Research-Based Market Segmentation: What You Should Know

By Kate McPherron, technology evangelist

According to BusinessWeek’s March 2007 “50 best performers” issue, top performers have something in common: research-based market segmentation. Businesses that target specialty markets will promote their products and services more effectively than a business aiming at the “average” customer. From Coach to Oracle, many companies realize the importance of segmentation and are realizing positive results. Coach’s new travel bags accounted for 15% of sales and Oracle won 30 new retail customers in 12 months, including Wal-Mart, Nordstrom and Perry Ellis.

In today’s connected world, segmentation is considered a must for marketers to identify the proper targets and appeal to those targets more directly. By targeting prospects directly, marketers can gain higher and quicker return on their marketing investments. Marketing resources can be effectively deployed by matching investments to the segments that hold the greatest promise. Additionally, segmentation models enable businesses to gain an understanding of the geographic and demographic attributes of their target markets, allowing them to implement strategic sales and marketing initiatives. Or so the theory goes.

This 9/18/07 Marketing SIG session offered practical insights into the reality of market segmentation, providing three different perspectives on market segmentation from executives who have been there.

Opening Remarks

Moderator
Paul Abel, founder and managing partner, Blue Research

Panelists
Shauna Pettit-Brown, manager of exploratory market research, Intel
Laurie Cremona, vice president, BEA Systems
Desh R Urs, managing partner, iBridge

Paul Abel introduced the subject by defining segmentation as taking a market and dividing it into groups with distinct characteristics, behaviors or needs. He illustrated this point with a 3-minute video clip from PBS’s “The Persuaders,” showing how political campaigns use segmentation to deliver messages customized for different demographic groups – a necessity in a “narrowcast” world. The video also described how much data is currently tracked and available on consumers that can be used for narrowcasting.

Paul explained that segmenting aims to find clusters of potential customers with a lot of commonality within the group, and lots of differences between groups. Ultimately the goal is to be able to more efficiently and effectively serve the needs of target segments as opposed to trying to serve a heterogeneous market, which can be very inefficient indeed.

Shauna Pettit-Brown: My introduction to segmentation was at Gartner, where I did primary research. At Intel, where they have a lower budget for research, we’ve had to use different ways to do segmentation, although it’s still the same basic process. Key things it’s important to do:

• Do background research: get to know the target audience (market) through secondary research, interviews, etc.
  
• Develop hypotheses about what separates groups: variables will distinguish one group from another.
  
• Determine market potential: Once identified, groups can be sized to determine market potential.

The basic notions that Intel had weren’t sufficient to identify or size the groups.

Laurie Cremona: BEA was faced with moving from enterprise computing to a service-oriented architecture (SOA) model. We placed a big bet on SOA. Since the CMO was very data-driven, we applied marketing more scientifically than in the past, identifying different levels such as job function, vertical market, enterprise size. We then extended the developer-to-developer program into those audiences, targeting each audience with specific messages. In the process, we developed hypotheses, looking, for instance, at whether they opted in for newsletters, whether they were customers or competitors’ customers. BEA chose GCR to do a brand study to look at the strength of the brand and the audiences. One audience GCR identified as the “bullish on SOA” audience was one of the segments. BEA hypothesized that this group would be positive on BEA, but we learned a great deal more that proved useful for high-value messaging.

Desh R Urs: My experience with Acxiom taught me that external data – if used incorrectly – leads to unsustainable inferences. For instance, the most-bought data elements are age and income. And what do people lie about the most? Age and income. If the data is often incorrect, how do you get to be market-driven and customer-centric? First, depend on secondary research when you can’t afford primary. Second, assign someone as the keeper of the knowledge. It’s important to keep looking at information. For instance, seeing that people bought flat-panel TVs at big box stores but cables at Radio Shack, it took some time to realize this trend was not about price, but about accessibility.

Questions

How do you get people in the company to look at segmentation?
SPB: Use the promise of market sizing – because you can identify the size of segments. If you didn’t have this data before, and are trying to forecast, you need the data. Also, be clear why you’re segmenting: Is it for developing messaging or for developing new products and services?

LC: When you have a limited budget you can’t afford to do broad-based advertising, so you need to focus and have the right messages for the focused groups. I was lucky because BEA was looking for that focus and internal audiences needed the data, so they became interested in and champions of the project, to be sure the end results would be useful

DRU: When we had a large budget, people were driven by fear – absolute, abject fear that if you don’t get buy-in by the executive team, they cut your budget. So decisions were driven by consensus, so that when a new product came out the sales guys knew what they were selling.

How do you proceed once you’ve identified the segments?
SPB: One of the deliverables research should provide is variables. Be sure to get these, put them into screenings for follow-up testing and tie into the customer database.

LC: Some hypotheses are really identifiable, like vertical, product used, etc., and you can validate and leverage these. Then tie the easy-to-identify attributes to unidentified attributes.

MySpace is now microsizing – what effect will Web 2.0 have on segmentation?
LC: We’re just touching the tip of the iceberg now, with the level to which people are collaborating. We don’t even know how social computing will be applied to the enterprise.

You can get data that are unreliable – how do you make data actionable and assure reliability?
DRU: When you’re selling business-to-business, it’s all about ROI. To validate information, use segmentation analysis and do differential campaigns into the marketplace. Segment and do call-to-actions and measure. You’ll have to learn very rapidly, from mistakes or successes, and adapt. Of course, you must have metrics in place to measure the successes.

How do you communicate the results to people who need the information?
SPB: To be actionable, data must live beyond marketing’s walls. In Intel’s case: new product development. We found the concept of personas to be one of the best ways to bring the segments to life. You know the personas’ needs, motivations, aspirations, barriers, successes, etc.

LC: Involve the end users from the very beginning. They have their own hypotheses and what they’re thinking already, their biases, can cause them not to believe what the research tells them. Those who didn’t share in the research, who have different biases, will have a hard time believing the research. The reason we did the research is for the “aha, we didn’t know that” reactions. It’s important, too to push the research out to higher, executive levels so they use it.

DRU: How you actually disseminate is important – it’s important to share with all. Some use a secure portal.

LC: Social computing with knowledge management – housed in the portal.

Does there tend to be a mix, one or all three of the outcomes of segmentation: messaging, changing how you sell, new products?
DRU: Each company has its own culture. Some are research-driven toward the financial outcome for the quarter. For others like us, customer attention goes across CRM, product, professional services. In either case, it doesn’t happen simply; it’s a combination of training, education, knowledge, really listening.

LC: We started with messaging, trying to identify habits like buying decisions, publications people reads, events, etc. What did the audience value? We also used research to extend into pricing, and as a sales tool: how segments could be addressed by the sales team.

What’s the lifecycle of research and segmentation – how long is it useful?
SPB: Hopefully, two to three years, although that really depends on how fast the market is changing – take the influence of mobility on all of us. You have to consider how volatile the market is, and whether behaviors will change based on market changes.

LC: The consumer market will be faster; really fast in the case of telcos. Enterprises are moving more slowly.

DRU: Applebee’s America by Douglas R. Sosnik, Matthew J. Dowd and Ron Fournier is a must-read for anyone interested in market segmentation and go-to-market plays.

Is there a maximum number of variables?
DRU: It’s practically infinite on the consumer side, on b-to-c. There’s lots of information available. On the corporate side, there’s a big landscape, and there simply aren’t 200 elements of data.

LC: Aggregating data together – groupthink – can reduce the number of variables.

SPB: Which variables are really distinct depends on what kind of segmentation you’re doing, but usually four, five or six for “back of the envelope” segmentation. With more complex, advanced analytical programs you can put in more elements. It’s still limited by logistical constraints.

LC: Identify the independent variables – which are the critical few? You want to boil it down to 3 or 4 critical ones.

PA: We think that 3 important ones are: resources (can they buy?), need (will they be motivated to buy?) and reach or accessibility (can we find them?)

Any final advice?
SPB: If you’re not doing it, even just back of the envelope, you should. It’s the key to figuring out the market landscape. If you do more segmentation, invest the time to do good background research so you know what the variables need to do. Be clear on the business objectives.

LC: Start early by just looking at your customers. Involve as many people as you can from the start so they buy in – because it’ll either be that they already know it or don’t believe it.

DRU: Arm the sales team with accurate, relevant data that they can act on.

Biographies

Moderator
Paul Abel, founder and managing partner, Blue Research
Paul Abel, Ph.D. is founder and managing partner of Blue Research, the region’s only market research consultancy focused on the high-tech, science and healthcare industries. For over 15 years, Paul has researched and surveyed a variety of audiences: from CXOs, to industry luminaries to engineers to middle managers. Blue Research services include focus groups, executive interviews, online and offline surveys. Paul teaches marketing at the Oregon Graduate Institute/Oregon Health Sciences University (OHSU). He holds a PhD in neuroscience/psychology from the University of Washington; BS degree from UCLA.

Panelists
Shauna Pettit-Brown, manager of exploratory market research, Intel
Shauna Pettit-Brown is the manager of exploratory market research in Intel's Health Research and Innovation Group. Shauna's role at Intel is to bring trends and market insights to bear on the development of new products for the Digital Health Group. Shauna has worked on a number of market segmentation efforts in her 10+ year market research career and recently completed a consumer health-oriented segmentation project. Prior to Intel, Shauna worked as a primary research consultant for Gartner Custom Research, specializing in qualitative research and new product development. Shauna graduated Summa Cum Laude from the MBA program at Babson College and received a BA in political science and economics from McGill University.

Laurie Cremona, vice president, BEA Systems
Laurie Cremona is a vice president with BEA Systems, where she and her job-share partner Elaine Miller oversee the worldwide services marketing organization. She is responsible for all aspects of marketing BEA's education, customer support and consulting services, including SOA Thought Leadership and BEA’s Global and Enterprise Account marketing initiative. Prior to BEA, Laurie served as director of marketing operations and intelligence for Global Field Marketing at Sun Microsystems, applying Six Sigma to design and implement management infrastructure and a market intelligence model. Laurie first practiced Six Sigma at General Electric, where she acted as a Six Sigma Master Black Belt for GE Financial Assurance. Laurie joined GE from Citadon, where she was both vice president of professional services and the Six Sigma quality leader. Laurie holds a masters of science in civil engineering and an undergraduate degree in architecture and urban design from Stanford University.

Desh R Urs, managing partner, iBridge
Desh R Urs is managing partner of iBridge. He brings over 20 years of entrepreneurial, startup and Global 500 corporate experience in sales, marketing and general management to the customers of iBridge. He has led sales organizations as SVP at Qsent, Inc. and VP at Acxiom Corporation, focused on usage of data in data distribution, direct marketing, fraud prevention and law enforcement. As a VP of global sales, services and marketing at Silicon Graphics, Inc., he managed engineering and non-engineering functions developing solutions in sciences, telecommunications, manufacturing, media, business and defense intelligence to support several billion dollars of revenue. During his tenure as VP at Think Tools AG and Brio Technology, he ran business development and alliances providing solutions in business intelligence and decisions cycle management to Global 100 corporations worldwide. In the late eighties, Desh founded Indus Systems, Inc. in Florida, which was acquired by a regional systems integrator. Desh serves on several advisory boards and is a member of the board of directors of companies in the United States and India.

About the author
Kate McPherron, a technology evangelist, has helped technology and industrial firms manage and market their products and services for 20 years. She can be reached at klm54@cornell.edu.