40 thoughts on “How to Find Product Market Fit – CS183F”

  • Andrea Trocino says:

    Detailed Summary of this Great Talk

    (1:40) – Imagine a future world and set a vision is amazing for research but it doesn’t really work for business. Funding a company is about what customers want. A lot of companies get marketed after facto.

    (2:30) – "The most common mistake startups make is to solve problems no one has". You have to solve a speficic problem that customer have today, now.

    (3:40) – 80% of all founders failed to find product market fit. That's 4 out of 5.

    (6:00) – It's actually harder to find the first problem than finding a route out. Also turns out that once you solved a problem for a customer he/she will come back with more problems. Once you find the first thing to fix it is almost trivial to find the second one and so on.

    (7:30) – It's very important to save as much cash as possible before finding product market fit to extend your runway. Step 1 of this approach you build, launch and iterate on several ideas until something magical happens…and that is product market fit. After that everything gets easier, customer shows up and you can feel momentum.

    (9:26) – 3 Chapters

    1) Deconstruct category leaders
    – (10:30) – You can build a category market leader when you build a platform. It is not just about the product you are selling, it is when your product can be used by other businesses to build their business on top of yours like Salesforce, Amazon, Facebook etc. Salesforce has what is called AppExchange that reveals all the data contained in Salesforce for other applications to be built on top.
    – (11:30) – Peter Thiel says "Not a platform until you have made $100M in revenue"

    2) What bad fit feels like
    – (15:45) – Today Segment is a customer data platform but we started as an educational tool for teachers to collect student feedback during classes. After convincing some professors to use the tool we realised that we were ignoring any test for product market fit. We realised that none of students were using the product because when students were on their laptops they were distracted in doing something else. At that point we raised $600k from Y combinator with a great vision but that wasn't what users wanted. The students didn't get much value. Forcing customers to use your product is not a good thing.
    – (19:00) – After that we called the investors asking if they wanted their money back or they wanted to reinvest in the team, and they said they believe in the team and said "go and find another idea"
    (21:00) – We started building an analytics platform, we spoke to potential customers and they were vaguely interested, they were talking to us and ask for product updates but they were probably not interested in buying it. We tried for one year and half without success, spent $500k without doing anything.

    3) What good fit feels like
    – (26:10) – We open sourced an analytics JS library that was pushing data to multiple platforms including but it didn't have much traction until my co-founder come up with the idea of building a product around that and I thought that was a stupid idea and I should have tried to kill it asap. We decided to build a landing page, put it on hackernews with a register your interest form to see if anyone would have been interested or not. We reached the top of hackernews, 326 upvotes, 3000 stars on GitHub and people asking to get bumped up in the beta list! Comparing this to everything we did before everything changed!
    – (28:00) – That's the key thing when it flips from you pushing to the customers to them asking for it and for more than you can deliver.
    – (28:40) – With the previous ideas we were thinking how classrooms or analytics tools should operate while this was the opposite, from a little library that solved a real problem without any vision associated. It was a tiny problem but it was real. The key is the market doesn't care at all of what your vision is, the market will get what it wants.
    – (30:00) – If you are questioning if you have product market fit or not, you don't. The first 2 times I thought we had it and it wasn't true, the third time when I thought that Analytics.JS was too small to be significant it had market fit because it was solving a real problem
    – (31:20) – Question about monetisation of the open sourced library. Answer from Peter: "The library by itself doesn't solve the problem, what we found was a disconnection between Marketing and Engineering teams so with our library we solved the engineering problem of pushing data into multiple tools used by the marketing teams. We had an opportunity to build on that.
    – (33:20) – Two examples of finding product market fit after the Analytics.JS, so short intro of what does is basically to collect data from web / mobile apps and create segments and push that to all the different tools you need. One of the places where we can send data is Amazon S3 and we started to noticed that lots of our customers were using that and we went to visit them to understand what they did with it. Everyone said that the data engineering data team was extracting data and pushing it into data warehouses. All of them said the same things and when we built a direct integration our revenue started increasing (sales graph @35:20)
    – (35:40) – One more story before opening for Q&A. 5 months ago we had 5 ideas for products we thought would be exciting for our customers. We visited a few of them and we went through those ideas and for 4 ideas they said "that sounds interesting" and in reality that meant they didn't care, when they saw the 5th idea and they said "wait, you can do that? We have to set a follow-up with this team, that team etc."

    (38:00) – The problem is not about finding the idea, it is more about killing the wrong ones fast and you will probably find something.

    (39:20) – Question about pricing since it was a small problem solved. Answer: We under-priced our product because we thought it was solving a small problem but when we realised what was solving for our users we started adapting our pricing. Now they have customers paying over hundred thousand dollars per year. The size of the business problem doesn't have anything to do with the lines of code written. We tried experimenting with pricing after we had 1000-2000 companies using us and we hired a sales advisors that helped us in finding the right price.

    (41:00) – Question about how do you know if you are talking to right users or not. Answer: It is very hard and it is important to recognise when we are shifting the product or the audience.

    (42:40) – Question about to teach people when to kill an idea or when to go after it. Answer: It gets easier after finding the first market fit and it gets a bit easier to train a product manager to go after a fixed audience and improve the product.

    (44:00) – Question about whether there is a script to use when talking to users. Answer: Biggest mistake is trying to pitch and selling your idea, instead of trying to understand there their problems are, especially when you already have a product.

    (45:30) – Question about how you find the first customers. Answer: At the beginning is via spamming, searching for intro via social / Linkedin / investors etc. Cold mail will get very low conversions, single digit probably.

    (47:00) – Question about vision for Segment. Answer: From existing connectors to a platform where all the integrated tools can build on top.

  • Unicorn Launching says:

    39:38 "The size of the business problem has almost nothing to do with the amount of code written [to solve it]." – Reinhardt

  • Dmitriy Fabrikant says:

    I put together a thread with highlights from the video:

  • Peter Reinhardt says:

    Many folks have been reaching out after watching the talk. For those of you searching for product-market fit, happy to help:

  • Sorcerer Stone says:

    It is wrong to state using Wayne Gretzky strategy for predicting market/development direction is wrong @1:34. All hockey players use this strategy successfully – past and present. Presenter idea is correct but used the wrong example.

  • Vikrama Dhiman says:

    This was by far the best lecture of the series. Peter Reinhardt is prepared but not rehearsed and there is an honesty to his talk without doing the shenanigans of modern presentation – no theater, no inserted jokes and made up stories. Thank you.

  • Christopher Horton says:

    Why did he not iterate on the first product with the students? I think there’s something in that idea and was kind of disappointed that he gave up before evolving and trying to find “fit” with the product

  • Many comments are missing the key point, which in my mind is being honest with yourself and focusing on working through the problem tangent as compared to the solution tangent …

  • This video is more about how to recognize product market fit than how to find it. Any more particular advice on the topic?

  • Arthur Gopak says:

    Thanks a lot for such an insightful lecture, Peter. So far, this is the best lecture out of the whole class.

  • Very interesting talk. But he has thrown the baby out with the bathwater on the value of vision.

    Like, with Sam's question: how do you know which ideas to test.

    He answers it like: not an issue, just kill bad ideas fast enough. But that is wrong. if you test ideas at totally random, you can suggest absolutely anything. Pick your nose, paint the house purple, eat rocks for a week. Of course ideas has to be constrained, and ideally constrained towards some goal. And that is your vision.

    So, it actually is impossible to know which ideas to test, unless you have some way of constraining the ideas you test.

  • Mind blowing talk! Forces you to rethink. Love the way he emphasises on product market fit is something you 'feel'

  • Saurabh Hooda says:

    Summary of the lecture:
    1. Keep your startup alive for a long as possible by keeping your expenses bare minimum.
    2. Talk to users and keep trying. (Trying new startups/features/tweaks/etc.)
    3. "Hope" that you'll find product/market fit. How do you know that you've achieved P/M fit? You'll get a flood of need-met users & market will extract features out of your startup.

  • My favorite lecture so far, including the ones posted back in 2014. Funny presentation that was easy to understand and to get something out of, especially thanks to the variety of relatable, real life examples he used. Love how we also mentioned the Lean Startup book, proving how useful and gamechanging books can be.

  • Evan Kozliner says:

    Such a cool video. His cofounders proposition to double down on the analytics library was very perceptive!

  • FUNcationNation says:

    last call to signup for T-shirts We have over 70 so far

  • This is definitely one of my favorite videos of the class, especially since a lot of the recent ones have been lacking.

    I really appreciated that in this video, Peter used his past experience to create actionable items for us as students (something that the most recent videos haven't been doing).

    Loved the speaker – very smart while still being humble. His lecture really added a lot to Startup School.

Leave a Reply

Your email address will not be published. Required fields are marked *