Marketing is hard
Jun 17, 2025
Ayush, Autumn Co-Founder
Are founders ever happy with their product? There are so many different bugs, ideas, features and requests that I would love nothing more than to spend all my hours on. Designing product is my happiest state.
Unfortunately, while I like to believe I have good taste, the product is the easy part. Getting people aware of what we're doing and convincing them its worth trying is what matters in today's world.
90% of my life goes into thinking about GTM. This is a write up of what we're exploring and some useful resources.
Who are we targeting?
When we initially launched Autumn, we were targeting seed and series A VC backed startups. Our GTM was cold outbound, main KPI was revenue. It was actually growing pretty well. But we realized 2 things:
Getting people to integrate was heavily timing dependant
There was a small segment of inbound users using it for side projects
We pivoted to make a bigger bet on a longer term strategy of becoming the next go-to in payments. Since April, our strategy has been to target day 0 builders and founders.
How are we doing it?
Our retained users are growing ~11% week on week. It's a broad, bottoms-up GTM.
The main channels we spend time on are:
X/Twitter
This is where most people find us. We've had some success posting demos and launches on @johnyeo_'s account (2K followers) and have a smallish following (600 followers) on @autumnpricing.
Our accounts are active (2-5 posts per day) and have reasonable engagement. 3 months ago, we had ~200 followers in total. If we were starting again we would:
Follow relevant accounts in our ICP, and other small active accounts trying to grow. YC is an advantage here as we already have a network of people like this.
Engage with these accounts a lot: 100+ replies per day, and follow even more accounts.
People will follow you back and engage with you too, which is good for the algo. I personally follow back anyone who (1) follows me and (2) engages with me.
Possibly even start a group chat to help keep accountable and comment, repost and bookmark each other consistently (if you're doing this, we'll join!)
Note: all of this only really works because there are people that like our product and are interested in what we're building. If you don't have a likeable product you should fix that first before shouting about it
Our most successful post was honestly total nonsense clickbait: a 1 minute demo where we built "cursor for stripe". When you spend enough time on twitter, you start to get a sense for what a good tweet looks like and what the "current thing" is. AI video content doing something novel works the best, with a snappy tweet that jumps straight into it.
Going viral is great. When it works, it works well, but is super variable. Chasing virality over and over again is a repeatable GTM for consumer apps, but harder in the tech twitter sphere. A lot of the time it feels like shouting into the void.
Resend's heartbeat framework resonates a lot with us. We have to constantly be building, writing about it, and shipping.
Hacker news
We had 1 blog post of ours that hit front page of HN for about 3 hours, and it drove more traffic to our site than we'd seen in the whole month. Even though most of it was negative (people found the title to be misleading), it drove a lot of sign ups.
Blogging (like this)
Our blogs probably get about 300-400 reads per week, which we're pretty happy with as we've just started. We're trying to lean more into creating content that we'd find interesting, hoping that other founders and devs also like it. When we write, we share it on twitter, our email newsletter, and a few relevant subreddits.
Word of mouth
We're fortunate enough that some people have really experienced the pain we solve and like what we're doing. Because of that, we often find ourselves being recommended in twitter threads, and people saying friends said good things about them. This is one that feels more out of our hands.
Other than a few upcoming platform launches (Launch HN, Product Hunt), we want to try a few more channels properly like LinkedIn and Reddit. The advice we've heard is that you never really know what's going to work for your startup, so try a lot of things.
We are also thinking about how we can bake marketing into our product a bit more (eg Autumn branded receipts on purchase).
What is our message?
I used to think this wasn't important, then I heard Ant (Supabase founder) talk on the scaling devtools podcast about how they went from 8 hosted DBs to 800 over a weekend, just by changing their messaging from
Real-time postgres DB -> Open source firebase alternative
And launching again on HN.
This is one we're working on. Some angles we've gone for are:
Any pricing model in 6 lines of code
Never deal with billing again
Open source infrastructure for pricing and billing
What supabase did for Postgres, we're doing to Stripe (?)
A lot of devtools took off by iterating on the product over time, and really crystallizing who they're targeting. Shoutout to Clerk who grinded for 3 years, then found success by going all-in on frontend, Next.js devs. Polar.sh (another payments company) had a lot of success by pairing their setup with better-auth.
Anyway
Marketing is a slog. It's trial and error, but with a long and hard to measure feedback loop. I wish we could be one of those teams that just builds it and they come. But until we hit PMF we'll just keep shitposting 🫡