I frequently use segmentation approaches to help my clients make better decisions about where to invest their media dollars. I developed a knack for this about 20 years ago when we replicated an attitudinally driven segmentation that a client had conducted and found that we could align factors from our database with theirs and recreate, with good confidence, the segments they uncovered. What resulted was a clearer picture of who our prospective customer was with a direct linkage to their media habits. Media planning and buying has never been the same for me. You see, we’re brought up in this business using demographics as target audiences. As soon as a client briefs you on a project you run off and look at generic media usage of a demographic grouping. But ever since this experience I’ve become jaded about using demographic targets exclusively. Don’t get me wrong, I know when it comes to negotiating broadcast buys the currency is a demographic target because that’s what Nielsen measures. But we can plan and select vehicles that align with the mindset/attitudes/behaviors of our segments. And if we select media this way we’re doing a better job for our clients in investing their marketing dollars instead of simply spending them.

Demographics alone are a poor indicator of whether or not someone will buy your product—with rare exception; point-of-entry for denture and the feminine hygiene product categories? You can say that most people who buy life insurance are over 35, but do they wake up on their 35th birthday and say “I must buy life insurance today”? Of course not. There’s another trigger. Maybe someone their age passed on and left a family without financial independence. Maybe they had a new child or a child enter a new lifestage. The fact that these people are over 35 has less to do with the decision to buy life insurance than the emotional factor. Their demography is happenstance. Planning and buying media on their demographics alone will lead to poor decision making—it’s a blunt instrument.

There are multiple methods and statistical techniques you can use to create a segmentation. Many organizations offer ‘off-the-shelf’ values based segmentations that predefine groups based on prevailing personal values and characteristics. VALS and PRIZM are great examples of this type of segmentation. PRIZM uses socioeconomic factors more than psychographic ones so it doesn’t fall into the same space as VALS. However these approaches fall short in some areas. That’s why I prefer to use a behavioral approach that is custom built based on the way people interact within a specific category. I don’t think people are as one-dimensional as the values based segments suggest and they act differently in different categories because of that. The chart below shows you the main difference in method here:

Both are markedly better approaches to developing media plans but will yield different outcomes. For example, using a behaviorally based approach may result in a segment of the population who are non-users of your category. It doesn’t matter what these people’s values are if they will not buy your product. The behavioral approach also gives you an opportunity to identify size of business value (dollars and cents) of different groups.

What’s your method?