I came across Juicero again recently and I still cannot understand this one.
Juicero built an internet-connected machine that squeezed packets of chopped fruits and vegetables into juice.
The machine originally cost around $700.
And investors put roughly $134 million into the company.
Google Ventures was an investor. Kleiner Perkins was an investor. These were not people who had never seen a startup before.
Then came the part that made Juicero famous for all the wrong reasons.
Someone discovered that you could simply take the juice packet, squeeze it with your hands and get almost the same result.
You didn’t really need the machine.
Bloomberg actually tested it. In some cases, squeezing the packet by hand worked about as quickly as using the expensive machine.
That part is funny.
But I think the more interesting part of the Juicero story is not the machine.
It is the decision-making behind it.
What did smart investors see?
It is very easy for us to look at Juicero today and say:
Who would fund something like this?
But that’s hindsight.
At the time, I can imagine the presentation looking quite impressive.
Connected device.
Subscription business.
Recurring revenue from juice packs.
Health and wellness.
Hardware + software.
Consumer data.
Premium customers.
Large addressable market.
Put all of that into a good presentation and suddenly a very expensive machine squeezing a packet of vegetables starts looking like a technology company.
And that is what interests me.
Because somewhere in all those discussions around TAM, technology, subscriptions, hardware, data and scale, there was a much simpler question:
Does the customer really need the machine?
Not whether the machine works.
Not whether it has impressive engineering.
Not whether people like fresh juice.
Not whether health and wellness is a large market.
Do I have a problem for which I need this particular solution?
That is a very different question.
A working product doesn’t necessarily solve a problem
This is something we often forget while building products.
You can build something technically impressive.
You can solve a complicated engineering problem.
You can create a beautiful product.
And still not solve an important enough customer problem.
Juicero apparently had considerable engineering behind it. The press could generate enormous pressure. It connected to the internet. The system could even identify its proprietary packs.
Technically, there was quite a lot happening.
But if I can achieve the core outcome by squeezing the packet with my hands, the technology becomes difficult to justify.
That’s the part of this story that stays with me.
And then I look at what is happening with AI today
Every technology cycle seems to produce this behaviour.
During the dot-com boom, adding “.com” to an idea made almost anything sound like the future.
Then we had mobile.
Then blockchain.
Then Web3.
Today it is AI.
I am actually very bullish about AI. I use it extensively myself and have built products using it.
But that probably makes this question even more important.
Put AI in front of an ordinary idea and suddenly it can sound transformational.
AI-powered this.
AI agent for that.
Autonomous something else.
The technology may be genuinely impressive.
But underneath it, I think founders and investors still have to ask the boring question:
Is the problem important enough?
And perhaps another one:
If I remove the technology from the pitch, does the customer still desperately want the outcome?
Juicero shut down in 2017.
The company is gone, but the lesson isn’t.
Every funding boom will probably create its own Juicero.
The difficult part is identifying it while everyone is excited about it, rather than ten years later when the answer looks obvious.
Twenty years from now, I wonder which of today’s heavily funded AI companies we’ll look back at and ask:
Wait… that got funded?