Since starting PostHog in 2020, we’ve learned a bunch about what does and doesn’t work when it comes to marketing to engineers. There are plenty of articles about ‘developer marketing’ out there, but most of them are either generic or focused on later stage companies.

In this intro article, my intention is to share specific, actionable advice for early stage startups building dev tools. I’ve bucketed advice into pre- and post-product-market fit.

I’ve also included a couple of channels that we haven’t nailed but know other successful dev tool companies consider important. We’re still learning, and definitely haven’t figured everything out!

Trying to go head-to-head with much larger competitors on content output is a losing game, so focus on quality instead. You are not Cloudflare (yet). 1 great article >>> 25 mediocre ones. Quality is more important than hitting an arbitrary content schedule – and makes the writing process a lot more enjoyable. In terms of types of content, we write approx. ⅓ SEO articles, ⅓ tutorials, ⅓ anything goes.

High quality pieces perform better in the long run, especially SEO ones. Treat your SEO portfolio like a VC treats their investments – invest time updating your best SEO pieces, don’t just write something and forget about it.

This applies to everything you write. SEO articles are not exempt from this rule – saying ‘oh it’s SEO so being useful is less important so long as we have good keywords’ is bad marketing. If you wouldn’t be proud to share the content on your personal site, you shouldn’t publish it.

This means keeping it separate from marketing, which is the opposite of what most companies do. While our marketing and website teams work closely together, the website team has the final say on what appears on posthog.com. This stops the website becoming too marketing-y and falling foul of marketing ‘best practices’ that most developers (and people in general) despise.

You should outsource the stuff that you can’t do internally, not the stuff that you can do but don’t have capacity for. If you don’t, you’ll spend an inordinate amount of time managing a freelancer – when this goes wrong, it’s harder work than just doing it yourself.

For example, we work with a freelancer on deep dive technical topics that are interesting to our users but not strictly about PostHog itself. However, we’ve wasted a bunch of time hiring freelancers to write mediocre SEO content that we should have just done ourselves.

Every single piece of content at PostHog is edited by at least one other person. The writer has final say on what gets shipped, but it is everyone’s responsibility to give each other feedback.

Line edits are easy and less useful feedback to give. Instead, start big picture and then go into the detail. If the article isn’t genuinely useful or interesting, punctuation doesn’t matter. Feedback can include ‘ditch this article entirely, it’s not working.’

We’ve found that hitting the front page results in a giant, ego-boosting traffic boost, with a noticeable but small signup boost. It’s not dependable though – even if you’re great at it, you’ll have a 1 in 10 hit rate. You need to graduate to repeatable marketing at some point.

Don’t upvote your own content, and don’t ask other people to – post it and pray. There are no secret tricks.

You can’t rely on UTM parameters to tell you where a user actually first heard about you. Example: user reads an article about PostHog on Hacker News -> searches ‘posthog’ -> clicks on a Google Ad. Our analytics will tell us “wow, Google Ads are awesome!” But that’s not the whole picture.

In your product signup flow, include an optional free text box asking people where they first heard about you. About 10% of signups usually fill this in. Read the data and report on it every week. It’s manual but vitally important info.