Andrew Chen, Quibb, Uber

Read what else Andrew Chen is reading for work

Quibb lets you share what you're reading for work. Use Quibb to share news about your industry, discuss what matters, and see what colleagues are reading.

Our mission is to connect professionals over business news and informed commentary — targeting every industry, profession, and geography.

Likes Comments
Startup Edition, Product Hunt

Great post, Andrew.

Thankfully, we can now validate new ideas and build product faster than ever before. I'd like to see some data trends on TTPMF. Is it decreasing? Increasing? Staying the same? What are the characteristics of or tactics used by products that have a short TTPMF?

Reply · Tweet ·
Quibb, Uber

I imagine that TTPMF has two main components, maybe more:
1) time to build the product (and each iteration)
2) making good decisions

Clearly #1 has decreased over time, but #2 probably hasn't :)
A lot of startup advice has focused on iterating quickly, etc., which is about #1. And the great value that customer development has really brought to the table is to increase our ability to deal with #2, but people probably know how to do #1 better than #2.

Reply · Tweet ·
Startup Edition, Product Hunt

I agree. Re: #2, that's why I'm such a huge proponent of personal growth and reading other's shared knowledge.

It's great to see people like Nir study human behavior and articulate his learnings into actionable processes to improve #2.

Quibb, Uber

FWIW, the 3 articles/videos that have influenced me the most about picking products are:

One of my favorite things about Silicon Valley- at one point I drove out to Steve Blank's place on the California coast and we had a 2 hour long discussion about market types and what does it really mean to be in an existing versus new market. Blew my mind and dramatically changed the way that I think about picking a new product. Such a great experience.

Scout Ventures, Dozen Digital

Solid post. It's hard to find that balance. And it somewhat depends on the market, the challenge in finding people who currently don't have ANY solution to their problem, or is the challenge in switching people from one solution to yours. But for both, its important not break too many of people's existing mental models (UI/UX, flows, features). And then, either establish just enough of a value proposition to get them to stay/pay (for markets with few competitors), or establish just enough of a differentiated value proposition to get them to stay/pay with your competitive service.

I agree that differentiation is secondary to PM-fit. Especially for the vast majority of us that can't just raise a bunch of money to spend a while educating a new market on something totally new.

This is a loose analogy, but somewhat relevant: when I was in high school, I played on a very competitive basketball team, with many players going to division one schools to play college basketball every year. I had a good jumpshot, and was told by the coaching staff to just shoot the ball (instead driving to the basket or playing much inside, or anything else). That was their mental model (Jim = shooter). Until my senior year, I didn't want to "just" be the shooter-guy, I wanted to do other stuff, and I didn't get playing time. I finally realized my senior year I had to not be stubborn, do what they wanted, and then would get the playing time necessary to try out new stuff and expand my game a bit. Lesson learned. Sometimes you can't get to your grand vision until you've done a good job keeping things simple. Then once your in the game (have customers, traction, PM Fit), you can start experimenting and differentiating a bit :)

Startup Edition, Product Hunt

Agreed and it definitely depends on the market. While copying 80% of an existing solution and differentiating in the remaining 20% can be a sound strategy, it may be unrealistic based on your team, run rate, and maturity of the market.

For example, if you want to build a replacement to existing analytical solutions like Google Analytics, Flurry, Omniture, etc. you'll have a long road ahead to reach parity to meet their v1 needs. A wiser approach, imo, would be to hyper focus on one problem within that domain, similar to what is doing.

Reply · Tweet ·

I completely agree w/ this post, Andrew Chen! Ryan Hoover, agree that trying to build out lots of functionality might be difficult to displace an incumbent. But, in this day and age of APIs, do you think it's possible to displace incumbents using their own functionality or other 3rd party APIs? E.g. power v1 with your competitor's functionality while, in parallel, building out your own? (heh, seems to be doing just that... -- can you do that with other markets?)

Startup Edition, Product Hunt

Spot on, Elizabeth, and yes I believe this strategy could be applied to several different markets, including the consumer space. That's why Facebook and Twitter have been increasingly protective of their core offering, banning apps that could potentially siphon their user base.

Reply · Tweet ·

Solid Advice, Andrew. Having worked on products in the recent past that didn't get to PMF and died, I can tell TTPMF and TTR (time to revenue) are the *only* thing that matters. The 80-20 idea is also solid. Not to do a sales pitch here, but we recently launched an app for craigslist ( with that exact motivation - 'lets reinvent craigslist with great UX', and make money along the way. Doing well so far.

So I do think the strategy should be - early idea validation --> validate if people will PAY --> build MVP (with minimal effort) --> test/get people to use/PAY --> and then build the 'nice' product and iterate.

Building product scares me. Finding PMF *fast* is the most important thing.

Tech Lead/Manager at Google

Sounds like you like one end of the trade-off: lower risk with the accompanying lower rewards.

Quibb, Uber

Are they lower rewards though? Facebook wasn't the first social network, Google wasn't the first search engine. Even Microsoft wasn't the first OS developer or Apple the first smartphone maker...

Intuitively the higher risk higher reward idea makes sense, but am not sure it actually applies in practice to tech companies.

Reply · Tweet ·
Founder at GoAnimate

Also interesting to see how incumbents often fail to catch up. FB tried to kill off snapchat but that didn't work. Hipmunk seems to be doing well too.

Reply · Tweet ·
Partner at Heracles

Nice post. I think this underlines the difficulty in doing a start-up while employed full-time with a "day job". Reaching product-market fit quickly is critical, and working with a restricted time budget is a serious burden and handicap. Once the MVP is out, reaching product-market fit is a race against the clock.

Product Lead at Flurry

Nice post. I haven't done a startup per se so take my $0.02 below with a grain of salt but I do have relevant experience from 2 failed businesses inside of Google (print ads platform, radio ads platform), two mildly successful ones (internet radio ads, tv ads platform - although Google killed the latter because it wasn't big enough) and scaling Zynga mobile poker.

I think the first thing that strikes me is that even an 80% clone (20% innovation) can be damned difficult to sell / position / get off the ground. For example, we knew there was $20B annual print ads industry in the US (this was in 2006). And we had some successes (we got most major pubs like NY Times signed up), got decent advertiser interest but we couldn't get the pubs to abandon the old ad sales model for our platform. So even in the "mostly copy" TTPMF scenario, it's still really hard. We had all the funding we needed, 20+ really bright people, a few industry vets we brought on and we still couldn't crack it (also, on a semi-related note never convince yourself you're going to go take a big chunk of a big, but declining industry. Not fun.)

Chief Product Officer at Fullbottle

Andrew - Great analysis. The numbers are interesting, and disturbing.

It seems you're saying the only realistic path to success is instant traction - but isn't that a unicorn? (Or is that actually your point?) I mean, even if a startup does what you suggest - "80% copying, 20% substantial reinvention" - it seems highly unlikely to have instant product-market fit. It seems, to be successful, a startup must have made substantial progress before initial investment, and before this 3 month mark.

If the economics you describe are true, it suggests that angel investment should only be made after PMF has been demonstrated. I mean - PMF must be right around the corner, if you are to be successful, or it is not and you will fail. If the latter, one shouldn't invest; if the former, the entrepreneur would be much better off demonstrating PMF first, both for his sake and for the investor's. In other words, the economics suggest that angel investment, is no longer for helping find a scalable model. If we accept Steve Blank's definition of a startup (a search for a scalable business model), it leads to the slightly strange conclusion that angel investors no longer (or should not) invest in startups. If rational, they are really not qualitatively different from venture capital. Angel funds are for scaling to 1 million; Series A for beyond that.

There's something depressing in all of this for dreamy entrepreneurs. And it both leads from and concludes with the idea that there's nothing new to discover. Arguably, 4sq's success, such as it is, is predicated on learning from failure after failure in the location space. If this is true, it means that 4sq is the beneficiary of cycles of foolish investment in too ambitious ideas, which were trying to invent 80% instead of 20% (because they had to at that time). The gloomy conclusion is that in a rational world, where all investment is for scaling, PMF can only be found by irrational risk takers.

Hmmm… anyway those are just some initial thoughts. I have to give it a nights sleep to see whether I stand by any of that.

I wonder how many entrepreneurs without a failure in their resume go through the exercise of a) qualifying an market, b) planning strategy backwards wrt time-to-revenue or TTPMF.

"Why now?" Asks Sequoia. "B/c now it's the best time to TTPMF" should be the anwer :-)

Reply · Tweet ·
Director of Marketing at Playerize

Failure certainly makes it crystal clear that spending as little time and energy in between 'now' and 'TTPMF' is the highest priority, because goddamn is that some painful time:)

Reply · Tweet ·
Founder at Fundamentum

My counterpoint to this is that you will end up with mostly incremental companies this way. What was the 80% existing product that amazon copied? Or to look at something more recent - twitter?

Can you make something big by only being 20% different?

Reply · Tweet ·
Quibb, Uber

I think the goal is- Make it the right 20% :)
Ideally the differentiation is baked deeply into the core of the product, not out on the edges. So even if you see that all social networking products are public and anonymous, then you go with something private with real names. But you still have profiles, friend connections, and the other things that people would recognize as a social network product.
With Twitter, you might argue that a lot of features were already well understood within a blogging product: the stream of posts, being able to subscribe to others, customizable profiles, etc. But the 20% that could be different was the 140 characters.

Where I agree with you is that if the product is basically completely the same, but the 20% is out on secondary/tertiary features that aren't used much, that's probably a recipe for a commodity product.

Furthermore, I'll agree with you that there's many companies that don't follow this recipe to reach success- Groupon essentially created a new set of mechanics that had never existed before, to create their business. Foursquare too, and Amazon, and eBay. I'm just saying that this is a way that I understand and feel is more repeatable :)

Chief Product Officer at Fullbottle

Andrew, to your point. Tony Conrad (True Ventures) was on stage at Launch on Monday, and seemed to confirm your thoughts. He stated (somewhat dyspeptically) "Entrepreneurs come and show me product market fit. I don't care about product market fit. I want to see serious traction."

Reply · Tweet ·
Editor-in-Chief at Quibb

I was there too - Tony's thoughts were really interesting to hear.
Summarizing from my notes, he said that investors are starting to realize that it's not enough to have traction, or to be at PMFit. Today, investors are looking for more and more traction - but also an understanding of what it takes to get traction. They're looking at the product, to see what's built into the product itself that will lead to further traction and growth. They're looking for baked-in viral loops, and even further to features/product, they want the entrepreneur to be able to articulate how that loop and how growth happens in the product.
It was an early comment (maybe within the first 20min of the conference) and came up again a few times later in the day.

Andrew, you've shared another great framework for evaluating progress & likelihood of a positive outcome. Aneil and Chris's comments are equally thought-provoking (and sobering).

Considering this thread makes me wonder if the answer to shortening TTPMF doesn't lie in the up-front customer development phase.

While the 80/20 discussion is interesting to consider, how many success stories are anchored in a reverse engineering exercise (Samwer brothers aside)? It may be naive, but I feel that sticking to the fundamentals of CustDev gives you the best chance of nailing PMF quickly.

As to the comments at LAUNCH, VCs seem to like nothing better than to ask for the moon. If you've already achieved massive traction, your risks have dropped proportionately, which should then reflect in the investors' upside potential. Wonder how the LPs would feel about that.

Emeka Ajene, Founder & Managing Partner at Africreate Emeka Ajene
Founder & Managing Partner at Africreate
Sunil Subhedar, Growth Marketing Manager at Uber Sunil Subhedar
Growth Marketing Manager at Uber
Karthik Hariharan, Engineering Manager at Uber Karthik Hariharan
Engineering Manager at Uber
Yuanxuan Wang, Software Engineer at Uber Yuanxuan Wang
Software Engineer at Uber
Jay Ottovegio, Strategy & Analytics at Uber Jay Ottovegio
Strategy & Analytics at Uber
Hao Truong, Engineering Manager at Uber Hao Truong
Engineering Manager at Uber
Mark Reeder, Senior Software Engineer at Uber Mark Reeder
Senior Software Engineer at Uber
Florent Crivello, Senior Software Engineer at Uber Florent Crivello
Senior Software Engineer at Uber
Paul Clayton Smith, Lead Experience Director at Uber Paul Clayton Smith
Lead Experience Director at Uber
Nikolai Cornell, Senior Brand Designer at Uber Nikolai Cornell
Senior Brand Designer at Uber