October 5, 2018

Crossed 10 subscribed users! Not too fast, not to slow. Sharing lessons learned.

Growing slowly, but steadily, stackdraft.io crossed 10 subscribed users last week. Currently at 12 subscribed early adopters, $120 MRR.

Lessons learned:

  1. 💡Know your target

Have to admit I did expect a bit faster growth, but I've shot myself in the foot by initially targeting B2C customers only. Business clients are holding back, because I don't have a "Team plan" just yet (will have it soon). Had a few questions about sharing accounts with co-workers. I did expect more freelancers, but I feel it makes much more sense to target B2B customers than B2C, at least with this specific product, which is a tool.

  1. 🐢It's okay to grow slowly

As mentioned, I was worried Stackdraft is growing too slow. But as days passed, I learned this allowed me to polish the product, fill in the missing bits or fix minor issues, that I very much preferred to have fixed early on the way. That way, much less work was involved in getting things fixed and sorted out for a few users, than it would if I had more users. Pretty obvious, but in the end, I appreciate I didn't suddenly have 100 users.

  1. 🏃User feedback is invaluable

I work hard to get my users opinions. Whenever I interact with them, including my Early Adopter e-mail updates, I ask for their opinion. While only a few come back with suggestions and ideas - when they do, it's pure gold. In those few cases, quite often the feedback is very detailed and focused. I can see they also care about where I take Stackdraft further, which is very uplifting :).

  1. ⚡️Watch out for 3rd party library failures

Only for tax purposes, I convert USD to EUR of all invoices based on European Central Bank rates. They redirected their daily EUR rates site from http to https and a currency conversion package, pointing at http://, started throwing exceptions. All previous month invoices failed to generate, because the conversion was right before invoice generation and I failed to cover that with a test. Dropped the package, now talking to ECB from my custom code and lesson learned - more tests covering 3rd party package failures.

  1. 👀Monitor

This relates to the previous point. I failed to notice for over 3 weeks the invoices didn't generate. I guess it also means at this small $10/mo rate most customers may have not care about those invoices, BUT I should have noticed that earlier. I intend to hook up exception monitoring (which I simply didn't have time to do yet - hey, MVP!) and pay much more closer attention what's actually going on on my servers.

  1. 📈Know your value metric (last, but not least!)

I'm considering changing the value metric from per seat (was just easier to start like that) to per diagram (you get up to X diagrams for $Y), because that's where the value of my product is. If you're a business with 500 people in your cloud department, you neither want a 5 or 10 user plan (too few users) nor 500 user plan (too expensive!). Or you don't want to pay for a manager's seat, who's only there to watch the privately shared diagrams. Plus, it makes sense to allow in ALL of the company users, because the more diagrams they create, the more value Stackdraft provides (and therefore makes sense to charge more).

I hope you find this post helpful! Would love to hear your feedback on the proposed value metric, thanks!


  1. 1

    Hey @mkarnicki thanks for sharing your lessons.

    1.May I know how long it took to acquire your first customer after you launched the product?

    2.Any particular marketing strategy that has worked well for you so far?

    3.Having experience working with freelancers and agencies, I think you are better off targeting agencies. For a number of reasons

    *Agencies have money to spend on products which helps them improve efficiency and workflow.

    *Agencies generally work on large projects with clients and sometime it's necessary they share the architecture diagram with clients and something like this would certainly impress the client.

    *The CTO's of agencies have decision making power whereas freelancers in certain projects maybe just a team member and wouldn't have the power to convince the team to buy or use the product.

    *Freelancers come and go in a project/workplace and have weeks/months where they shift between jobs whereas the CTO or tech lead stays for a longer period at an agency so your lifetime value of the product is likely to increase.

    1. 2

      Thank you the recommendations why targeting agencies makes more sense. This is very informative for me!

      1. I launched on September 10th, got my first 3 customers (one from IH!) on the same day. I think a very affordable $10/mo early adopter price helped. Another customer subscribed 4 days later. Another 1 day after that. Week of silence, and another. And so on. I know I got my last 3 (described in my other comment below) by a referral from my user who got a free account earlier.

      2. Actually, I think I did pretty badly with my launch. I feel I should have prepared an enticing video (I had a basic video demoing the presentation mode with no sound, not very impressive). So I can't recommend on the marketing strategies. I did receive very good feedback for the e-mails I personally send when my users register. But as far as outreach, I think I should finally launch on related subreddit, which I haven't done yet. I'll probably hold up until I have team plans.

      I will add that, I have certainly learned it doesn't matter how small anyone is. I'm a solo dev from Poland. Yet, I was privileged to hold a hangout with Director of Google Cloud ~3 months before my launch. I reached out to him on Twitter and it clicked. It really strengthened the thought anything is possible :).

      1. One again, thank you for your pointers regarding comments. These are valuable insights for me!

      Have a great day @Nakkeeran :).

  2. 1

    Thanks for sharing Michal. Good points that aren't revolutionary but always good to hear them again!

    Can I ask, how did you find your first users?

    1. 2

      Hi Chip, thanks for replying! Yes, I agree they're not anything new, but I guess we all have to learn at least partly on our own mistakes :).

      Sure! I'll start by saying I did give away a bunch of free accounts, both to a few friends and to some higher positioned people. I organically connected (with no "network connections" of any sort) to the Director of Google Cloud and VP of Product at DigitalOcean via... Twitter :)! Funny story is that one used bad postal code (payment failed), one used American Express (which I didn't support at the time [1]). So I have them free accounts. I've given a few others to people I knew could help shape the product.

      Currently paying customers are a slightly different story. My first customer is, believe it or not, from IndieHackers <3 ;). A couple probably found me a few days after the launch on HN and PH. Another few came from organically exchanging comments here and there online. Last 3 came from a referral on an internal Google mailing list by a user, who received a free account (because he too, at the time, only had American Express). It's an interesting pivot - it wasn't his fault he only had AmEx card, so that same user later generated another 4 leads by referring the product (of which 3 users are subscribed; one immediately removed his card - it happens when somebody wants to check the product out; I require card up-front, at least currently).

      Architecting cloud infrastructures is quite a niche market, but that's exactly what I'm going for. One of my recent customers plans to use Stackdraft presentation mode (animated) on her upcoming keynotes :).

      Cheers!

      [1] American Express payments in USD have to be registered for separately when you're using Braintree... it's weird, but I finally did that roughly ~2 weeks ago.

      1. 1

        Sounds like the initial networking is what got the momentum going initially then. Fair okay! Keep it up. 🙂