Day 694: Hacker News Launch Retrospective

marketing

It’s been a week and a half since the Browserflow launch on HN, so let’s do a retroooooooo!

I posted on Hacker News with pretty high hopes because I felt confident in the post title: “Show HN: I made a Chrome extension that can automate any website”

It had all the ingredients of a successful post:

  • “I made” - casual and personable, not corporate
  • “Chrome extension” - Hacker News freaking loves browser extensions
  • “automate any website” - Hacker News also loves automation / productivity

So I posted and waited.

When I post things to Hacker News, I don’t tell anyone. I don’t ask friends to upvote because I don’t want to worry about whether I’m being sneaky enough to avoid the upvote ring detection algorithm. If I submit a post and it doesn’t do well, I can always post again some other time.

An hour later, Browserflow had fallen off the first page of /new and I figured that this chance was over. Next time! I went to take a nap because I hadn’t slept well the previous night.

When I woke up a few hours later, Browserflow was on the front page of Hacker News with dozens of people on the site. Oh snap!

It peaked at the number 5 spot, and throughout the day, there were always around 200 people on the site at any given time. Pretty amazing.

Here’s the graph from two days after the launch:

Over 80,000 pageviews and 42,000 unique visitors!

Here are the things that went well with the launch:

  • I came up with a title that made people curious enough to check out

  • The website held up to the torrent of traffic

    • The marketing website is a Next.js site hosted on Vercel, and it performed beautifully. My bandwidth usage started getting a bit close to the monthly limit of 1TB which was concerning because it starts getting very expensive once you cross that threshold, but thankfully it stayed under. (The reason why there’s so much bandwidth being used is because I hold the demo video directly on Vercel for performance reasons.)

  • The product was SOLID

    • The comments were uncharacteristically quite positive and many people commented on being surprised by how good the product and the onboarding experience was. They’re seeing the result of countless iterations over 1.5 years, but they don’t know that. The fact that the product worked without issues for everyone (as far as I could tell) and that it was easy to get started made me happy about all the time I spent on the details. This is not to say that I think you should work on products until they’re absolutely perfect before launching — I definitely could have launched sooner but I didn’t so I’m glad that at least the long delay was somewhat justified by a more polished user experience.

  • Lots of signups

    • Browserflow’s user base more than tripled from less than 500 to over 1500 users

  • Welcome email

    • When a user signs up for Browserflow, they get added to an email list where they immediately receive a short welcome email from me asking them why they installed Browserflow. I received a ton of responses, and this was great because I got to meet my new users, find out what kind of people they are (developers? non-technical users?), and what use cases they had. It was a lot of emails to get through, but I’m so glad that I provided an easy way for people to get in touch.

      • Does this mean that in the future, I should consider adding a live chat bot (e.g. Crisp, Intercom) to the site when it launches? It would certainly increase the number of conversations I have, but I don’t think I could’ve handled the load by myself, especially since those conversations have an expectation of a synchronous response rather than an asynchronous one.

  • Unexpected connections

    • Besides users of the product, I had people like other founders reach out to connect which was super cool

Here are the things that could’ve gone better:

  • No sales

    • Despite having lots of people sign up and write nice things about Browserflow, no one upgraded to any of the paid plans for the first two days after the launch. I thought that was a pretty bad sign, but in retrospect, I think people just wanted more time to use the tool and it took some time for some of them to hit the initial usage limit on the free plan (30 minutes)
    • I felt like I started really benefitting from the launch when I released a special launch deal in response to emails I received from individuals who wanted to use Browserflow a ton locally but couldn’t justify the higher tier prices (because they’re priced for businesses). Once I created the deal and sent out an announcement, I started getting more subscriptions.
    • If I were to do this over, I would’ve put the special launch deal before launching so that more people would’ve known about it. It might’ve helped more of the 40k visitors give Browserflow a try and possibly upgrade because I’m sure some of them looked at the pricing page and decided the product wasn’t for them.

Overall, this launch went as well as I could’ve ever hoped and I’m grateful that the product resonated so well with the HN community. Even though launches don’t matter nearly as much as makers often think they do (they’re momentary blips and aren’t a sustainable source of growth), this one was definitely a welcome boost. Now the real work begins!