Sales Safari and JTBD only work after you have found signs of demand (but you don’t know why they are buying). How you find that demand in the first place is idiosyncratic. (That is … you get lucky 
). The Heart of Innovation, on the other hand, is the one framework in the Demand Series that helps with pre-demand search. But it helps in a very odd way — it prevents you from lying to yourself.
2 Likes
Thanks for the extra clarity on this! I’d previously not actually read the Heart of Innovation article yet so had mistakenly lumped that in with Sales Safari and JTBD. I understand the nuance that you’ve pointed out now and can see how tools like “the not not”, situation diagrams, and DPI can help pre-demand while having conversations and building a picture.
I can see how DPI and customer interviews in particular can apply to a “zero to one” business where you’re trying to find demand that is extremely lucrative (venture scale/$1bn) and build a product that is defensible. It’s often too difficult and costly to “just build” under so much uncertainty and with such a high bar to meet.
But how about situations with smaller outcomes (e.g. $10m)? Is this still the optimal approach in your eyes?
Two “lenses” I’ve personally fixated on in the past:
- The Samwer brothers (Rocket Internet) model - take an existing business model that’s taking off elsewhere and localise it to your own country. One of the products the Sanwer brothers had success with for example was Alando (eBay for Germany)
- The “copy and pivot” model (h/t Richard Harpin of Homeserve) - take an existing product or service and make it better (ideally faster, cheaper, or more convenient)
My theory on why these approaches COULD be better for a bootstrapper with modest ambitions is that they get you closer to PMF without the exhaustive and uncertain groundwork of customer interviews for months on end. Maybe you’re less likely to get a huge outcome, but more likely to start generating some revenue quickly which can sustain you in looking for more lucrative demand and/or be compounded?
Seems interesting on the surface, but these approaches have the issue of being derived in a top-down way - copying and pivoting is not always easy without insider information, and any “improvements” you make are solutions in search of a problem until you validate that there’s demand for that. In light of learning about effectuation and “action produces information” I’d be inclined to clone a business to begin with and then effectuate forward into a better solution; a better approach than trying to theorise on how to build a better business on the outside with no skin in the game (I’ve been down that route many times and I often talk myself out of it).
Here are some interesting examples:
- Cal AI - took MyFitnessPal and made it easier by adding AI processing of images to automatically calculate calories. Doing $2m per month in revenue (debatable how much of that is profit, but the revenue is very impressive).
- Powersheds - £11.5m in revenue in 2023 (£1m profit) selling high-quality garden sheds online.
- Aybl gymwear - £32.5m in revenue in 2024 (£4.5m profit) selling gym clothes. This one is interesting in particular because they weren’t particularly early to this market, or innovative. But their growth is explosive. Their advantage was likely routed in experience and knowledge (their exec team are some of the best in the biz).
There are countless examples of these kinds of businesses, but I chose these ones in particular because they feel so simple. They didn’t build an innovative VC scale solution on the back of months of customer interviews, they just took something that was already working and put their own twist on it.
Taking the other side of the argument, I guess you could say:
- Cal AI - perhaps a case of survivorship bias and getting lucky. Question marks over profit and whether there is actual enterprise value there
- Powersheds - I believe founders worked in the industry prior to founding. They likely discovered some secrets in a less formal format than customer interviews?
- Aybl - perhaps fashion is special in that the value is intangible (brand) and extremely talented operators was enough of an advantage in this specific scenario to create a desirable product
Maybe the most compelling criticism against these businesses, in the context of the CommonCog concepts, is that these businesses have no compelling moat and are susceptible to competitive arbitrage, especially from competitors that already exist - at least Cal AI and Powersheds (Aybl less so due to brand affinity). Whereas maybe a bottom up approach gives you a head start and more time/opportunity to take first mover advantage and build moats?
Apologies, I rambled a lot here, but hopefully you can see what I’m getting at with all of this? These are the questions I’ve been asking myself as I’ve been going through the materials over the past few weeks.
2 Likes
Hmm, I think the high order bit here might be:
- You want to get to a point in your life where you can make ‘unlimited’ affordable loss bets.
- The way to get there is to start with smaller, easier businesses.
On the businesses you cited, it’s true that these businesses have no compelling moat, but that might be ok if they’re ‘stairstep businesses’ (Rob Walling’s term, not mine). That is — you use them to practice and gain business skills. Notice how Rob’s framing is an instantiation of effectuation! The stairstep approach is what he did before he built and then finally sold Drip (and is now kinda retired). And in fact, in the most recent episode of Startups for the Rest of Us he mentions how he had done a few ‘stairstep businesses’ by the point he had his two kids, so for the first few years of their lives he cut back and put everything on maintenance mode. Then, when they were older, he ramped up again.
I think you could say that by the time he had kids, he was in a position of ‘infinite’ affordable loss bets.
2 Likes
Great article! I read it in 2020/21 and it, along with other similar content at the time (e.g. Nathan Barry’s ladders of wealth creation and Daniel Vassallo’s small bets course which is also another framing of effectuation), is what convinced me to go into freelancing as my first small step. It’s now my full time gig and opens up the opportunity for some affordable loss bets with the flexibility I now have.
Given that those businesses I mentioned would all be hugely successful outcomes for me personally, perhaps the top down approach is king for earlier steps?
I did already execute on The Samwer brothers model twice last year. Both businesses were failures for various reasons. I’d say the business fundamentals rather than the approach for the most part - but executing this kind of project did open my eyes up to the fact that it’s not as straightforward as it seems to clone something in another location. You need to take into account cultural differences (I think the Samwer brothers were particularly good at this) and there may also be hidden reasons why the business worked in the first scenario that can’t necessarily be replicated in your clone (e.g. in my case, one of the businesses that I cloned was built on founder-led marketing with an already big following and I underestimated the influence of that on the outcome).
3 Likes
I meant to respond earlier, but wanted to get through the veritable reading list and then Thanksgiving got in the way. I *love* this answer and the articles linked. I think the truth is that we all want the formula but there just isn’t one. I interpreted your article as a call to try to start small markets companies, but am reading it now as a call not to ignore small markets companies if that’s what the founder + market implies, because there is value there. It’s actually pretty freeing to instead think about my own circumstances/skillsets for the type of company to try to build (also: it’d be awesome to have a “nest egg” to fund a company that’s not encumbered by external investors and their narrow expectations from day 1!).
Just read an article about David Rubenstein, a founder of Carlyle, and the article made it pretty clear that he was uniquely positioned to start that company (and thus why it would be somewhat foolhardy for a different person to try to “replicate” the same strategy and expect it to play out the same way, even not withstanding some ethical dubiousness of the strategy).
2 Likes
Hi David — one more thought here: I think cloning an existing business probably[1] addresses the question of customer demand (which I believe is a mysterious and basically non-deterministic thing[2] to try to guess at), so having that out-of-the-way at the start definitely is a huge advantage compared to a business that hasn’t solved that mystery.
But I also think executing is really fucking hard too. And with a business that is a clone of another one, I assume execution is waay more important than with a business that happens to find a gusher of untapped demand.
Having said that, I do think that execution can be more deterministic — with enough time and practice, you should get better, and you should have a sense of if you’re getting better. I’m probably 1/3 through my career, and I feel a lot of conviction that I’m at least marginally better at execution now than at the start of my career, while I feel zero conviction that I’m any better at predicting new types of demand.
[1] I assume it’s easier, but when you move into a new market like a new geography, I think the guarantee of demand is more tenuous. E.g., the Samwer brothers have gotten it wrong quite often too, because customers are just fundamentally mysterious.
[2] I’m probably too extreme in this belief, but I think demand for new offerings is truly non-deterministic — like I don’t think Laplace’s metaphorical demon (i.e., understand velocity and position of all particles in the universe) could not successfully predict demand. Humans are just too weird.
5 Likes
You could be right about execution being harder if demand is already pre-established.
Alex Hormozi has an interesting heuristic about roads you can go down in business which is kind of related:
- Solve a problem better than others (execution risk)
- Solve a problem no one knew existed (product-market risk)
- Solve a problem everyone knows about but is hard to solve (technical risk)
We know that creating a business in an existing market solves for product-market risk and technical risk already, but the nuance which you rightly point out is that the absence of 2 and 3 could increase execution risk.
I’d posit that the degree to which execution risk could be amplified could also be dependent on other factors such as age of the market, size of the market, sexiness of the market (sexy ideas = more competition) and tendency towards winner takes most/all.
3 Likes