Profiling, Targeting and Segmentation

By: John M. Coe, President
The Sales & Marketing Institute, Inc.

Executive Summary:

In marketing, as in many other disciplines, the first question is,"where do I start to solve the problem?" Frequently, marketers start where their strengths lie or feel comfortable. Database marketers start with data, creative types start with the creative, etc. Frankly, that's why so much marketing misses the mark and does not produce the desired results. Years ago, I was watching Ron O'Brien (a two-time NCAA diving champion and coach of many of our Olympic divers including Greg Louganis) learn a new dive. After his many attempts to"hit" a 3 ½ forward somersault with 2 twists, I asked,"how do you learn a new dive?" His answer was both insightful and appropriate, and one that I've taken to heart – Ron said,"If I put my head in the right position, the body soon follows."

When we develop marketing campaigns, we frequently miss the mark and should take Ron's advice. But what is our head position? My advice to marketers is that profiling and targeting combined with a segmentation process is it, and if we get it right, our marketing"body" will soon follow – and do it much more easily.

The reason I believe this process is so important is that in B2B, much marketing planning and execution is done with little solid information and/or according to anecdotal input. It's not uncommon to have a company"identify" a target market and launch a program based on only several sales successes. Everyone gets excited and charges off to call on more of these"kinds of companies" without further analyzing the overall market to see if, in fact, that is a segment worthy enough to target. As we move from sales-driven strategies to more integrated ones, there needs to be a much more disciplined method of targeting and segmentation to achieve the best results. Overall, it's a three-step process.

• Profiling – Where have you been?
• Targeting – Where are you going?
• Segmentation – Who's going to get you there?

This white paper deals with profiling, targeting and segmentation, which are the three critical first steps in any marketing program, and the required"head position" to achieve maximum results. Companies that have not yet undertaken this three-step process may well find that it's a difficult, but also a very rewarding effort.

Within the B2B market, there are a vast array of products and services that each have their unique characteristics from commodities to complex and engineered solutions. None of these situations are beyond using these three process to improve the overall marketing strategies and tactics leading to improved results.

Profiling: Where Are You Now?

The first step, profiling, not only confirms what you know about your market, but also adds insight regarding where you have had success in the past, in a more quantitative manner. Once the profiling analysis has been completed, it can be used as one of the inputs to target future marketing efforts.

Anecdotal examples of success almost always come from sales. While most companies have a strong sense of not only what markets are good targets, but also where success has been achieved in the past, in almost all cases in which the"insight" on markets exists, it has not been well defined and quantified. Too often I hear of a new customer who is in the"XYZ" industry, and find that a major marketing and sales initiative was launched to find similar customers, without any marketing analysis being conducted. This is typical for a sales-driven company. For many years marketing's job (frequently called Marcom) was to create advertisements, develop brochures, build trade show booths, and perform similar duties. The mission was not to define, for the sales group, anything about the market and/or customer base, and where sales calls should be directed. This was the job of product and sales management, and frequently, the determination of who were the potential customers was left solely to the salesperson covering the territory.

Profiling: The Process

Profiling should be a basic marketing process that is performed routinely by all companies. Unfortunately, my estimate is that fewer than 20 percent of B2B firms have actually profiled their customer and/or prospect base. In fact, recently I was consulting with a very successful software company on its marketing programs, and found that this twenty-year-old high-tech company with more that sixteen thousand sales transactions had never profiled its marketplace. Here's how the marketing communications manager answered the"where do you sell?" question:"We sell large Fortune 500 companies and government agencies." That was it; no other detail on the makeup of the market was forthcoming. Management didn't even know how many of the sales came from the same company, since the sales records were by location and were not tied to a corporation or parent company. As a result, the marketing objective was to sell more"large deals", but nobody knew where these"deals" were, as no real market analysis and profile had been developed. Therefore this high-tech company was flying blind, except to direct the sales organization to"large companies. Needless to say, that company is now engaged in a detailed profiling project to identify the best industry segments.

So, where do you start the profiling process? First and foremost, the customers need to be profiled, for the obvious reason that past success is almost always a predictor of future success. So then the question arises – on what do we base the profile? In my experience, two demographic pillars support the basic profile, and they are industry type and size of company.

While there are other profiling approaches (e.g., sales revenue or some other key characteristic), this basic demographic profiling is a required first step. The support behind this approach is that these two data descriptions are also the"data bridges" to outside databases from business information compilers. We need these data bridges to find companies that match the profile in these compiled databases. It doesn't do any good to develop a great customer profile, and then find that no other data source has this information. You become"marketing-locked" if this happens. Once this demographic profile is established, other, more company-relevant, profiling approaches are insightful and actionable.

Industry Type

The standard definition of industry type has been the Standard Industrial Classification or SIC for the past seventy years. In 2000 this definition became a legacy system, as we are transitioning to the North American Industrial Classification System, or NAICS code. The newer NAICS coding has several major advantages over the old SIC coding:

• It includes the new technology companies that were not specified in the old SIC codes. For example, Cisco did not have a specific SIC code that described their company's products, and was assigned the famous 99 NEC or Not Elsewhere Classified, number.
• The new coding system is NAFTA-consistent and aligns with both Canadian and Mexican coding systems. It is also the basis of an initiative by the Department of Commerce to establish a worldwide coding system that, as of this time, is not yet in place.

There are other improvements and revisions to this new system of classifying companies. To keep updated on the transition, visit www.census.gov/eped/www/naics.html Note that even though the government has officially changed to the new NAICS coding, the industry has not. At this time most companies, who have coded their database with the SIC code, have not taken the time and expense to re-code to NAICS.

Therefore, as market demand has not caught up with capabilities, the transition is not as widespread as would be expected. However, if you are now considering which coding system to use, my recommendation is that the newer NAICS code would be better, even though the transition in the industry has not been uniform. The transition to NAICS is available from data compiling firms, as they have build"data bridges" from the old SIC codes to NAICS.

Once you understand the industry coding system and what it provides, several other issues need attention before you proceed. First, how deep into the industry description should you go to define the company type? In the old SIC system, a four-digit level fit most situations, which equates to the six-digit level under the new NAICS codes. This decision, in part, rests on how many customers you are profiling. Many B2B companies only sell to hundreds of companies, in which case, going to a six-digit NAICS level will most likely be too granular and not provide the insight necessary to identify the best industry concentrations or clusters. On the other hand, if the need is to define your customers in as detailed description as possible, then the six-digit level is best. This is a decision that needs to be made before a profiling process in undertaken.

In addition, a customer base that is composed of large companies requires further consideration, since they will most likely have numerous NAICS codes to reflect multiple operations and product lines. Just think of how many NAICS codes IBM or DuPont have that describe their business units. On this issue, the site or location of the customer may be a tip-off as to what code should be assigned – but not always. Because this is such an important piece of data, in some cases a manual check may have to be made of the specific code assigned to the customer. Incorrect coding will lead to faulty profile analysis. The good news is that companies' NAICS codes rarely change or decay over time, so once you've accurately recorded the numbers, you can rely on them. Determining the right NAICS code, and assigning it may be a daunting challenge if the customer base numbers in the thousands, but it is worth the effort, as this basic demographic piece of data will be used over and over again to set marketing strategy and sales direction.

Company Size

The next key issue is how to define the size of the companies to be profiled. To most people, this means sales revenue. While this definition may be the most desired, it poses several problems for accurate profiling. Here's why. Most companies in the U.S. are private. In fact, only about 10,000 of the more than 12 million companies in the United States are public, and are required to file factual information on their sales revenue that the public can access. This means that at least 99 percent of all U.S. companies keep their sales figures confidential. Don't get me wrong: revenue numbers for private companies exist in compiled databases, but they are often estimates, modeled on some other factor, such as number of employees at the company and nature of the industry.

Therefore, to determine the size of the company, the number of employees may be a better gauge. This information is usually not considered secret. In public databases, it is often reported within a range of employee sizes.

When developing your own information, it would thus be satisfactory to also use ranges, as exact employee size is not as important as the general size of the company. The standard employee size range breakdown is as follows:

• 1-4 employees 100 - 249 employees
• 5-9 employees 250 - 999 employees
• 10-24 employees 1,000 – 2,499 employees
• 25-49 employees 2,500 – 9,999 employees
• 50-99 employees 10,000+ employees

Unlike the NAICS code, the size of a company often changes from year to year, and therefore becomes a data element that needs continual updating particular if you are selling small to medium sized businesses. Nevertheless, I prefer using the employee count for the company size criterion, since this question will most likely be answered fairly accurately. While data compilers do update revenue number and employee size definition in their databases, the question regarding number of employees is one that we also can ask our customers and prospects and receive a reasonably accurate answer to update our own records. The sales, customer service, and telemarketing groups or other types of customer facing activities can include this question, and therefore, we don't have to solely rely on data compilers for an update.

Another choice on size of company should be made at the outset of profiling process. While the preceding employee size categories are standard in the industry, it may be wise to combine several categories, so that when a market profile matrix is created there are fewer cells in the overall grid. Here's a typical example of three segments – small (e.g.1-49), medium (e.g. 49-999) and large companies (e.g. 1,000+). The specific size definitions depend on the marketing situation you face and how granular you want this definition to be. By the way, the government's definition of small companies is 500 or fewer employees, so be careful in throwing around these terms, as a"small company" definition for one person may not mean the same thing to another individual. In fact, one of the most important issues in B2B marketing is to have common definitions of terms.

How to Profile

The profiling process relies on matching your customer or prospect record to public information from data compilers and enhancing your record with the selected information (e.g. NAICS code and company size). That may sound easy but it's not! The problems lie primarily in matching your customer record with the corollary company record in the public database. In many cases, the way in which the customer or prospect record was inputted may not match the way in which the same company record is listed in the outside database. Let's use DuPont as an example. There are multiple ways that DuPont could be listed and here are a few:

• DuPont
• DuPont & Company
• DuPont de Nemours
• The DuPont Company
• DuPont (fill in the name of the division and all the variations)
• E.I. duPont de Nemours and Company (the official name)

If the data entry of the name differs from the spelling used in the outside database, the software matching process may not find the match due to this small but significant difference. Several years ago I got a call from someone; who had attended one of my seminars. I used the DuPont example in the seminar sessions, and her company, sells to DuPont. So, when she went back to work, she looked DuPont up on their customer database. It had 73 listings for DuPont -- an obvious problem if they wanted to see what their total sales were to this customer. Among the 73 listings were several duplicates of the same plant, in addition to multiple divisions of DuPont.

This does not even take into account the mailing address for a company, which has its own set of problems in addition to the"official" address of the location. Large companies can have as many as three legitimate addresses. There is the"front door" address (usually the one that the public databases record); the billing address, which may be a PO Box; and even the shipping address, which could be the receiving dock. The last two addresses are frequently found on the customer records, particularly if they come from accounting. As is easily seen, the possible variations are multiple and can potentially cause the computer-matching software program to conclude that the two records are different, even though they are actually the same company and location.

Because of this problem, match rates found between customer records and outside databases are usually only in the 60-70% range, even though the information on the internal company information also appears in the outside public database. This difficulty is known, and many computer service bureaus have written special algorithms in an attempt to solve the non-matching problem. These programs work better than consumer matching software does, but still not well enough to get much above the 80% matching rate on the first pass.

Now that you're aware of the problem, what do you do? First of all, insure that the address of the customers in your database is postal-certified by running them through CASS certification and, of course, the National Change of Address (NCOA) process. Then, assuming that you are using an outside computer service bureau or database provider to perform the matching and enhancement process, be aware of these issues and inquire as to their unique processes and capabilities to deal with the matching problem. When the matched and enhanced list is returned, the second step may be to manually enhance the non-matched records. This may require a look-up process, as it is important to have the NAICS code and company size definition on as many records as possible, and certainly for all the large customers, to properly complete the profiling process.

Now, let's see how this type of analysis might be shown once completed. The following table shows a real example of an attendee profile that was done for a seminar firm who held sales training courses. This firm had initially rented lists to fill the seminars based on response criteria of individuals who had attended a seminar similar in nature, and by the geographic area where the seminars were to be held, in this case Ohio, Indiana and Michigan. These response lists had no industry selection criteria. After a number of seminars had been held, they wanted to improve their targeting and list selection criteria by analyzing the actual attendance by industry, as they were expanding the seminars to other states. It was decided that the size of the firm was not key to their analysis.

The first step was to match the attendee list with a compiled database, and enhance it with the SIC code (their master database was in the legacy SIC code, and therefore we kept with this as the coding system). The match rate of the entire file was only 67%, which was satisfactory for this level of analysis. The first chart ranks the top 20 industries by 4-digit SIC clusters based on the total number of attendees to their seminars.

This first result, by attendee count, clearly uncovered four-digit SIC code clusters. The firm's initial reaction was to rent lists based on this analysis, and select all the companies in each of these 4-digit SIC codes in the states where their new seminars were planned. This would have been in addition to their continuing to rent high performing response lists. While this might have improved their response and attendee conversion rate, this is not where we stopped, since by only looking at the total number of attendees by SIC is a one-dimensional analysis.

Penetration Analysis – the more insightful picture

The first level analysis was based on the number of seminar attendees present the four-digit SIC description. As a comparison of one industry type vs. another, it is a solid approach. If the total number of customers is relatively small, and there are less than 10 per cell, then the analysis may well stop here. On the other hand, having many customers in one industry may not provide the most accurate or insightful picture.

The next step, or penetration analysis, compares the industry counts against the total market size by the same SIC cell definition, and determines a"share" of the total market, which will be a more accurate measure of market responsiveness. The problem for most B2B marketers is that the total number of customers is frequently not large enough to calculate reasonable percentages (remember the old rule of significant figures). If there are enough customers in each cell then a penetration analysis will provide a much clearer picture of where market success has been greatest.

Here is the same statistics displayed by calculation of a penetration analysis. You will see how the ranking of attendee SIC clusters change.

As can be clearly seen, there are different rankings based on the penetration analysis. As an example, only one out of the top five"old" rankings remained in the top five. Four new SIC codes rose to the top as the most"penetrated" clusters.

As a penetration analysis is developed, a key issue is defining the scope of the market. As an example, this seminar company chose to compare the attendees in the three states to number of companies in this geographic area where the seminars were given. Any good database of U.S. companies can easily be sliced by geography, making the penetration analysis more relevant to the actual customer acquisition effort.

If the base number for this analysis is too large (such as comparing it to the total US), then the analysis will not have statistical reliability, as the percentages will be too small. It's your market situation and customer/prospect count that will drive the decision as to whether to try a penetration analysis or not.

In addition, don't forget that this analysis can form the basis of other calculations. The most common of these is sales revenue by industry. Once the analysis is done and most, if not all companies are placed in their respective industry cells, it's a simple matter to then add up the total sales volume to then obtain another insight into which industry is has the highest revenue. Further calculations of product sales or even gross margins by industry are also common.

Final thoughts on profiling:

Most of this description has dealt with customers, and rightly so, as customers have displayed a real behavior – they bought! At times the profiling for inquiries or leads might be also prove to be helpful, as typically there are far more leads than customers, resulting in more reliable statistics. In addition, inquiries are fresher and might be more reflective of current market interest and provide additional insight than a customer base that has been achieved over a long period of time.

On caveat to consider is that many companies have directed sales activity to target specific markets, and like other things in life, this can prove to be a self-fulfilling prophecy. Naturally you should see success where these marketing efforts are concentrated, and this will be reflected in the number of customers in each industry cell.

Therefore, the analysis is not solely a reflection of customer interest in your product or service, but rather a result of dedicated sales effort. Just be aware of this influence so that your marketing insights and conclusions from this process are not"uniformed".

TARGETING – WHERE SHOULD YOU GO?

Of all the marketing processes, the one that will be the shortest in this white paper, but is absolutely critical for any firms' success, is the targeting process. This process relies on four different inputs that point to the target markets for subsequent sales and marketing effort. This marketing planning process uses these four inputs, but is more a matter of actually"doing it" rather than writing about it, as B2B situations are quite different from one another. The following brief description outlines these four basic approaches to targeting, but only disciplined internal discussions can produce the final list of target markets and their priority for marketing campaigns.

A brief warning is in order at this point. The result of the targeting process is a definition of the target markets that represent the best chances for sales success – that's obvious. But there can be a fine line between a target that is too large and undefined vs. one that is too small and granular. As an example, I frequently hear marketers and/or management individuals announce that they want to pursue"small businesses" or"the mid-market" as a target. Unfortunately this description is undefined, and covers too many companies to constitute a true target market. Yet I see large budgets being spent (particularly in the banking industry) in an effort to sell these types of"target markets". These efforts typically do not produce the results desired or expected.

On the other end of the spectrum are target markets that are too small to constitute a market of any size. Frequently, these target markets are derived from a sales success, and the one or two companies sold are then held up as the kind of companies to target. So marketing accepts this description and proceeds to develop a campaign to reach this"market" and does no further checking to determine if it is really a market large enough to justify a sustained marketing effort. Results may be achieved for those few companies that fit the description, but at a high cost per sale.

Two criteria should be considered before deciding if the market described should become an official target. The first criterion to use is the economic value this target market represents to your company. Obviously, the greater the value, the more everyone will see it as legitimate, and worthy of sustained marketing and sales effort. The second criterion is the difference it would make to the potential customers if you approached them with more relevant communications and offers based on your targeting. One of the keys today in breaking though the clutter and insuring that your message is heard is the degree of relevancy it contains. Therefore, careful targeting can be the basis for the development of relevant messages and offers that will better connect and drive the desired behavior.

Four Inputs for Development of Target Markets

1. Profiling results: The effort to profile has several purposes, but the most important one in the targeting process. At minimum, the profile will uncover the industry segments in which sales success has been achieved. While the profile of the customers and/or prospects may be too granular if the 4-digit SIC or 6-digit NAICS codes is used, this coding system rolls-up to a broader definition of each industry by going to less granular definitions. In all the years that I've done profiling for companies and even while I was at IBM, there never was a time that it didn't produce an"ah-ha" from the product management, marketing or sales group. Frequently, the comments made were,"I never knew we sold these types of companies" or"I didn't realize that these firms represented this much of our business!" Obviously, this new insight altered the targets for next year and drove marketing investment.
   
   
2. Markets of known opportunity. Most companies focus on these markets, as they are already known as target markets in which there is a proven sales opportunity. For this input, the definition of the target market can become difficult. As an example, if the target is"small businesses" then we have to be more specific. Frequently, the definition of a good target is not one that can be easily quantified with a convenient list of companies.
   
   Recently, a sales force automation client was targeting mid-market companies with at least 10 salespeople. Mid-market wasn't so difficult to define by employee size, but finding a list by how many salespeople work at companies was next to impossible. So, one of the first lessons in targeting is to define the target in a manner that can be tied to outside databases and company lists. Otherwise, the definition may be internally correct, but it's not actionable.
   
   
3. New product/service introductions. Every company has new products or services they are planning to introduce. One of the main jobs of product or market management is to define the marketplace and opportunity even before effort is expended on product development. Logically, there should some target marketing and research done prior to development work done on these new initiatives.
   
   That's in the ideal world. Unfortunately, far too often the product development is driven out of technology, with the hope that a large enough market exists for this product or service. Sometimes, a market develops that is entirely unexpected. My first job at BF Goodrich Chemical found me working as a product manager for a product that was developed as a denture adhesive (not successful), and was sold in large quantities as a thicker for cosmetics (very successful). Be alert to new unexpected opportunities.
   
   
4. Competitive opportunities: The fourth, and least frequently used targeting approach, is based on competitive opportunities. All sales and marketing people know that competition exists, and are prepared to address it individually when it arises in the course of a sale. In fact, many marketing organizations perform a through competitive analysis, and forward it to sales, so that they can individually offer appropriate counter features and benefits when faced with a specific competitor. This is a common view and usage of competitive intelligence. On the other hand, why shouldn't marketing define the broader market in terms of the competitive product or service being used, and then position the communications directly against that competitor or type of competition?
   
   

   Here are two examples of competition that make the point:

• Direct competition is always defined as those companies who are offering the same and/or equivalent product or service. Recently, a large Chicago regional bank saw an opportunity to"steal" away small business customers of another large regional bank that was being acquired by one of those mega banks (names withheld for obvious reasons). Clearly they saw the opportunity based on the perceived loss of service for these small businesses and positioned a series of service offerings and messages to convince those"acquired" customers to move their relationship to their"more local and friendly" bank. Not only did the advertising campaign address this, but also a highly targeted direct mail effort as well. The key was to obtain the names of those businesses customers of the acquired bank. This was available through a data source, and the campaign was launched and good results achieved. A current and real life story of targeting by competitive opportunity.
  
  A number of years ago, I worked with a less-than-truckload national freight company. They had several national competitors, and had the vision ask the sales people to record which competitor was sharing each key account. The rationale was quite straightforward – it was the labor contract with the Teamsters. If one of their national competitors was engaged in either a tense labor negotiations or a strike, then a targeted message could be quickly sent to those key customers in an attempt to gain"share-of-customer". They used it several times with great effect, and much to the dismay of the competitor.
  
  Clearly, if this type of competitive analysis is undertaken it then becomes a field in the database. If the direct competitive set contains multiple competitors, then all should be listed.
  
  
• Indirect competition is defined as the offset product or service that replaces the need for your product or service, but is not the"direct" competition. In the last 10 years, technology has been responsible for many of these indirect competitive situations. This form of competition is particularly dangerous to companies, as they focus so much time on direct competitors that they frequently miss the indirect form of competition, and lose entire markets.
  
  One classic example in the 1980's was FedEx vs. fax, as these were two ways to deliver hard copy documents to distant locations. FedEx even opened fax centers when fax machines were not common (one was even across the street from my old office on Michigan Ave. in Chicago) in an effort to compete, but abandoned it quickly, as fax machines were bought quickly by all businesses. This form of competition changed the structure of their market forever, and is a classic example of indirect competition.
  
  A more current example is the indirect form of competition the US Postal system faces with email. The significant drop in firsts class mail combined with direct mail is fueling a very rapid drop in mail volume and revenue. They have no capability to fight this form of indirect competiton.
  
  For the purposes of selecting target markets, the definition of that indirect competition that might be an easy target can become a great targeting opportunity. It does require accurate data on competitive usage, but that can be obtained in a number of ways including a telemarketing survey. Once the required data is obtained, it can become a great target market, as the degree of relevancy of the message and offer can be very high.

Some final thoughts on targeting:

There they are – four methods to develop target markets. This process is absolutely critical, as it directs much of the following marketing activity. This process is usually done at planning time, and sets the stage for segmentation, as typically these"targets" are too large and need to be sub-segmented for effective marketing communications.

It is a company wide concern and process and it can't be under estimated how important this is to solid performance as sales and marketing productivity starts by going after the best and most profitable target markets.

SEGMENTATION FOR COMMUNICATIONS

The critical role that segmentation plays

Salespeople have been making their message highly relevant for as long as sales people have been around, as they choose what they do and say based on their current knowledge of the customer and/or prospect – in essence true one-to-one marketing. In fact, much of professional sales training is focused to equip them with the skills; both technical and professional, to"size" up the company and individual, and proceed accordingly. Therefore, if the marketing group is to fully integrate with sales and assist in the coverage of the territory, then they need to communicate more like salespeople. The problem is that marketing communications is not a face-to-face activity. It needs a process to develop and build the basis for any communication, which then targets companies and/or individuals. The process of more finely segmenting the market coupled with recording this information in a database are the primary methods for the marketing group to replicate the sales interface as closely as possible.

Clearly, this process of segmentation, data gathering and communication using this data is not the same as sales calls, and is certainly not a replacement for it, but it's a dramatic improvement over the past methods marketers have used to develop communications to targeted audiences. The capability of using a database of information has enabled marketers to act more like salespeople and come closer to that one-to-one communication.

Segmentation – three choices:

The term segmentation has many meanings. The most commonly used definitions are; macro, micro and one-to-one segmentation. The one we will focus is micro-segmentation, as this is the"active ingredient" in highly relevant marketing communications.

Micro-segmentation – definition and benefits:

Micro-segmentation is just what it implies – a further splitting of a market segment into smaller groups or clusters that are communicated to with highly relevant messages and offers. These micro-segments and clusters are composed of companies and/or individuals that have something in common with each other as they relate to your selling situation. Here's a helpful definition:

"Micro-segmentation is a process of grouping together individuals and/or companies into cluster that sharing common characteristics relevant to the sales of your product or service, selling process, or company."
The benefits to micro-segmentation are multiple. The key ones are:

1. It reduces the size of the target audience, which likewise lowers the communication budget.
2. It creates the opportunity for very relevant messages and offers based on the clustering criteria –the more highly relevant the message, the better chance it has to break through the clutter and connect with the target audience, driving the desired behavior.
3. By focusing on the micro-segment, it magically causes the discussion to revolve around the customer or prospect and, in a surprising result, creates the sought after"customer driven marketing" processes that companies frequently talk about but don't actually do!
4. Once the micro-segmentation process has been completed, it can be used to define the data elements required in the database.
5. Finally, a defined micro-segmentation matrix of the market can be used to screen inquiries as they are received to see if they fit the criteria of the best segments, and therefore deserve to be more aggressively followed-up.

Micro-segmentation – dangers:

There are other benefits of this process, but there are also some dangers as well and the key ones are:

1. Too many micro-segments can create too many projects to effectively manage. Frequently, once exposed to the concept, marketers can go a bit overboard, and want all markets finely segmented with individualized communications sent to each. While this is technically possible, it's not practical based on the physical number of projects and associated cost that is required to develop the message, execute, and follow-up each micro-segment. It's far better to hit a few important clusters thoroughly than try to launch too many communications with insufficient message quality and follow-up.
2. The criteria selected for the sub-segmentation is wrong. At times, one person's view of a market may be flawed to the point of not truly understanding how the market is really structured. Good market research and even profiling should provide the insights to select the criteria that will accurately reflect the market structure. But still too many segmentation schemes are created based on one or two individuals'"opinion." The small business sector is particularly hard to segment. The lesson here is that once selected, the criteria used for segmentation are difficult to alter. So, be sure that it really reflects a difference between sets of companies that, when translated into different messages and offers, produces results.
3. Many sub-segmentation schemes sound great in concept but when it comes time to record them in a database, it becomes apparent that the data required to describe the segment is either not widely available or easy to update. Here's a classic example. Salespeople, at times, want detailed information about a potential customer that gives them a reason to call. In the chemical industry, for instance, the pounds or gallons of the product purchased would be a great way to segment the market, yet this type of data does not exist on available public databases. Even if we were to telephone all potential customers and obtain their consumption data, this information would change quickly over time and cannot be updated easily. Therefore, a segmentation scheme based on product consumption doesn't work unless there is public reporting of this data – a rare find, to be sure.
4. Even the best segmentation scheme needs support of others in the organization. In B2B, primary support comes from the sales and product management groups. Clearly, if they don't buy into the scheme it will most likely die. As the sales group is the eyes and ears of the company, they need to support any segmentation approach. Simply, they must use the information and knowledge that is derived from the segmentation scheme in their daily sales activities and provide feedback on results. The most common disconnect with a sales group is based on their geographic coverage vs. an industry-based segmentation approach. Individual salespeople may or may not have these types of customers or potentials in their territory, and if they don't, then they really don't care about what marketing feels is important. If salespeople don't buy into the segmentation definition, they won't record the information and nothing you can do will change that.
5. Resistance may also come from product management, if the segmentation approach is contrary to the traditional view of the marketplace and how the company has grown over the years. Don't forget that we are attempting to bring change to the traditional sales and marketing roles and this"new-fangled" thinking may not be embraced readily. This is particularly true if the company has a strong product management focus. Product groups segment the market through the eyes of the product, - not the customer. This has lead to much tension in organizations, as the trend is clearly to move to a customer-based view of the market rather than a product-based view. Micro-segmentation is one of the key mechanisms to alter how a company approaches the market, and may be difficult for traditional product management people to accept. Universal support for any customer-based segmentation approach has to be achieved within the company or it just won't work.

Eight Micro-Segmentation Approaches

There are many different approaches to develop clusters for marketing communications campaigns, whether the target is a potential or current customer. The following eight micro-segmentation approaches may all be useful for your marketing and sales situation, but not likely. Select from those that make the most sense, and then build your process and database around the best ones.

1. Demographic Segmentation:

   This is the most common way to segment and should be done by every company, as it becomes the basic data platform of micro-segmentation. Here are the common demographic data elements to consider:

• Geographic: We all know that the ZIP code describes the geographic location. The problem with ZIP codes in B2B is that they are too small a geographic area to use even if you're targeting small businesses. The first three digits of the ZIP code are called the SCF or Sectional Center Facility – an arcane post office definition. Thus the SCF generally becomes the smallest geographic description that is useful in B2B. This becomes important when selecting lists since the ZIP code is usually too granular for geographic targeting. Don't forget that an the telephone area code is also a form of geographic description, even though they are now being revised frequently.
  
  
• Industry type or SIC/NAICS code and company size: These two demographic descriptions have been thoroughly detailed in this white paper, and are the first basic building block to understanding and segmenting any B2B market.
  
  
• Location type: The type of facility may be an important piece to help define your market place. Most companies are single-location firms, but as the size of the company grows so does the specialization of the site. Here are the most common site definitions for larger companies:
  • Headquarters or HQ
  • Plant
  • Research Center
  • Retail store location
  • Regional or District Sales Office
  
  Most compiled databases contain these definitions or similar ones. Your decision is to determine if site definitions are important and, if so, record them in your database.
  
  
• Fiscal Year: Most companies are on a calendar fiscal year. For example, the IRS requires that any private or Sub S Corporation conforms to a calendar/fiscal year. On the other hand, C Corporations are allowed to select their fiscal year, and about 80% are also calendar/fiscal. The other 20% are spread among the other three quarters for their fiscal year start.
  
  Why could this be an important demographic data element? Well, if you're selling a product or service that needs to be in your customer's budget, then you better communicate to this company before or during their budget cycle or you will miss the biorhythms that take place in a company during planning time. When I was with Quaker Oats, we were on a fiscal year that started on July 1st. That meant if you wanted to sell me sales training programs, you better start talking to me in the spring. All too often, we make the assumption that the fall of the year is when companies are in the planning process and launch campaigns at that time. By having the fiscal year defined, you can then segment and communicate to those companies based on their unique planning cycle and beat the competition to the budget.
  
  
• Year founded: In my seminars I listed this as a demographic element to consider, but over the years it began to seem meaningless. That is, until I once said it was useless in a speech, and by coincidence, a marketing manager from National Pen Company piped up, and said that it was a key demographic way to segment their potential customer base, as they sold pens to companies on their 1-, 5- or 10-year anniversary date. So I continue to list it. Also, some people claim that newer companies need different products or services than older ones, and if this also describes your marketing intelligence then include year founded as a demographic data element.

2. Relational Demographic:

   This is an enhancement on demographic segmentation, and is defined as factual information that is related to the sale of your product or service. The most common relational demographic elements are equipment or process in use by the targeted account or micro-segment. This single piece of information may well completely change your message, offer, and even your total marketing approach. It can be an extremely powerful piece of information, as it will make what ever you do more relevant to the targeted audience. In fact, salespeople have been using this type of information for years without ever calling it by a name. It's just smart selling to adjust the approach and presentation based on information about the company.
   
   To make this clear, just think that you were selling a plastic material such as polyvinyl chloride, or PVC. Knowing if the company was an extruder, blow molder, rotational molder or calendar operation would make a great deal of difference in not only what you said but even the type of PVC you would sell. Another example is sales and marketing software. Whether the company has outside or inside salespeople and/or how many sales people there are makes a big difference in the software opportunity. This then becomes a relational demographic data element.
   
   While there may be numerous relational demographic facts for consideration, I recommend the following process. Call or visit one or two of your best salespeople and ask them the following question –"What single piece of information would you like to know about a company that would make the most difference in your sales effort?" Make sure they know you are looking for demographic information. They will often be your best source of input on this and many other database elements.
   
   This relational demographic fact is an untapped opportunity for most companies to use for micro-segmentation and can be very powerful in constructing highly relevant communications that cut through the clutter. In addition, don't think that your competitors are not working on segmentation schemes and, just like you; they will add the standard demographic data to their files and be able to segment the market using this data. But, if you go one step further, and establish a relational demographic field on the database and populate it, then you have a segmentation scheme that is not easily duplicated. In other words, you might be able to perform"stealth marketing" and out-smart the competition. Yes, it is true that the data will have to be obtained on a company-by-company basis, and will most likely need the involvement of the sale group. But, I guarantee that it will be worth every dollar of effort. To start, try for only one relational demographic element, and then use it to good effect.
   
   

3. Sales Cycle Segmentation

   There is a commonly accepted sales cycle in all companies. It may vary from industry to industry and depend on the nature of the product or service being sold, but it does exist. Therefore, one of the most logical forms of micro-segmentation is to define where a prospect or customer sits in the sales cycle and then communicate to them accordingly. As the customer typically also knows where they are in this sales cycle, the message resonates well with them.
   
   Here's a general treatment of a sales cycle with brief descriptions. Don't try to create a new sales cycle, as the sales group will resist this effort. All you're trying to do now is define what your company's sales cycle is.
   The sales cycle may vary from segment to segment in your market place, so different sales cycles may co-exist within the same company. In addition, the common definition of terms will also be a great value internally when discussing the stage of the sales cycle.

• Suspect: These are companies that you have some reason to believe should want or need your product or service. They have not indicated that they are interested, but do have the proper profile of a potential customer. Frequently, these suspects are on lists that you order from list compilers, can be found on trade-show attendance lists or even directories of companies in an industry association. No matter how or where you find these lists - they should be called suspects, not leads.
  
  Unfortunately, some list compilers refer to these types of lists as"lead lists". This does a major disservice to marketers, as none of these lists of un-contacted companies are"leads." They are suspects, and should be referred to in this way. Finally, at this stage of the sales cycle, suspects are most always companies and not individuals.
  
  
• Inquiries: An inquiry is an individual who has"raised his or her hand" in response to some form of marketing communication. They still are not leads, as you have no idea at this junction whether or not they have serious interest or can even qualify to buy your product or service. Inquiries arrive from different sources and methods, and until they are qualified as leads they should be called inquiries. One of the biggest problems for marketers is to call an"inquiry" a"lead". There is a big difference between an inquiry and a lead! Based on years of experience there are generally no more than 10% of the inquiries that move on to leads.
  
  
• Lead: Now we come to the common term"lead." Unlike the previous definitions, a lead is an individual who represents a company that can buy your product or service, and is seriously interested to do so. There may be several lead categories, including the always famous"hot lead." A common lead category is"demo" or"sample request" which is usually a big step up from just someone expressing an interest and capability to buy. Needless to say, the definition of the term"lead" is the one where the most conflict has occurred between marketing and sales, and should be the focus of dedicated effort within your company to reach a common and accepted definition.
  
  
• Proposal/quote: In most B2B situations, there is a proposal or quote sales step. This is where the process of obtaining new customers is quite active and handled by the sales group. As this is such a sales-intensive process, it may be important to cease marketing communications during the proposal phase, as not to confuse or disrupt the"close." It is at this sales cycle step that many sales people suppress or"red flag" a record in the database as a direct message to marketing to"stay out." This is very understandable and should be honored, but the"red flag" should be removed once this process is completed and the prospect either is sold or not.
  
  
• First purchase: Customers are not just defined as"customers", since they remain in the sales cycle and need further definition as to the type of customer. The first of these customer definitions is"first purchase." The reason that this"first purchase" category is important is that when someone buys from your company for the very first time, there may be a need to focus even more on them to retain their business.
  
  Customer satisfaction calls, welcome letters, etc. all should be directed at first-time customers to do one thing – insure the second purchase. Studies have indicated that in commodity areas (e.g. office products) there is little correlation from first purchase to second purchase. These companies may, in fact, be trying you out and, in their mind, do not really think of themselves as customers yet.
  
  A big mistake companies make is to assume that the salespeople will pick up the new customers and treat them properly. The ironic truth is that this may be the exact time more marketing communications is required to help confirm the decision to buy on the part of the customer. These same studies have indicated that there is a strong correlation to repeat and long-term purchase once a customer has bought the second time.
  
  
• Repeat or Good Customer: Simply, a general customer definition that distinguishes this customer group from the new or first purchase group. Here is an opportunity for more granular customer definitions based on your business model.
  
  
• Past Customer: Here's a question all marketing and sales departments must answer – when does a customer become a"past customer"? Is it some artificial accounting period like a year? Or is it when the customer feels they are not going to buy from you again and, in their mind, are a past customer?
  
  At IBM, one industry group was asked this question and answered as follows – if a company hadn't bought within an 18-month period then they were officially a"past customer." When these same"past customers" were surveyed the surprising result was that almost 50% responded that they thought of themselves as an IBM customer. A disconnect to be sure and if we had begun to communicate to them as"past customers," we would have surely been sending the wrong message, and ironically creating a self-fulfilling prophecy.

4. Behavioral Segmentation:

   To coin a phrase, it's what people do rather than what they think that really matters. As a direct marketer, this is music to my ears, and should be to yours as well, since we spend our marketing energy to drive behavior or responses.
   
   I know advertising and public relations experts are concerned with creating top-of-the mind awareness and positive brand image, and there is much value in that goal. But, as a direct marketer, I want to cause individuals to do something vs. just thinking about doing something. To be effective, we have to drive the behavior, capture it, and then respond to it appropriately.
   
   A micro-segment of individuals who have responded for a Web seminar have something in common that can used as the basis of the next communication to drive – you guess it – more behavior. Here are some behaviors to capture. In your selling model there may well be others, and if so, add them to the list.

• General inquiries: At times, advertising and even PR generate general responses or inquiries. While it is difficult to determine why they responded, there should be"gold" in these inquiries. They may arrive by phone or from a registration or inquire on the Web site. Record the media that the individual used to inquire on the database.
• Responses to specific offers: As most marketing communications today will have a specific offer, it is important to not only record the inquiry, but also the offer that the individual responded for. They likely will remember and so should you.
• Trade show"stop bys": I call this category"stop by," as that's exactly what most trade show results are – nothing more than someone stopping at the booth, and allowing their card to be taken or badge to be swiped. While there are good leads in these"stop-bys," most will not even remember doing so. Just ask yourself how many booths you visited at the last industry trade show and how many you can specifically recall? Not many, I'm sure. Thus, this should be a separate category of behavior.
• Seminar attendance: Whether it's a Web or traditional seminar, the experience is quite different for the individual vs. just inquiring for an offer or stopping by a trade show booth. This category of behavior also shows a serious commitment on the individual's part as everyone's time is scarce, and to attend a seminar means that they are very interested in the topic. Always record the date and topic of the seminar for future reference and usage.
• Multiple-responders: Individuals who respond multiple times and in a variety of ways may be, by themselves, another important category. Rather than just keep all individual behaviors separate and therefore miss individuals who are multiple responders, their actions all should be brought together. This group may, in fact, be the most likely to buy as they demonstrate continuing interest.
• Digital behavior: A new opportunity to track and capitalize on behavior is in the digital arena. Much has recently been written about this new form of behavior as it should. Also, new tracking software is now available to perform this analysis. While this is a good source of behavioral information be careful that it can be tied back to the individual taking the action and not just a company URL.
• Customer or purchase: This list would not be complete without at least referencing customers who purchase. As this is a progression that is likely based on prior behaviors, it may be important to tie these past actions together and analyze them to predict which ones seemed to be more predictive for sales conversion.
• Calls to Customer Service: Here's a behavior that many marketers miss, as they are focused too heavily on the acquisition side. Customer service calls may be just that – calls for assistance. On the other hand, they may represent additional up-sell or cross-sell opportunities, depending on the nature of your product or service. Think about how these calls can be translated into knowledge for future sales efforts or loyalty campaigns. One word of caution here, be careful in tasking good customer service people with up- or cross-selling responsibilities. They won't like it and will not do it well as it's not in their nature to be salespeople.
• Other behaviors: There may be many other behaviors unique to your company that should be recorded for future reference in communications to those individuals. As an example, attendance at company-sponsored events like golf tournaments would be valuable, as those who played a round of golf or attended a pro tournament as a guest of your company can be communicated to in a highly relevant manner.
  
  A new form of behavior feedback is now found on Twitter, Facebook and even blogs. These can be misleading, but should be considered if these forums make sense for your business.
  
  Think of all the behaviors relevant to your sales process and allow for them to be recorded on the marketing database, as it will pay dividends.

5. Competitive Segmentation:

   All products and services are sold in a competitive environment, and as discussed in the targeting section, competition should be carefully defined for that purpose and for potential marketing communications as well. As a quick review, two forms of competition have been mentioned in the targeting section and here are two more in addition to direct and indirect competition.
   1. Direct competition: This form of competition is always the one everybody focuses on, as it's the most obvious and consumes the attention of product management and sales in an effort to beat it. This was discussed in the targeting section.
   2. Indirect competition: One of the most difficult forms of competition to manage is indirect competition. Following up on the indirect competition FedEx faced from fax machines, they are now seeing indirect competition from the Internet and the sending of documents via e-mail.
   3. Budget or lack of it: Not competition exactly, but more of a roadblock. Many times your target customer does not have the budget to purchase your product. It may be that they will never be able to spring loose the dollars, but more than likely it is a matter of timing. Recording the details behind the budget roadblock can lead to more effective communications.
   4. Status Quo:"Why should we change?" is a very tough form of competition faced by all salespeople. Your product or service fills a specific need, and assuming it's not a breakthrough, then the need usually has been satisfied by an existing product or service. Yes, your solution is better but – why change?
      Defining the competitive set and recording it on the database provides great opportunity for micro-segmentation and highly relevant marketing campaigns.
      
      

6. Analytical segmentation:

   While this is not a white paper on advanced database marketing, a brief mention should be made about the use of analytics to segment the market. The problem in B2B is that the data needed to adequately support analytic techniques is poor, and therefore the results can be inaccurate or at worst misleading. The primary purpose of these analytic techniques in direct marketing is to predict the likely response rate and conversion to sale between one set of customers and another. In other words, which segment of the customer database might respond better to a specific offer or message?
   The most common analytical segmentation is RFM, or Recency, Frequency, or Monetary. RFM has become an extremely important process in the catalogue business and, of course, can be used for business cataloguers as well. It applies only to customers, as an actual sale is necessary to perform the RFM analysis. Here's a brief summation of the technique. Take the customer file and analyze it as follows:
   
   
   • Recency – the customers who have bought the most recently are at the top and cascade down based on oldest purchase date.
   • Frequency – the customers who have bought the most are at the top of the list and those just once at the bottom.
   • Monetary – the customers who have purchased the most in the form of dollars are at the top and decline to those who have bought the least.
   
   For each list, cut it into quintiles, or fifths, and assign a score to the top fifth of 5, then 4, etc. until all customers have received three scores. Total them up and the ones with the highest score will more likely respond to your next promotion at a higher rate than those with lower scores. Sounds plausible and in consumer marketing, it has been validated many times.
   
   The problem in B2B is that many business models do not fit this simplistic look at the sales relationship. As an example, what about customers who are on yearly contacts and release shipments monthly – have they bought once or 12 times? How about distributors who are, in effect, buying to meet end-user demand and/or inventory? What about sales of capital equipment where one sale can be in the millions, and will not be repeated for several years if not longer? But, if this fits your business model then use it, as it can be a very powerful predictor of future behavior.
   
   Even thought RFM may not apply for your business situation, there is one element of the three that you may want to seriously consider when constructing a list of customers or even prospects for the next campaign. This element is"recency." The more recent the inquiry or purchase, the more likely you are to get a response or additional purchase. This is the most predictive of the three elements in the RFM formula.
   
   

7. Need or pain base segmentation:

   There's been a lot written about need-based segmentation and, in theory, it makes great sense. Now if it only could be executed! Let's first distinguish between assumed needs and real ones. In basic sales training, you are suppose to"assume a need" of the potential customer based on what you know about the company. Then, through probing, you determine if that need is real or if other needs are more important. We, as marketers, can do the same thing and assume a need for a certain micro-segment, and then position and pitch our product or service in relation to that particular need.
   
   The problem in B2B is that need for a purchasing agent is quite different from that of a plant manager, even within the same company. It rarely is true that all the decision-makers and influencers will share the similar needs in the same priority. In addition,"needs" are situational, and can change with management or economic changes. What would happen to an organization upon the change of top management from a revenue-growth CEO to a cost cutter? I think you get my point – need-based segmentation is difficult to execute.
   
   
8. Job functions or functional psychographics: As we market to individuals who play various roles in their respective organizations, an excellent way to market is to think of all the individuals who share the same function as a micro-segment. Then instead of marketing vertically to companies with different individuals and functions, segment horizontally and develop campaigns directed at functional job segments across all companies. The message to a CFO vs. a plant manager about your product or service may be quite different, even though the product is not.

In Summary:

In my over 35 years in B2B sales and marketing I've seen both spectacular mistakes and great successes. Some of each was achieved without a targeting and segmentations process. On the other hand, when this process was applied, the successes far outnumbered the mistakes. Today with the increasing demands by senior management on measuring results and justifying marketing expenditures, it is almost inconceivable that success will be achieved without a disciplined process of profiling, targeting and segmentation.

This may well represent a process change for many sales driven companies, as no longer can firms solve their marketing problems by sending salespeople to make"one more call a day". This is now too expensive and inefficient. Better direction is needed not only for marketing campaigns but for the sales staff as well. Hopefully, this report starts you on the road to marketing success. View it as the start and not the finish!