Learning to Look Outside — Discovering Uncrowds & ROI

Jul 6, 2020 · 5 min read

This article was originally published June 25th, 2018

If you’re going to invest in anything, especially if you’re going to be among the pioneers using a new metric; it doesn’t matter how much theory and passion the vendor shows, you need to have either nerves of steel; or you need robust ROI models to make that investment a watertight win.

We’ve thought endlessly about those models since the creation of Friction/Reward Indexing (FRi™) and finally we have them. It took a total change in the way we thought about the challenge to crack this one. Our solution is simple now we have it but as it is so often, the road to simplicity was paved with twists and turns, and overcomplications.

First let’s look at how FRi is built from a combination of understanding the maths of shopping preference at four levels:

  • - Channel*
  • - Touchpoint
  • - Shopper mission**
  • - Need-state combos

Short version is that we can tell you which retailer a brand-sensitive, premium-preferring, localista, moody-bollocks shopping for a canoe on their phone is most likely to choose; and we can tell you if you are more or less likely to be chosen and why. Doesn’t have to be a canoe, doesn’t have to be those need-states or on a phone but we can tell you.

So here’s where the twist in thinking solved the ROI challenge; every time we looked inside the product, we got more and more stuff out of it but each layer of new abilities made the models more complicated. So one afternoon, frustrated and grumpy and as a last resort we put down the product and instead looked outside of it. That was the eureka moment, that was the breakthrough because by ditching our needs and instead going back to customers, retailers, and shopping, we could see clearly again.

What did we see? That need-states don’t live inside models they live inside customers. We saw that we could identify all the common combinations of these need-states independent of anything else and START with these clustered groups of need-states instead. Once you know how many of these clusters exist in the world and once you know roughly how many people are in each of them at any given time then you can easily build ROI models from there.

Even better, we found that most clients already had their own research on need-state clusters without always realising it. They often already knew things like ‘12% of our customers are price-sensitive and time-rich when they interact with us’,

Even better is that these clusters are of need-states not people; it’s not ‘Billy Smith is 41, drives a Mercedes and is so brand-sensitive it hurts but has no time to shop’ because sometimes that’s not true of Billy Smith, he doesn’t give a monkey’s about what brand of dishwasher tabs he uses and he spent four days recently researching and choosing a new bag for work. We’re lying to ourselves whenever we try to pin-down individual customers as fixed and knowable people, or when we use sociodemographics to attempt to group Billys together, or worse when we apply attitudinal and persona segmentation to a customer base and expect that to be particularly useful beyond providing the notional shred of comfort that we have a clue who our customers are.

“Marvellous Maureen is 34, likes to whistle when others frown, owns a dog called Peanut and once dated Ed Sheeran’s Dad. No! Maureen is a symptom of a marketer’s growing insanity and it is Marvellous Maureen who will visit them in their dreams at the moment of their death…”

What we do know is that over the entire UK population, there are roughly 27 useful need-state combinations, that six of these are very common, and we know broadly how many people at any one time are in these clusters.

We also know that need-states are independent of all the usual segmentation; two people with similar jobs, the same education, in the same postcode and with the same family structure and even from the same political party and the same age can be in wildly different moods with wildly different need-states at the same time. In terms of how they shop, even when shopping for the same things, they are more different than they are similar.

You don’t need to be in a tribe to share need-states for a coffee.

Our population need-state clusters capture what makes them the same and does so at the most important time: when those customers are on their shopper missions. That changes everything; these are clusters of needs and emotions that you can sell and market to. Because these groups are absolutely NOT tribes; we call these 27 clusters the Uncrowds.

A hairy arsed motorcycle gang member can share exactly the same mood and need-state when buying themselves a new knife as a Vicar might when considering getting some nice new vestments. They are resolutely not in the same tribe but they are an Uncrowd and you can create customer experiences that win the preference of each once you know that their Uncrowd is one of your Uncrowds.

And so from that it becomes easy to build ROI models for Friction/Reward Indexing investment — once you have identified your key Uncrowds, then using FRi to win more customers in each of them becomes a straightforward KPI to which you can attach chunky ROI numbers. FRi identifies what each of those Uncrowds considers friction and reward in each channel for each shopper mission. That’s then your map to building customer experiences that win you increased preference in those Uncrowds. The right CX for the right Uncrowd, increase preference by 10% and BOOM, there’s your lovely fat ROI.

We love reducing friction here while delivering big reward so get in touch and we’ll build an ROI model especially for you that shows how much you’ll gain from thinking in Uncrowds and using FRi to win their preference and favour.

* God I hate that word but common usage forces me into it
** ARGGHH, THEY AREN’T MISSIONS, NOBODY THINKS LIKE THAT but that’s what the industry calls them so we have to

The Uncrowd

After the customer analytics herd there is the…