September 3, 2020

I'm shutting down DecisionBee

Thibaut Patel @thibpat

After a good month of hard work, I'm shutting down the app I was working on.

I have written more about the context and what I've learned on my blog: https://thibpat.com/blog/shutting-down-decisionbee

The main lesson is that I've been building something that address a problem that isn't important for my target customers. I was lucky to find this out during during the first month of operation, thanks to customer interviews.

More than ever I strongly believe in these Y Combinator mantras:

  1. Make Something People Want
  2. Write Code, Talk to Users

Off to work on the next idea!

  1. 8

    Hat's off for making the decision - it's definitely not easy! Especially since we all have a tendency to fall victim to the sunk-cost fallacy.

    I ran into something similar with kampsite.co - but luckily decided to instead talk to my existing users to figure out actually what problems they have and am now in the process of a rebuild / pivot to actually address these problems.

    I only bring this up because I wanted to ask if there was any hints from your existing users that maybe would indicate there is a similar / different problem to be solved?

    Or another way to think about it is that if you had to reverse engineer a problem out of your solution, is there potentially another way to solve it which does solve that problem? (Obviously assuming it's a problem worth solving). Taking a snippet from article you listed below:

    First, the lack of visibility on others building or having built the same thing as my team. Second, the tech and architecture debt accumulated due to different teams building things very differently, both approach-wise and quality-wise.

    Is there another way to solve this problem differently from RFC's? Again this might not be a problem worth solving but the point I'm trying to make is that the problem might be legit, just the solution might be wrong.

    1. 2

      Thanks for the feedback! I like your idea of trying to tackle the problem in other ways.

      The issue as I see it is that the RFC process is a great way to solve the problem. Teams that I've talked to are using Google Docs or GitHub/GitLab to store their RFCs and collaborate. And after talking to them, they are very happy with the current setup!

      I guess a good business to tackle this problem would be a consulting business where you spend time with teams to introduce them to the RFC process.

      1. 1

        I was reading this and wondered how can you improve over either a Google Doc or GitHub Pull Request

        1. 2

          I was thinking about several ways:

          • remove inline commenting to make it easier to read the document and keep the discussion in one place: at the end of the document
          • provide easy tools to make sequence diagrams and gantt charts in the document
          • list the document in review stage / in draft

          However the conclusion I came up with is that these features are just a small improvement over the existing tools. So it will be really hard to convince anyone to change tool and even harder to make them pay for it.

  2. 2

    Queries from my end:

    1. Having gone through your website I couldn't understand what it exactly does ?

    2. How much is the overall finance needed to maintain the site ?
      Is this one of the reasons ?

    3. What was the visitors response when they came to your site ?

    1. 1

      Good questions:

      1. This is a tool to manage your Request For Comments process. If you're not familiar with the RFC process, you can read more about what it is and why it's important to use there: https://blog.pragmaticengineer.com/scaling-engineering-teams-via-writing-things-down-rfcs/.
      2. There isn't a lot of costs to maintain the site, but given that no one is using it, it's costing time.
      3. They came once and never came back 😅

      As I've written in my blog post, the problem is that I've made a tool that isn't solving a important problem for my customers, so they're not interested in using it.

      1. 3

        Worked in teams for 12 years. Never heard of RFC being used in everyday language. Whatever you work on next, make sure you use language or terms that target users are already familiar with on your landing page. if it requires a google search to understand a landing page, people will not do the search.

        good luck though.

        1. 1

          Thanks for the feedback, it's definitely part of the problem!
          I thought I could tackle this issue by doing content marketing, sharing the benefits of this approach. But even folks already using the process don't need a new tool.

  3. 1

    Hi Thibault, great to see you here. I think we met at an efounder event in Paris.
    I am also obsessed by the "Don't build hoping for interest" - would you have time to talk about this soon?

  4. 1

    I am happy for you to learn something important, in deep, but I am sorry for you to close a project. And thank you to let us learn from your experience.

    1. 2

      Thanks @pydatageek :) I'll make sure to share about my new project soon!

      1. 1

        I am looking forward to it. I like people pushing the door even after failures (!) On the other hand, there is a project of me and I have shared milestones of it on indiehackers, here. I am getting comments from people about it. I have received from a potential user that there are tons of similar websites and I will not be able to be publicly seen on Google because of them. Also he said that I need more solutions/products that they don't offer. I agree with him. What do you think?

  5. 1

    Can you elaborate a bit more on what an RFC is?

    1. 1

      A quick blurb that I shared in the app:

      An RFC is a Request for Comments. It is a decision making process for engineering teams. It is based on documents that suggest changes to a group. The goal is to gather feedback on future changes to improve them, and to keep a record of these decisions and discussions.

      If you want to learn more, feel free to read how Uber uses Request For Comments to scale their engineering team: https://blog.pragmaticengineer.com/scaling-engineering-teams-via-writing-things-down-rfcs/

  6. 1

    Big kudos to on having the courage to pivot and be willing to change in the face of evidence.

    I faced a similar thing with FlowCraft and wrote about it too.

    A couple months later one of my friends said she loved the authenticity.

    All the best with whats next

    1. 1

      Thanks Jay! Feel free to share the link to your article so that I can give it a read!

        1. 2

          Thanks! I find it challenging to keep working on a newsletter, congrats on pulling the plug to focus on new projects!

          1. 2

            Thanks, we get a few swings at this

  7. 1

    Well done for closing it down after only a month. The sunk cost fallacy is a big one and "pivoting" can simply be an expression of that. My own project is constantly "on the brink" of success and I have done numerous pivots. The question I ask myself is whether by pivoting I am simply exploring a wider part of a landscape where nothing is commercially viable. Killing the whole project and doing something new might have a better chance of success.

    It does not help that I was writing software to manage events just as Coronavirus hit! Good luck with the next idea.

    1. 1

      Sorry to hear your business has been hit by the lockdown. I hope things get better soon!

      Agreed that was a decision that I struggled a bit with. The trigger was customer feedback so that felt reassuring to shutdown based on data.

  8. 1

    Must be a bummer, hopefully next one is a homerun!

    1. 2

      Thanks I'm hoping the same!

  9. 1

    It was surely tough decision to be made.

    Good luck with next projects Thibaut!

    1. 1

      Yes it was tough, but I'm happy I took it today instead of in 6 months! Thanks!

Today's Top Milestones
  • Am I doing user research wrong?
    TLDR; I got 10 waitlist signups yesterday, personally emailed 8 of them, and only got 1 response so far. Would appreciate some thoughts about my proce
  • Built MVP for Automated User Testing
    Fed up spending an excessive amount of time test user flows in the browser, I decided to build a tool to automate this process and save me hours of ti
  • Launched
    Writing code at until 2 in the morning for a week and ruthlessly following my target finally resulted in today’s launch. I learned a lot about integra
  • 63 copies sold in 9 Days
    Hello Everyone, Excited to share the progress of our brand new ghost theme InoLetter. In 9 Days we sold 63 Copies. The Coupon (FC50OFF) of 40% discoun
  • 70 domains and a new domain appraisal service
    Today I'm happy to announce that there are already 70 domains in the Upfordomain database ready for rent/sale despite no marketing or promo. I have al
  • +130 integration tests and first blog post
    NetcoreSaas v1.3.0 update highlights: -130 Integration and 7 Unit Tests -Clean Architecture -.NET 5 -TailwindCSS 2.0 -Other: --Impersonate users --Upd
  • First paying user after 100 emails
    Inksprout is the second product I'm working on and I wanted to validate the idea from the very beginning. I ran 2 experiments: 1. Paywall 2. Freemium
  • Posted to Product Hunt
    Posted My Site Archive to Product Hunt in hopes of getting some exposure. The fact that you're limited to 260 characters for the description forced me
  • First customer for the new services!
    During this week I've been working on the re-pricing and offer of Tailry, focusing more on one-time tasks and simpler subscription plans. I removed th