“I am forty. I have wrinkles and grey hair and would like to have a walking stick to help me walk.” If you’re in your thirties or forties, these comments from 8 year-olds are pretty funny.

They’re funny because — hopefully — most 40 year-olds can walk up and down hills without feeling they’ll “soon die”.

For an 8 year-old, the experience of a 42 year-old is impossibly remote. 8 year-olds have no “projectable experience”, which is one reason why we don’t employ any under 10 year-olds at Fluxx.

But when a 35 year-old designer is asked to consider the needs of a 65-year-old customer, are they any more insightful? With only stereotypes to rely on, what laughable prejudices might they display?

Similarly, older people can perpetuate stereotypical attitudes towards the young. I can’t say for sure that recent attempts to interest ‘the youth’ in the EU referendum debate were not originated by a crack team of millennials, I’d be very surprised if any were actually consulted ahead of the campaigns creation.

Segmentation and user personas, based on good research data, are good tools to help fast-track product teams into the mindset of consumers — to help them understand the attitudes and behaviours of ‘typical’ customers and identify new opportunities.

However, segments and personas often become too stereotypical — detached from the rich and messy reality of everyday life, and this can really limit our understanding of other people’s experience and context.

Lazy gender stereotyping continues to lead to ‘Pink it and Shrink it’ product design for women and ‘Large It’ for men. Occasionally people manage to mix both together in a complete and unholy mess, as achieved by Bic last year .

All of which serves to illustrate the observation by Jessica Ivins that, ‘When we design with gender in mind, our instinct is to resort to extremes instead of solving real problems’.

Obviously, trying to ‘identify’ at a distance with a large and diverse group of people based on age or gender leads to no real identification at all. However, if you’re willing to swap the clinical stereotypes for immersion in the messy business of observing real life, you might just stumble on genuinely desirable solutions to people’s problems.

Tesco’s Click & Collect scheme fitted into the lives of parents who liked the convenience of online shopping but didn’t want the inconvenience of having to wait in for deliveries. Parents could combine collection with other tasks such as the school run.

Based on this, Tesco looked at how to extend the value of collection to this group elsewhere in their busy lives, so experimented with mobile collection points (essentially a Tesco delivery van) at places like junior league football matches.

One of the most popular collection points was at the Portree on the Isle of Skye. Customer loyalty is strong enough that there is now a 2,150-member Facebook group called “WE WANT TESCO IN PORTREE, ISLE OF SKYE”.

Tesco was looking at the edges, a little bit away from the mainstream, just out of the corner of its eye.

The edges are messy and nuanced and, well, interesting.

The edges tell us things that might be hidden from us (possibly in plain sight) by more sanitised accounts and representations of human behaviour.

In other words, the edges are real. People behave in ways that it would be hard to predict without doing the research.

We recently spent time at the edges trying to understand how homeowners might use a diagnostic kit intended to help them to get the most out of their home heating.

It turned out the real world was very different from the one we imagined before we went out into the field.

A clever high-tech product to tweak your heating system meant very little for a busy, house-proud mum, spinning plates and in a home occupied by kids and pets with towels and wet clothes draped over every radiator.

One analytical, efficiency driven, data-loving economist we met loved the idea, but in a freezing/baking flat with huge windows and poor insulation, subtle heating tweaks just won’t cut it.

The idea is still a good one in theory. But real life isn’t theory. So we step out into the real world to develop the insight needed to make a good theory a great product.

At Fluxx, we do things like spending hours on the road with plumbers, fixing toilets from Walsall to Westminster, handing out leaflets on windy commons and setting-up stall at pet shows to meet with pet owners.

We once spent three days trying to sell a fictional Smart Home proposition at a massive consumer home exhibition.

We gamble our own money to experience the highs and lows of online gaming; and joined a knitting group to better understand the interests of Woman’s Weekly readers for Time Inc.

When Argos gave us a simple challenge “Get to know our customers and colleagues”, we had a few choices: 1 ) read some market research, 2 ) put out a survey , or 3) go and work in some Argos shops.

So we spent a week working in a variety of Argos stores from Edinburgh to the Old Kent Road, just like any other shop colleague. We proudly donned our blue shirts, Argos name badges and health and safety training, then met a lot of fantastic people on both sides of the counter, we assisted with about £12,000 worth of sales, and even sold one extended warranty. But more importantly, we walked away with dozens of real life stories and conversations that we could use in the work that followed for us.

One customer walked into the store carrying a blue bucket filled with shattered glass. “This is my shower screen. It just fell apart.” they said.

Talking to people is always mind-expanding, a reminder that their experiences are unique and often impossible to fit into neat boxes and categories. It’s this real life experience that allows us to bring a segmentation analysis, or a big piece of quant analysis to life — to put real names and faces to otherwise anonymous data — and therefore, give ourselves the best chance of really understanding what is happening.

This might make us seem like oddballs at times, and it definitely takes some nerve to repeatedly leave behind the comfort of desk research and pre-existing segmentation but it really is the only way to generate interesting insight into other people’s lives.

Dean Wilson is a Senior Consultant at Fluxx, a company that uses experiments to understand customers, helping clients to build better products. We work with organisations such as Lloyds Bank, Royal Society of Arts, the Parliamentary Digital Service and William Hill.

If you enjoyed this, you might enjoy “31 Things We Learned in March 2016”