Segmentation is tough, but here’s a practical approach

Large companies spend crores of rupees doing research to fine-tune appropriate segments in a market that is increasingly getting fragmented.

Segmentation is in everyone’s mind whether one is in marketing or manufacturing. Marketing gurus from Kotler to Ries stress the importance of segmenting customers and markets for better strategic thinking. I advise most of my clients to undertake an intense session of segmentation. It is a critical aspect to drive business growth.

SMEs across the world have probably done some basic level segmentation of their markets and customers, leading to some broad strategic planning for targets such as volume and profitability growth, market share, etc. Large companies spend crores of rupees doing research to fine-tune appropriate segments in a market that is increasingly getting fragmented.

However, many marketing folks find it hard and at the end of a cumbersome segmentation exercise, they are still left with the question, “Now, what do I do?”

How do you make practical the segmentation journey? How do you make use of the results beyond just a document for internal consumption? To really deliver actionable findings from the segmentation work, you must evaluate a number of possible dimensions.

• Consider both current and potential customers or targets. Any practical approach to segmentation must have this consideration, to start with. The segmentation exercise will be limited to those chosen for the survey. Since it is not feasible to survey a larger and endless population, considerable pre-thinking should be given to selecting samples to research. Most segmentation work gets limited to "category users," but other sub-populations like lapsed customers or non-users need to be considered too. A discriminant function analysis (any market research agency would help you understand this) run on these sub-groups will help discover the degree to which they hold attitudes, values or needs similar to those in key segments of the whole market. Those who do are more likely recruits for a brand.

• Identify and freeze a reasonably good sample size to allow for a natural segmentation. Any segmentation exercise, if done right, can lead to the identification of a segment that may be as small as 10 percent of the target group. Without a sample of respondents large enough to provide full details on a small segment, you may end up with the wrong profiles without significant differences. If so, and if this smaller group happens to be one of your critical segments – one with a large proportion of current or anticipated users – the value of the research outcome is seriously diminished.

• Compare, profile and categorise the selected segments to the remaining population. Researchers try to profile the segments against the taken sample in between the full research. This approach is not optimal since it may reduce the potential differences between segments. You should instead examine the segment against the balance of the population by asking, "How is this segment different from the rest of the target?" An analytic approach will help define the differences more clearly. The behaviours and desires can thus be more readily identified.

• Conduct significance testing at several checkpoints. Statistical significance testing is a desirable part of all good research in order to take bold decisions. Choose a more stringent level for what you want to profile the segments. Also, analyse the data at a more qualitative level too to bring out aspects where differences may be subtler and more difficult to recognise. Even if the agency is not certain of multiple level significance testing, you should insist on these rules set early in the project. When you spend a lot of resources doing this work, get the right insights.

• Propose product/service attributes for better appeal to different segments. Your offering attributes that are persuasive to one segment may not be appealing to another. Your researchers must identify specific combinations of taste, styles, colour, features, price, etc that optimise appeals among all targets. Make them aware of this requirement.

• Rank and prioritise the segments. You must have a clear sense of priorities for segments that are most critical to current and future growth. The ranking should be based on all available information from the research. Also from current purchase patterns and use situations, potential for future buy and loyalty, positivity to your offerings currently, size of the market, and similar or "like-minded" attitudes. These priorities shall dictate the course of short- and long-term future efforts against the segments critical to the growth.

Segmentation is always tricky. The power of segmentation is when all stakeholders with their expertise plan to hold workout sessions with the research team to make them understand your business goals. Take the responsibility to ensure that the results are not only insightful but also practical.