Ten out of ten start-ups don’t segment their markets well or at all.

Oddly, a number of temporarily successful companies don’t either.

Segmenting is one of the essential marketing strategy disciplines and yet the least practiced. Many start-ups skip it thinking that they intuitively know their market segments or, worse yet, that they will try-and-pivot their way to success. Of the companies that at least take a stab at segmenting, they often do a poor job. One of our earliest clients – a post-IPO success story facing an industry shake-out and forced pivot – confided they had chosen an industry verticals segmentation plan because “It’s as good as any other.”

Such lackluster segmentation sentiment is dangerous because segmentation is the foundation for your go-to-market strategy, and the one that accelerates early revenue while solidifying a survivable niche from whence to grow. Failure to segment is tantamount to failure. Let us list where people most often get their segmentation schemes wrong and what they can do about it.

Bad vector, Victor

Vectors are the quantum used to measure the most meaningful criteria for product judgment within your total market. Vectors can be anything, which is why most people choose them poorly by defaulting to well-known but irrelevant vectors. Were industry verticals the right vectors for the client mentioned above? No, but software development platforms and mission criticality of applications were. Finding the right vectors is not trivial, which is why we at Silicon Strategies Marketing have developed a six-point check list for segmentation model viability (though two of the check-points are not always applicable to certain industries).

This check list helps with the other important factor in defining a segmentation model: assuring that it is sane. Some companies will create a segmentation model that fails due to each sector not being of a viable definition. For example, one of our six check points is determining if participants in a segment freely communicate to one another. A segment where buzz cannot be generated is not a segment and cannot be dominated. Segment modeling must be validated or the segment model is likely invalid.

Keeping score … cards

When segmentation is defined, knowing which segments to assault – and which ones not to – is critical. Entering a saturated segment with strong competitors is suicide. So is entering a large, uncrowned, yet unprofitable segment.

Score carding is the process of defining a rational scoring system for your specific segmentation scheme. Some of the things that can be scored are segment size, sales and price points, time to sell, segment saturation, competition and whole product fit. Many of these factors then model sales likelihood, sales cycle duration and sales costs.

There are two reasons why many companies don’t score card their segments. First, sore carding is slightly more painful than amputation without sedatives. It is also very market and product specific, with a great number of assumptions at play. Yet not performing this degree of diligence leads many into the wrong segments, which in turn leads them into bankruptcy.

Improper priorities

Lacking a good segmentation model and/or a good segmentation scoring system then leads companies into poor prioritization for segment acquisition. Not that this matters to most start-ups. They are convinced they can sell to their entire total market and all potential segments, so they skip the prioritization process entirely and leap head first into the Silicon Valley dead pool.

Not prioritizing is a twofold problem. First, weak or non-existent score carding leads to unclear opportunities – all segments look equally attractive. This in turn can lead a company to attacking too many segments at once. In our recent survey of founders, over 30% targeted three or more segments which is a near guarantee of start-up failure (indeed, an equal number of respondents said their companies either had a short run or fell directly into insolvency).

Other companies do not incorporate segment-level competitive analysis into their score carding and thus enter huge segments to fight giants. A far better strategy to enter an empty though smaller segment and dominate it thoroughly, earning fame, buzz and competitive protection.

Perhaps the saddest outcome from poor segmentation is the chasing of unicorns or dragons. Unicorns are those fanciful dreams of the perfect product in the most desirable segment. Such day dreams lead to unrealistic assessments of potential and calamity. Dragons (as the “chasing dragons” drug abuse metaphor alludes to) are Moby Dick obsessions, typically held by founders and their desire to be a player in one or another segment. Founders have been known to disregard even great segmenting data to chase what they can never catch.

Summing segmenting

Segmentation is a science, a discipline and difficult. Yet it is something you don’t want to leave to a whim. Do it, because short of selling and ringing the cash register, it is the single most important step you can take to prevent disaster.