July 2019: Hacker News Launch
- Subscribers: 2,163 🎉
- Users: 28,353 🕵️♀️
- Pageviews: 124,026 👀
- Trends: 1672 📈
- Insights: 54 👁
Check out the live stats: Open Page
Our second month was massive as we launched on Hacker News and the response was fantastic. A lot of quality discussion and Trennd even reached #1 on the front page (albeit very briefly!)
This Show HN went on to have a huge domino effect onto Twitter. Dozens of people tweeted about Trennd.
Amazingly we even got a shoutout from Rand Fishkin. This made my month. 🤩
All of this combined to rocket us above 2000 email subscribers. I’m ecstatic and excited to build this thing out further and see where we’re at in 12 months.
Show Hacker News
So Friday 12th at 10pm and I’m sat at my desk in Japan ready to post to Hacker News. My wife’s already gone to bed, but I figured this was the best time to post such that people were awake in both US + Europe. Plus I could reply to any comments and keep check on things through to Saturday morning JST if I had to.
So I drafted my opening thread comment and hit submit.
"Sorry, your account is too new to submit this site."
Oh man. I'd been thinking about this submission all day. And I couldn't even post. 🤦🏼♂️
Anyway, I emailed the mods at Hacker News to explain and they very quickly and kindly marked it ready to go through. My account was made in 2017, but turns out “too new” can also mean not having participated much in the community.
Flash forward to the following Friday and this time it went through.
We’d made sure to phrase the title to resonate with the audience. On Hacker News (and in most places I suppose) simple, clear and humble works best. Big shoutout and thank you to fellow indie maker Vincent here. He helped me to get this title spot on, along with a bunch of other great advice that made a huge difference.
The post immediately moved onto the best of Show HN page, and then quickly onto the main page.
Traffic increased to 400 concurrent users on Trennd and we briefly peaked at #1.
It was at this point that Trennd got the infamous "Hacker News hug" and the site died. The free MongoDB Atlas database plan did not appreciate this flood of connections. So I had a frantic 15min where the website was down while I upgraded to the paid tier.
It’s a shame that I lost this quarter hour of Hacker News prime time. And I’ve since implemented some server-side caching that prevents the database from getting hammered.
But since I recovered quickly, we remained near the top of the front page for 12 more hours, which is good enough for me. Let’s not be greedy. 😛
I manned the comments section into the early hours of the morning, got some great discussion and a tremendous amount of positivity from the crowd.
Just as I started to think everything was all over and we were back to normality. The buzz started to spill over to Twitter.
Thankfully I’d updated the twitter preview image beforehand, so there was a pretty placeholder image and description when dozens of people shared and tweeted about Trennd.
And then, the cherry on top of it all was that Rand Fishkin tweeted about it! Unbelievable! Made my month. 🍒
The number of likes and retweets on the back of this was insane too.
I’m a twitter newbie, but it’s where everybody in the tech world hangs out to talk about new things. A perfect stage for Trennd, so I need to be more active there.
Other newsletters and blogs
This exposure also led to us getting picked up by the hackernewsletter and kottke.org which drove some good traffic our way too.
User Submitted Trends
Overall, it was really cool how ready people were to contribute and add trends of their own. It confirmed my hypothesis that we might be able to crowdsource the surfacing of new trends to some extent. In a similar way that Product Hunters crowdsource new products.
But with this comes the issue of quality control. One visitor for example added “Donkey Porn” as a new trend, and it may well be trending, but I had to moderate it away.
While people are adding trends, not so much insights. This is obviously because the amount of work it takes to write and submit something insightful is massively greater than punching in a keyword. But if I want there to be meaningful content around each trend page, I need to figure out how to better incentivise people to add insights too.
The newsletter is the one thing that I make sure happens every week. It’s great motivation to get things done so I have something to talk about, but also keeps me in touch with Trennd’s growing audience.
Feedback from people about newsletter though was that it was becoming uninteresting. Mainly because they didn’t know what to do when faced with 10 different links to unknowns. And I get it. Something had felt off with the format for a while.
So I changed it to 5 trends (instead of 10+) and started to add a small description for each.
This new format feels much closer to what it needs to be now. And I’ve even had a few people email saying “much better format” which was a nice affirmation.
Now we’ve truly validated interest in the project with over 2000 subscribers, marketing to get another X hundred subscribers isn’t going to change anything.
Now I need to make the platform more powerful. And I’m going to hold off on the Product Hunt launch until then.
The #1 challenge is the noise to signal ratio - a pain that’s already starting to be felt. With almost 2k trends already, nobody is going to click through each page of 30 trends to discover hot trends relevant to them. They need to be able to slice and dice the trends trends in more ways/dimensions that are relevant to them. Like for example sorting by absolute search volumes too.
More consistent categorisation and classification of trends would also be good, which in reality means an automatic labelling solution. I have a few ideas in mind for this.
But primarily, greater quality of trends versus quantity requires stricter curation on my part. And I’ll create some systems to help with this.
I’d also like to start monetizing the site so that it can pay some of its own bills and maybe even some of mine!
In the medium to long term, a premium version of Trennd is perhaps the way to go. But right now I’m still figuring out what this should look like and I want to do it right. I don’t want to damage any goodwill earned so far by pushing out a rushed premium product too soon that’s the wrong fit.
So in the meantime, I’ll look to find a community or affiliate sponsor with a great product that’s a good fit. This could potentially get a very small amount of income coming in - which would be a start!
All of this is a lot for one month though, so I admit it’s definitely going to take longer than that.
Any thoughts just let me know: [email protected]
Otherwise, see you next month!