacts with each. After contacting a consumer three or more times without a response the marketer suppresses the consumer forever (and wonders why there are not enough prospects). In contrast, some marketers never suppress any prior audiences. They want to maximize the size of prospect audiences and believe every new campaign exists in different circumstances (different economy, different season, etc.), so consumers will behave differently. In most cases neither of these extremes is the optimum view.  

"Optimum" may differ from brand to brand, but there is a strategic question that should be answered first: Fundamentally, what is the purpose of prior audience suppression? Most would answer that the purpose is to omit people who are unlikely to respond. That makes sense, but it is highly unlikely that every previously promoted consumer who did not respond is a bad prospect forever. If that were the case, every marketer would quickly run out of prospects to promote! That can't be the right answer.  

While the optimum practice concerning prior audience suppression may vary by marketer, two closely related principles are always true:    

1. Contact frequency should be optimized. Just as suppressing everyone who has been contacted but did not buy is unwise, so it would be unwise to mail every prospect a direct mail offer every day or serve them a display ad every second. If a consumer is contacted too often, it will not merely annoy them. It will also depress response and thus depress ROI. Conversely, if a consumer is not contacted often enough, the brand is "out of sight, out of mind," and as a result, the consumer will be unlikely to think of the brand when a need for merchandise that the brand offers arises. "Not too often, but not too seldom" is the principle. But what is too often, and what is too seldom?  

The optimum interval between exposures to the same consumer varies by channel. It may be 16 catalogs per year, 32 email messages per year, and one display ad per hour. Or it may be some variation of these intervals. A new brand should initially adopt and adhere to a contact frequency policy by channel that makes sense for the brand based on experience and good judgment, followed by careful testing of the policy. The number of contacts should then be increased or decreased based on ongoing testing that eventually leads to optimum frequency.  

As a rule, more frequent contact reduces revenue per contact but increases total revenue (and thus, increases marketing cost as a percent of revenue). If bottom line performance per contact is excellent then more contacts is usually a good strategy. How many more contacts? Usually, the right answer is: Don't shift contact frequency radically. Make minor shifts, and live with the shift for a while to determine effect. Thus, a catalog that is producing 14 editions annually would not typically shift to six or 20 editions. Rather, it might shift to 15 or 16 if bottom line profit is strong or to 12 or 13 if results are profitable but weak. Similarly modest changes in contact frequency in other channels likely are appropriate unless results are so incredibly good or terribly bad that more rapid change is in order.    

2. Campaign completion patterns should be well understood. Campaign completion curves show the amount of time that elapses from the point when a campaign begins until the campaign is substantially complete (i.e., the point when the campaign has produced most of the orders it will ever produce.) There are several reasons that a brand should understand its campaign completion patterns.

One reason is that most direct merchandising brands have peaks and valleys in order volume. Peak volume periods may strain a brand's order fulfillment operations. Valley periods are often times when the brand loses money: It is difficult to keep order fulfillment people busy; admin costs rise as a percent of sales; etc. Thus, valley periods harm the bottom line. Campaign timing should be based primarily on when demand will be strongest, not on operational considerations, but operational impact should be understood and, if practical, marketing staff should devise campaigns that supplement order volume during valley periods. But as a practical matter it is difficult to insert an extra campaign into a valley period. It will unquestionably increase revenue, but it may also harm results for both the previous campaign and the subsequent one; and, valley periods are sometimes low demand periods, making it very difficult to find success at such times.  

A primary reason a brand should understand its campaign completion patterns has to do with "clipping tails." Contemplate a marketing campaign, called Campaign 1, that is projected to produce $1 million in revenue. Then imagine that when only 30% of Campaign 1 orders have been received the brand initiates Campaign 2, addressing the same audience with a similar offer. Whatever actual revenue Campaign 1 was going to produce will now be reduced because Campaign 2 has "clipped its tail." Initiating Campaign 2 when Campaign 1 was only 30% complete was almost certainly unwise. But how long should Campaign 2 be delayed? While our Wiland Campaign Completion Analysis™ tool won't make this decision for a Multichannel Retailer, it will inform the decision maker and facilitate good campaign intervals.  

With good analysis in hand, marketing staff should set policy concerning campaign intervals. A good campaign interval policy might be something like this: "No major marketing campaign will be initiated until the most recent prior major campaign to a similar audience is estimated to have produced over 80% of its total projected eventual revenue, except during the holiday season when the 80% factor is reduced to 50%." The point here is not to argue for specific interval percentages. Rather, it is to encourage the marketer to do two things: study campaign completion cycles and adopt a good campaign interval policy.



Completion reports such as the one above are a subset of Wiland Insights™. The client simply provides the name of each campaign, its initiation date, and associated key codes.

A different perspective on campaign completion is provided in Wiland's Marketing Campaign Completion Cycle – Daily View™ reports, as shown below. This chart tracks the percentage of eventual total revenue produced by day since first order date (or hours, for email campaigns).



In most cases, campaign completion occurs in what might be described as an elongated bell curve. The cycle starts fast, roars up to a top in a pattern that is somewhat like the left side of a bell curve, falls in a manner that suggests a bell curve is forming, but then elongates. Some (but few) orders come a year later, or longer. The exact point when the bulk of orders have been received varies depending on offer, season, and contact frequency. For example, a brand offering holiday items in December to an audience that it contacts every three weeks will reach 90% completion very quickly. In contrast, a marketer of consumable supplies who contacts customers every nine weeks should expect a much longer completion cycle. The key here is to encourage every brand to understand its own completion cycles and take them into account when making prior order omission decisions. Some amount of "tail clipping" is inevitable because the very last order from a campaign may not occur until long after additional campaigns have been initiated. 

Armed with good knowledge of campaign completion patterns and how they differ from season to season, the marketing team should strategize concerning both contact strategy and prior audience suppression. Let's say, for example, that the brand currently has an interval of approximately 60 days, but campaigns are 85% to 90% complete in 30 days. This suggests that increasing contact frequency may be wise (more contacts to fill in the 30-60 day valley). On the other hand, if the brand has an average interval of 22 days and campaigns are 98% complete in 21 days it is possible that each successive campaign is excessively clipping the tail of the previous campaign. This may suggest that suppression of one prior campaign is in order, or that reduced contact frequency would improve the bottom line, or both. 

Arriving at the absolutely correct answer concerning prior order suppression is an ongoing pursuit because there are many variables, and the environment changes. In general, the following principles usually apply: 

A.      Prospect audiences from immature campaigns (e.g., those that will not have produced 80% of their eventual total by the time a subsequent campaign is initiated) should be suppressed from the audience for the subsequent prospect campaign. 

B.     No other prospect campaigns should be suppressed. None. A campaign that is 85% or 90% complete simply should not be used as suppression. The likely result would be elimination of many of the best prospects available; the exact opposite of what is optimum. Yet, many multichannel merchants, to keep the suppression decision simple, adopt a policy such as "always suppress the past three campaigns." Policies such as these, which lack consideration of completion cycles, are most likely damaging the brands that employ them. 

Some may argue that the 80% completion guideline proposed here will result in too little suppression. They may believe that this approach to suppression will lead to overexposure of a brand to prospects. Our long and varied experience suggests that this is rarely the case. Many times we have supplied prospect audiences previously provided for campaigns that are over 80% complete, and tracked them independently in order to empirically determine whether they should be suppressed. There are exceptions, but these prospects almost always produce above average results. Why would this be the case? Why would a group of prospects, all of whom saw an offer from the brand recently but did not respond, perform above average? The reason is simple: they passed the modeling screen in the last campaign AND they passed the modeling screen again, recently. They are among the brand's best prospects. 

If a brand is not accepting of the "only suppress immature campaigns" tactic, they should, at a minimum, avoid the extreme of suppressing numerous campaigns. At most, they should raise the maturity definition (to "85% complete," for example). Otherwise, suppression simply serves to prevent marketing to excellent prospects. 

A key related point is that suppression is often implemented in multiple ways. For example, a list broker may request that a list provider omit certain prior orders. Sometimes brokers do this based on their own opinion or their understanding of client preference, without discussing it with their client. The result can be excessive suppression, harming the brand. Also, the brand marketing team may instruct a service bureau to utilize certain prior audiences for suppression during merge/purge. Even if audiences have been ordered based on a wise suppression policy it can be trumped by an unwise merge/purge suppression policy. 

Finally, a marketer may give suppression and prior order omission instructions to each cooperative database from which they order. Too often, they fail to think through all the implications and establish a policy on such topics. Even if they have a policy, they sometimes fail to make sure it is implemented universally by everyone involved. They may, for example, decide to only suppress one immature campaign, or not to do any prior suppression because all prior campaigns will be sufficiently mature by the time a new campaign is initiated. But they will fail to inform the list broker who places the cooperative database order, so the broker requests prior order omission. The prior order may have been ten days ago or six months ago. If it is longer ago than the 80% campaign maturity cycle then suppression of prior orders is almost always a mistake. 

Every brand will surely benefit from the establishment of a clear policy concerning prior audience suppression, and which is universally adhered to for every campaign by all of the parties involved. 

Suppression of Customers & Owned Prospects

In addition to setting policy concerning suppression of prior prospect campaigns, a brand must also give attention to the question of which owned prospects should be suppressed from prospect audiences, if any. There is much variety in how this is handled. One company with 85 million customers and former customers across several brands insisted that the only prospects they would rent were consumers not among the 85 million owned records. Guess what! There wasn't much productive prospect audience left after suppressing everyone who spent a nickel in the past twenty years with any brand of the company. Such extreme suppression of owned prospects certainly is not the right answer. But renting prospects that a brand would market anyway, without rental cost, doesn't make sense either. 

The right answer is in the phrase "would market anyway." The mere existence of a prospect in a company's database does not qualify it for marketing. Imagine, for example, a multi-brand direct merchandiser choosing in-house prospects for an upscale merchandise offer. It would probably not select someone who ordered once, 49 months ago, from a different brand, and bought just one inexpensive item. That being the case, why suppress them if a cooperative database model identifies them as a recent upscale merchandise buyer in a highly correlated product category? 

We believe the right policy is to suppress only those owned prospects that a brand actually selects internally to be marketed in an upcoming campaign. Otherwise, the brand will unwisely suppress good prospects that they won't select internally simply because they have inadequate data. 

Multibuyer Allocation

Marketing professionals in Multichannel Retail companies wear many hats. One of them is the "keep audience cost down" hat. Another is the "make sure I use the best audiences and produce strong ROI" hat. Along the way, as they try to wear all of their hats at the same time, they are faced with a decision that has to be made for every campaign: When the same prospect is supplied by more than one source for a campaign, which source should be "credited" for having supplied the prospect? Several approaches are used:    

1. Proportional Allocation. Proportional allocation gives each source half credit for a two-source prospect, one-third credit for a three-source prospect, etc. Proportional Allocation is the right way to deal with multibuyer allocation if the marketer wants to thoroughly understand actual performance by source. And understanding performance by source is fundamental.    

2. Audience Priority. Audience priority allocation rank orders the sources. Source #1 gets credit for every prospect it supplies, even if one or several other sources supply the same prospect. There are understandable reasons a marketer might conclude that audience priority is a good method. The top priority source may be its largest source, and the marketer concludes that there is no point in using the other sources unless they are incremental. Or, the top priority source may be the source that costs the least, the logic being that there is no point in using more expensive sources unless they are incremental. This may sound good, but it has one major, fatal flaw: Audience priority virtually guarantees that the #1 priority source will produce the best result because it gets credit for 100% of the multi-source prospects, which are almost always the best prospects. Thus, even if the 2nd or lower priority source is actually superior, the marketer never knows it. The top source gets all the credit, gets paid the most, and looks great even if other sources are actually superior.    

3. House Credit + Proportional Allocation. Some marketers input far more owned prospects into merge/purge than they could ever justify mailing in an effort to achieve matches with outside sources. They credit all the matches to owned prospects, proportionally allocate the remainder, and deduct payment on rented audiences for house file matches. This is a highly questionable practice, but it is done. It makes a lot of sense to enter a large volume of owned prospects into merge/purge so that the net file can be optimized, and poor prospects dropped. But the fact that they are "in the merge" does not justify non-payment to the source that actually knew it was a good prospect. The best practice is to enter owned audiences into merge/purge in two groups: the ones to be marketed regardless of matching other sources (make these top priority for crediting); and those the brand won't market to unless they match an outside source or are optimized (don't give these priority).  

There are many variations of these three approaches to multibuyer allocation. Smart marketers, brokers, agencies, service bureaus, cooperative databases, and others differ on the pros and cons of the various allocation methodologies. But this much is clear: It matters. The way allocation is done significantly impacts outcome. Put the best source in 39th priority and watch it fail. Put a mediocre source in top priority and it may look like the best source. Clearly a marketer can obtain some advantages by allocating in a manner other than proportional. It can reduce audience cost in that campaign by enabling deductions for house name matches or by giving top priority to the lowest cost source. But in the long-run, obscuring the actual performance of acquisition sources does not serve the marketer's best interests.  

Conclusion

Whew. That's a lot of stuff about new customer acquisition. And here's the hard part: A top marketing pro on a Multichannel Retailer's brand team needs to keep it all in his or her head all the time. Or at the very least, they must set aside time to think about all of it periodically and honestly assess whether the brand is on top of things. A strong brand with a great offer may be able to stay in the black without paying close attention to all the new customer acquisition topics discussed here, but it likely cannot optimize its bottom line without doing so. 

In today's highly competitive environment, pre-tax net margin for many Multichannel Retailers is in the low single digits, breakeven, or slightly negative. These brands need to improve. Even brands with higher pre-tax profit margin may not be optimized. Moving into the future by carefully considering all of the topics discussed here can add margin points for virtually every brand; maybe just one point; but perhaps several. Regardless, we invite you to move into the future with Wiland. We deliver superior new customer acquisition methodologies that will improve your bottom line if you utilize them properly.  

We're Here to Help

Multichannel retailing is a vital component of our culture. Brands connect consumers to products they care about and that fulfill their basic physical, emotional and spiritual needs. They connect people to other parts of the country and parts of the world that they may never visit in person. In addition to providing a physical product or service, brands teach, inform and inspire.  

We are committed to helping Multichannel Retailers thrive in today's challenging and evolving marketplace by improving their customer base and supporting their brands. Through the identification of prospect audiences that generate superior long-term value, we contribute to the viability and vitality of every brand we serve.  

Contact us today to learn more about our entire suite of customer acquisition solutions and other services for multichannel retailers. We are here to help.

Contact Us