Imagine that you had paid a marketing company to do some customer segmentation for you and they came up with this:
You probably wouldn’t pay. An investor is in the position of having to pay you before you then go and build your company; your segmentation has to be better than this.
The simplistic view of segmentation is that customers can be divided by demographics. Gender, race, age, wealth, and so on. But your revenues and customer acquisition costs depend on customers buying things, not being things. So behaviour has to be the defining feature of a customer segment.
A clear example comes from pregnancy testing. There is a clear demographic target for manufacturers, namely women aged between 16 and 40. But behaviourally, Quidel, a manufacturer, separated out two different segments.
Hopefuls were women who wanted a baby. This segment was served by kits placed near fertility products in shops, branded Babystart, in colourful boxes with a photo of a smiling baby, and priced at $9.99 to reflect the inelastic demand that hopeful potential mothers have.
The product for Fearfuls was called RapidVue, in a plain white box selling for $6.99 next to the contraceptive products. This was for people checking they weren’t pregnant.
Same demographic, completely different reasons and behaviours, and thus two different segments with different products to serve them. Of course, the kit inside was identical in both boxes. (Use this as an example of proper segmentation, not of pricing strategy. Unethical pricing will always punish you in the long run, and it’s unethical).
Forming Segments
This is the first step, and it is done around two considerations:
- Why does a customer buy?
- How do they buy?
The objective here is to form groups of customers by behaviour, so that behaviour within the group is the same and across groups is different.
Profile Segments
Then you can determine the common descriptive features of the people within the groups:
- Who are they?
- What do they do? (in terms of their transactions, not jobs!)
The objective here is to identify the people behind the data, and the way they interact with your product or advertising, so that you can reach them more effectively and efficiently with the product or information about the product that is right for them.
Having profiled a segment, explain its numerical features. Number of customers, financial size, annual spend per customer on this kind of product, and anything else that is relevant.
Action
Only now should you determine what you are going to provide each segment, through the four Ps of marketing:
- Product
- Place
- Promotion
- Price
Doing this allows you to provide a use case for customers within each segment. How will they hear about the product and where will they go to buy it (Place)? Which product are they most interested in (Product)? How will they buy it, bundled or alone, and with what kind of frequency (Promotion)? What does this expenditure look like in terms of their available money (Price)? And as always, you will need some data to support all of this.
Prioritise
Once you have the plan of action for each segment, you can decide the order and priority for each one. This is based on the value each segment has for you, where “value” is used in a loose sense. Your real interest might be in selling cheap stethoscopes to multitudes of graduating junior doctors, but you might prioritise selling expensive ones to consultants first as you know they have a value in influencing your real target market.
For each segment you will need to do a new, mini, market breakdown, as the competitive forces are different. These breakdowns will both inform and explain your prioritisation choices.
Finally, you can decide for each segment whether they are a core customer, on whom you will spend the majority of your resources, or one that will you try to attract with a smaller amount of budget, or ones you will accept but not pay to acquire, or ones you simply don’t want. This labelling should be dynamic as you progress through your milestones.