Rethinking Segmentation to Discover Future White Spaces

It is 8:30 in the morning. People, half-bleary, stumble into a virtual meeting to discuss what people want. Before the discussion progresses, someone raises a question about what people want is reflected in their current choices, is it not?

The answer is both yes and no. Current choices are a function of not just what I want today but what are the alternatives available to me at the moment of choice, within the parameters of my own constraint. Let’s take the example of mobile phones. Forty years ago, if people needed a phone, they would have chosen landline, home connections. They did not have an alternative! If marketers had asked them, are they happy with what they have, some may have grumbled about things like connectivity, fees etc. but most would not have been able to say that they needed mobile phone. If market researchers would have shown them a prototype of one and listed the benefits of mobility, I am reasonably sure majority would have been flummoxed by its utility but a small percentage would have been interested. And sometime around the turn of the century, marketers who understood this cashed in on mobile phones.

 

Current choice was definitely not equal to desire then.

And therein lies the explanation of yes and no that I mentioned at the beginning of this piece. Most of us are, in general, not very articulate when it comes to what is it that we actually need and want. We can answer within the bounds of what we have been taught and seen (including marketing comms) but very few have the ability to lucidly explain what they actually desire. Mistaking this inability to articulate as the absence of desire for what mobile phones could do has been the downfall of many companies – from telephone to camera.

 

Question, therefore, arises how do we as marketers get to what people desire.

It is through choices which people make but not in the context of the current market construct. Choices that they make when we present them with options for the future plus the current. An example:

I have to attend a meeting which is one hour away by road. It is important that I attend it in person. I have to choose from the following modes:

1. Take an Uber. They charge a premium and offer evidence of sanitization before I board
2. Take the metro. It is fastest and I have a ready pass with free rides left.
3. Take my own car. I can drive in comfort and won’t need to rely on anyone else

If I make the first choice, it tells you that for me sanitization and the protection it offers are more important than the price.
If I make the second choice, it tells you I prefer efficiency and savings.
If I make the last choice, it tells you that I prefer not depending on others for my comfort.

Perhaps the choices above are rather straightforward and unidimensional but they do allow you to understand my needs and desires.

Now, let’s add another layer. Let’s assume that I do not own a car. So, if the above situation arises in real life, I will only be able to choose between options 1 and 2. But, if you were to present to me a future scenario, unconstrained by current alternatives, and I choose option 3, it will tell you what I desire, even if it is not my present reality. It tells you that in the future if I can, I am inclined to buy a car. For a mobility company, it is important to know how many such people exist to decide future strategy.

Some marketers tend to argue this works well for aspirational categories like cars and phones but would it apply to more mundane categories like, say bathing products?

 

Let’s do this exercise again:

I have to bathe in the morning before going to work. I have the following options:

1. Bathing crystals that lather when dropped in bathtub providing me with a relaxing bath while it cleans
2. A moisturizing body wash that does away with the need to apply moisturizing lotion after a bath, providing complete skincare
3. Shower gel that doubles as shampoo cleaning me from head to toe

I choose option 2. It tells you that I desire efficient and complete skincare but I do think that skin and hair need different products.

Option 2 is not a real product in my country. But it tells you that there is a need that is unmet. If you had directly asked me what I do I need from a bathing product, I would have answered with something generic like cleaning, lather, and fragrance, maybe. As a regular consumer, I am not going to imagine what the possibilities for bathing products are. It is the marketers who can imagine and it is the job of insights partners to tell which of these have any future potential and why.

We have now been using the above philosophy very successfully for clients across markets and categories. Using choice design, we ensure that a large number of alternatives/options like those in the examples above – comprising a product solution / reason-to-believe, functional and emotional motivation – can be shown across samples.

 

We also build a second level of nuance in the questioning, through the rather versatile qualitative technique of laddering.

 

Figures 1 & 2 are masked examples from an actual study that we conducted for a client who had been shrinking consistently in a big market and past attempts at segmentation – from attitudinal/psychographic to big consulting-led – only yielded the current market construct with little indication of how the category would evolve in the future.

The choice data thus captured is used to cluster/segment through machine learning techniques Singular Value Decomposition. This kind of technique is used in image processing.

Example:

 

Refer to Figure 3. The original image has a resolution of 400 X 400 pixels, which means that the image contains 400 unique rows of pixels or row vectors. Singular Value Decomposition selects fewer number of rows among the original 400 which approximates the image as prominent as possible. So higher the number of selected rows clearer the image will be.

This dimension reduction property of SVD is adapted as a clustering technique that:
• Incorporates relationship between attributes
• Non-fuzzy, discriminated segments

Through this approach, we are able to clearly give sense of how the category and spaces are going to evolve in the future – indicates where the people want to go, unconstrained by what is available in the market. As for our clients, they have been busy revamping their brand growth strategy based on these future white spaces to stay ahead of the curve and own the future.

Photo by Phuong Tran on Unsplash