What constitues a good startup problem

It seems deceptively simple to define what constitues a good market problem to work on as a startup. It is actually rather subtle and for some part very counter-intuitive.

startup problem - innovation copilots

Most of the startups participating in one of my trainings are initially shocked at the inordinate amount of time I spend working on ‘the problem’. I’m certainly not alone there. Everyone who is regularly dealing with startups gets eventually frustrated to see how they concentrate en masse on building a product and not focusing on what the market actually needs. And while anyone who ambitions to shake a market’s status quo shouldn’t be too pragmatic, as much rationality as possible should nevertheless prevail. But, very few are the startups committed from launch to tackle a clear-cut problem.

When pushed, they often go for fuzzy and wishful. They also prefer explaining their added value; the positive change they will bring to us all. Decreasing price, making better products, or enhancing satisfaction are the copy and paste mottos for all. Few realize before it’s too late that the price / quality / service approach is a huge misstep because it conveys ZERO information to the market. Who’s going for helping customers waste their money or make their life miserable?

When no one can say the opposite of what you claim, you say nothing. And don’t get it wrong: this is not a marketing problem. It’s a strategy problem; as in ‘you don’t have any’.

On the contrary, the benefit of starting with a clear cut problem for a startup is essentially that you lay a solid platform for managing the development of a solution that will have to pivot a few times. You can’t confuse anyone when you broadcast what you want to solve. While saying what you want to achieve, not knowing yet exactly how you’ll achieve it, is irresponsible (or at best very naive).

That being said, finding a good problem to solve is not trivial.
What does “good” mean, right? Well, it’s such a central subject that you can spend days reading all forms of literature on the subject.

For what it’s worth, here’s the qualities of a good startup problem in my +10 years of experience on the subject:


startup problem innovation copilots

1. There is a key metric and a demography

A good problem to solve could be described with its starting point(s), its perimeter, its nature, its evolving spread and reach.

The evolving part is tremendously important for scaling up the business obviously. It is most of the time a demographic element pointing at the underlying drivers of the problem which ensures that the future market you hope for, won’t disappear in the next ten years. It’s not a fad. As such, they are reason enough to take concern and invest ressources and cracking the problem.

These descriptive elements (among which demography is so important) can be usually wrapped around a central metric. Even if it’s an unfair or indicative one. The central metric used by the startup to frame the problem speaks volume about the business and technical acumen of the team. This is about the vision.

Let see what this all means with a typical example (the numbers are bogus, they are just there for the sake of the article):

Above 250 beds in EU hospitals, proper medical attention cannot be guaranteed to patients and nosocomial infections prevalence is doubled.

This is a metric that explain a critical point past which the market problem appears sharply. It also fits in a tweet.

Such metric may encapsulate the notion that at some point, corners are cut on hygiene, doctors don’t have enough time per patient, medical files start to get mixed up too much, etc. We don’t know yet but we can ask and explore the why. Observe how we’d actually love to challenge that. Why 250 not 500? Why the number of beds and not the square meters or number of nurses per patient?

See how strategy nests marketing, not the other way around?

Now for the demographic element:

Since the last five years, the number of beds in regional for-profit hospitals in Europe has been increasing 25% every year with a current average number of 380 beds.

(Still bogus data.)

Now we have a key metric AND a trend. We highlight a specific subset of potential target customers within a foreseeable value-chain (said hospitals, regional administration in charge of healthcare, doctors, families). There is no business plan yet, but the structural asumptions are laid.

The last step now is getting to the actual problem:

Every year, more people die from infections acquired in hospitals, than by car accidents.

(These are the ‘nosocomial’ infections, and this is not bogus.)

Say that in reverse and you have a crystal clear pitch for investors:

Every year, more people die from infections acquired in hospitals, than by car accidents. Since the last five years, the number of beds in regional for-profit hospitals in Europe has been increasing 25% every year with a current average number of 380 beds… (insert dramatic pause)… Past 250 beds, an hospital cannot insure proper medical attention to patients and nosocomial infections prevalence is doubled.

It’s a bit wordy and can be refined. It’s also a tad technical… But we’re in the healthcare market and that complies well enough with the corresponding business and technical culture.

This first point that I take time to explain over and over whenever I have the opportunity, also shows something rare and incredible: a startup that can come up with this information has probably done its damn homework!

Let’s go now to the second key point that defines a good startup problem:


startup problem innovation copilots

2. The framing is polarizing

If I follow-up on my previous example, you have realized that we can discuss strategy without immediately getting on-boarded on what solution the startup want to develop, its technology, etc. Is it an IoT infection monitoring tool, a new ERP system for hospitals, online training… ? We don’t care yet. That’s a good problem to work on. And if it’s not good enough, we also have the option not to waste time discussing further.

But the framing is also polarizing.

In my (relatively fictional) narrative about nosocomial infections, many shortcuts were taken. The problem is not so simple that the 251th bed is cursed. Numerous more factors are obviously in play, and the statistical significance or even the causality of it all can be questioned. To the extent that even if the numbers would be perfectly on point, many experts of the field will be justified to tear this problem framing apart.

That’s good.

If everyone would agree with on the problem, then it would be solved already. Consecutively, the startup would waste time engaging resources in this market. I even have a rule of thumb for that: no more than 20% of the experts in the field should agree with the startup’s problem framing.

Also, very often it’s not even about experts but the customers themselves don’t even acknowledge the problem they have…
I wrote earlier our quick classification on the four problems that innovators can solve dealing with this paradox, so I won’t spent time going back to this.

In the end, being in a zone where no one really agrees, gives is a non-null chance that the solution will come from an outlier approach.

And what are startups if not outliers?


startup problem innovation copilots

3. Google doesn’t care but others do

Being an outlier means striking a fine balance in term of risk.

First, you don’t want Google, or any other giant like Apple, Amazon, Alibaba or Tencent currently active in the field. If you should aim high and not just work on derivatives ideas, don’t try to solve how we pay on internet, how we exchange photos, how we commute… even if you get tons of public money for it. For the next ten years, this field has been burned to the ground.

On the other hand, disregard that advice if you’re Elon MUSK… Or, if you really think you’ll bring several more orders of magnitude at solving the problem (read: bringing dirt cheap Petabytes of storage on a mobile phone before 2020).

Also, don’t dismiss the problem if other major incumbents that are not GAFA’s caliber are roadshowing innovations that WILL solve it. Typically if Samsung or GE are demoing things at the CES Las Vegas, you’d pretty much guaranteed that customers won’t touch what they are doing with a stick. Consider the CES as a compass pointing randomly at anything between South-East or South-West of the future market. It’s not good enough to infer where to go, but you sure don’t want to go there.

All that being said, you still want significative competition:

It’s exceptionally rare for startups to be killed by competitors —so rare that you can almost discount the possibility. Paul GRAHAM

Incomplete solutions from research, other well ahead of you startups, or even well established incumbents are not a problem. They are a boon, in the sense that they are already probing the market, statistically failing at cracking the problem yet, and thus helping you forecast better where to go, or where not to go. Remember (or learn) that the myth of the first-mover advantage has been debunked long ago by the very team of Standfod professors that formulated it. In real life, pioneers have arrows in the back.

Consecutively, one of the stupidest thing that a startup could say is that they have no competitor. Don’t even try to be smart by saying ‘direct’ competitors just because you haven’t identified anyone with your technology or your pretty form factor. Innovators don’t compete on products but on cracking the market problem.


startup problem innovation copilots


4. It will be painful to crack

The fourth element for a good problem is that trying to solve will be disuasively difficult, or as I prefer to say… painful. Difficult is for thinkers and academics. Painful is for doers.

Painful means that there’s not only a level of intellectual challenge to overcome, but plain physical difficulties: too many early customers to talk to, too much money to invest upfront without clear way of monetizing yet, too many incidents to deal with when prototyping this kind of technology, too cumbersome regulations to go through, etc.

They are ‘good’ problems because they do call for ferocious determination and teams dedicated to beating the odds. People with some level of “fuck you” attitude that we politely call entrepreneurs. These kinds of problems are the ones that eventually sort the Uber and Airbnb out of the fray. And if choosing these doesn’t mean you’re the next Uber, it does mean that you’ll leave many competitors behind from day one.

Of course, one way to be able to overcome painful problems is also not to be aware that they are painful.

This notion of painfullness is probably close enough to what Paul GRAHAM calls schlep and what he says when he considers that “a company is defined by the schleps it will undertake.”


startup problem innovation copilots

5. You have scar tissue from the problem

What we’re saying up to now is that teams capable of framing a problem through its underlying mechanic (point 1.), not afraid of embracing dissensus (point 2.), finding a right timing as to when to address the problem (point 3.), and prepared to grind through unusual difficulties (point 4.) are teams more capable than others to succeed. As such, we can consider that the problem a startup will select does speak directly about the team quality.

If you’d want to change the title of this article to “What is a good startup team”, I’d be OK to go with it (and this is also why my take on innovation strategy is very much connected to Stéphanie’s take on innovation culture and mindset). And this is my final meta key point for selecting a good startup problem: it’s about you.

Now, I’m not saying that in a happy touchy-feely way, where if you select a problem that really matters to you, it’s going to be the best problem in the world. Nah, not really. I’m saying you should have scar tissue from this problem.

This now common notion of scar tissue for business is fantastically important. It means that you have endured intimate negative outcomes stemming from the problem you want to work on. Not at the level of real psychological or physiological trauma, but if you want to solve the problem of commuting in London, it’s not just about having worked in the City while living in Wimbledon for three years. It would be more about having held a key position in the IT department of the Transport for London authority (and getting fired?) THAT could have given you a inherently smarter understanding of where everyone else has been failing for the last decades.

A nice corollary of having actual scar tissue from the problem would be that you’d want to use your own solution as soon as possible. Eating your own dog food and directly gauging if you delivered added value or not, usually bode really well for the way you’ll be driving the project.


In conclusion

As always there is no real process or methodology for innovation. But however your personal experience may differ from mine, as experienced actors in this market we all end up using the same heuristics.

An innovation business is about changing the market. Changing the market means solving unresolved issues. These issues or pain points, may be brand new or long lasting… They define ‘the problem’. And, in real life we have finite resources, time and capabilities. It means that not all problems are as interesting as others.

Don’t be one of those startups.

Author: Philippe

Philippe has been training 200-300 startups a year since 2007, consulted for dozens of multinationals on rupture innovation or corporate incubation, and was directly involved in more than 150 startups building their market fit and scaling up their business. He also teaches business model innovation in key MBA programs whether they are in Paris or Shanghai. And since 2017, Philippe is now living in Amsterdam, one of the best European business hub around.