Are we solving problems that don’t exist?

Are we spending too much time on made-up problems that will never succeed as they are only imagined?

Remko Vermeulen (1) is credited with saying “75% of startups fail because they are solving problems that don’t exist.” Could this be true for every business, not just startups? If maybe not 75%, then still a significant 35% or 24%?

How can this happen? How can we so consistently invest in problems that don’t exist?

1. It’s not you it’s your model

Every decision is based on an assumption of what the world looks like. Nobody has an exact 1:1 model of the world in front of them when they make a decision, and so we use a lens and a filter to fit some information into a narrative of the real world which we would like to be true and that we could act on. We call these mental models (2).

“Mental models are how we understand the world. Not only do they shape what we think and how we understand, but they shape the connections and opportunities that we see (2).” — Farnam Street

In order to make decisions together, we sometimes externalize our insights and understanding into models that visualize and/or organize the world, for example, customer journeys, business models, personas, SWOTs, etc.

The challenge with models though is that they are simplifications, they have to be. They filter out information that might be significant, and they do this without us even noticing.

Because of this, every model has a bias. It decides for you which information is important and which is not (3). And if your model is wrong, your perfect solution will be solving a problem that exists in your model and nowhere else.

Your perfect solution will be solving a problem that exists in your model and nowhere else.

2. Customer models are the riskiest models

In 1898, E. St. Elmo Lewis published the AIDA model (3). It is a linear, sequential, hierarchical model with little empirical support and has been found to be a poor predictor of consumer behavior (7). Still, it has garnered massive amounts of attention and use, even as late as today, 126 years and several technological and societal revolutions later.

I assume the reason it’s popular is because it’s simple. Because we want the world and our work to be simple even if it’s not. But simple models are worse because they increase confidence and reduce the desire to challenge our own assumptions and learn:

“If our simple stories shape the data we notice and the way we package it to ourselves, our automatic feeling of rightness creates a lock down in our information search” — Jennifer Garvey Berger, Unlocking Leadership Mindtraps(4)

In 1997, McMillan and McGrath published the article, “Discovering New Points of Differentiation” (5). It introduced the Consumption Chain or what was later adopted as the Customer Journey. This model gave organizations an opportunity to articulate their own narrative of what they assume their customers are doing.

The company is both the investigator, imaginator and narrator of the journey which with almost robotic beauty structures the customers into a tidy process where they fold into lines and with the perfection of British queue culture walks cogly through the process and into the pockets of the company. — author

The customer journey visualizes the customers’ path to purchase and greatly simplifies a teams process of finding problems to solve.

Unfortunately, it’s hard to confirm that it’s a fair representation of reality.

As an example, in 2012, Google and Shopper Science investigated 3,000 paths to purchase and found that not even two customers were following the same path (6).

For example, a customer journey might portray the customer as living in a singular world with a simple problem to be solved on a linear path through a nonexistent problem. From a customer’s perspective, nobody is "buying, onboarding, and churning" products. They are Janelle after a long day at work wanting to sit down, relax, be mindlessly entertained and maybe rewarded with a laugh, chocolate bar, or glass of Pepsi.

In one of my own analysis a few years back I found that our customers were influenced by 35 different forces as they were making key decision related to a product purchase. The company I was making the analysis for wasn’t even present in the decision itself (even if all former models had reflected that they were present and sometimes even the only force of influence).

Often your customers are not at one place in a journey, but many places at the same time in many journeys (everything, everywhere, all at once). And so the logic of flow and maturity breaks down. As in: how many future travels are you planning right now? Or if you are a physician with Diabetes patients, how many patients are you treating right now (the answer might be tens or hundreds)?

3. A model paints the picture we want to see

Most customers aren’t customers at all; they are people (7). For example, at night they go to sleep, many of them on the same mattress for years or even decades. They use the mattress every night to rest and get re-energized. Some have back pain, some are stressed, some sleep together, some alone, some can’t sleep, some spend days in bed, some eat ice cream, some copulate, and some watch TV. The least thing they are is customers of new mattresses. And when they are what’s most important to them? It's not buying a new mattress, but using it.

Most customers aren’t customers at all; they are people

If the model changes the narrative from what’s important to the customers to how the business gets paid (the output of a transaction), it might still be the right model. We just need to be absolutely certain that it is — especially if we are dependent on our customers and want to make sure we are solving problems that exist for them.

4. Isn’t there an easy fix?

Yes, there is: choose a model that fits and experiment to make sure that it does.

A. Choosing the right model

“Data precedes framework.” — Dave Snowden (8)

The biggest challenge with models is that we choose them — they don’t choose us. We are familiar with a few which we tend to use over and over without assessing if it is the right one or not.

Having good knowledge of models (you can read more here) and having a process to assess what model has the right fit is a good start.

The biggest challenge with models is that we choose them — they don’t choose us.

Personally, I often use a causal loop diagram (9) to collect and map what the organization thinks about the topic we are investigating. This is a decently neutral way to identify the causes and effects that are assumed to exist and how they connect. Then based on the system that emerges we find the right model to fit, testing it along the way to make sure we are on a good path.

B. Experiment

Not every decision maker acknowledges that they are making assumptions; some see their synthesis of knowledge to be fact. By doing this, they undercut the organization’s ability to learn.

“If managers ‘believe’ their world views are facts rather than a set of assumptions, they won’t be open to challenging those world views” — Peter Senge, The Fifth Discipline (10)

The approach Riskiest Assumptions Testing (11), tries to solve this. It helps the organization ask and discover “what has to be true for the problem to be true?”. It produces a list of assumptions that the organization can prioritize, assess and test (12).

“The traditional model’s strategy is to focus on the question: What is true? A more effective model for framing and making strategy choices is to focus on the logic behind the choice by asking: What would have to be true?” — Roger L. Martin, A new way to think (13)

C. Simple is better, right?

While a model filters the signals from the noise and gives us clarity and confidence, it can also remove curiosity and willingness to learn.

The simpler our model is, the more important it might be to test it. Simple is not a sign of quality (14), it is only an emotional sense of comfort in the heart of the model makers and users.

“Your sense of being right about something, the sparkling clarity of certainty, is not a thought process, not a reasoning process, but an emotion that has nothing to do with whether you are right or not.” — Unlocking leadership mindtraps (4)

So, the simpler your model is, the more important it might be to test it.

Summary

How do we know if we are solving a real problem, an imagined one, or an unimportant one? We have to do the following:

1. Map out how we understand the world by externalising the influences and connections (between them) we know that leads to the outcome we desire.
2. Identify the key underlying assumptions that have to be true for our map to be true
3. Test them
4. Adjust the map

Sources

(1). Remko Vermeulen, https://remkovermeulen.com/

(2). Mental Models, Farnam Street, https://fs.blog/mental-models/

(3). AIDA-marketing, Wikipedia, https://en.wikipedia.org/wiki/AIDA_(marketing)

(4). Unlocking Leadership Mindtraps, Jennifer Garvey Berger, https://www.cultivatingleadership.com/book/unlocking-leadership-mindtrapslinkedin

(5). Discovering New Points of Differentiation, HBR, https://hbr.org/1997/07/discovering-new-points-of-differentiation

(6). The digitization of everything — its impact on the buyer’s journey and marketing’s role, https://www.sas.com/no_no/insights/articles/marketing/digitization-of-everything-buyers-journey.html

(7). A customer is just a body with a wallet attached to it, https://everythingnewisdangerous.medium.com/what-thoughts-do-you-wish-to-create-383a845f29e8

(8). The Origins of Cynefin — Part 2, https://thecynefin.co/part-two-origins-of-cynefin/

(9). Causal loop diagram, https://www.transentis.com/page/causal-loop-diagrams

(10). The Fifth Discipline, Peter Senge, https://en.wikipedia.org/wiki/The_Fifth_Discipline

(11). The MVP is dead. Long live the RAT, Rik Higham, https://hackernoon.com/the-mvp-is-dead-long-live-the-rat-233d5d16ab02

(12). This article includes a guide on how to break your problem into assumptions to be tested. https://everythingnewisdangerous.medium.com/how-to-write-a-customer-value-proposition-part-3-testing-it-and-personal-learnings-93aa90fdf434

(13). A New Way To Think, Roger L. Martin, https://store.hbr.org/product/a-new-way-to-think-your-guide-to-superior-management-effectiveness/10565

(14). Murray Gell-Mann argues that aesthetics is a sign of a good model, https://www.youtube.com/watch?v=UuRxRGR3VpM