Content for developers with Karl Hughes
Recorded in May 2022, as Oracle, Atlassian, and Twilio's share prices were cratering and VCs were telling portfolio companies to cut anything without near-term payback, Karl Hughes made the case that content marketing is one of the few genuine moats a devtool startup can build — because unlike an ad budget, a competitor can't buy their way past years of accumulated content overnight.
Karl Hughes is founder of Draft.dev, a content marketing agency for software engineers that grew from zero to over 80 clients in two years. He was previously CTO at a venture-backed startup and guest lectures at Kellogg Business School.
We get into why content is a moat that compounds the way paid ads and cold outreach never do, why he spends more time on the promotion checklist after writing a piece than on the writing itself, why developer content is quietly behind the curve on video formats like TikTok and YouTube, and why ignoring developers outside the US and Western Europe means missing most of the market's future growth.
The Q&A below is experimental — extracted from the episode transcript with AI. If you spot a mistake, let us know at jack@scalingdevtools.com.
It's May 2022, and Oracle, Atlassian, and Twilio's share prices are all down sharply, with VCs telling startups to cut marketing spend that doesn't pay back quickly. Content is seen as one of the more reliable growth channels for devtools, but also one of the slower ones. How should devtool startups think about content against that backdrop?
At Draft.dev we work with over 80 developer tool companies, almost all VC-backed, mostly Series A and beyond, so we'll see some scaling back of marketing budgets eventually. But in content specifically, I don't worry about it as much as some other service providers do. In developer marketing, paid ads and cold outreach are both genuinely hard channels to scale because of ad blockers and how skeptical developers are of both. If those channels are working for you and content isn't your strength, do what works — double down on your best fit, especially with a tight runway. But what people miss when they stop investing in content is that they stop building the moat that keeps competitors from catching up. Look at Cloudflare or DigitalOcean — both hit on stock price, but what keeps delivering new customers without throwing more money at it? Content. They've built a moat that no new entrant can replicate overnight. With paid ads, if a competitor raises a new round, they can simply outspend you and you can't catch up. Content you've already built up, they can't replicate quickly, and no amount of money makes up for time. Founders have to think about existing for ten years, differently than a VC managing 100 or 200 companies across shorter fund cycles.
For an early-stage company that needs the fastest possible payback, is there a case for deprioritizing content anyway?
There are exceptions to every rule, and for seed-stage companies there's a good argument to focus on whatever pays back quickest first. But some percentage of resources should always go toward long-term building, because a short-term spend might get you a spike of users, but without a replicable channel that gets cheaper to acquire users over time, you're a money pit that just needs more and more VC funding. Companies that find a flywheel — content, community, a viral loop — get the cheapest acquisition once the flywheel is actually spinning, and that's what excites the next round of investors far more than a short-term spike from ads or cold email.
Beyond acquisition cost, does content change conversion too, since developers often buy based on what they've read?
Content augments other channels. Stack Overflow's whole site is content, and developers know the Stack Overflow brand because of that content, not because of cold emails — though they might send some cold emails to push someone toward signing up for the job board, brand equity from content is what makes that work. On day zero, trying to get your first ten or hundred users, cold email might genuinely be the right call because you don't have content yet. But once you have thousands of pieces of valuable content and people already know you, augmenting with paid ads or cold email gets much easier. Content is also part of the product experience itself — developer relations teams often sit under support, product, or engineering rather than marketing, because content that shows free users the advantage of a paid tier, or helps a user self-solve a problem instead of booking a support call, works as a retention and upsell channel just as much as an acquisition one. Devtool products generally need tutorials, integration guides, and API concept explainers — content crosses over into documentation and helping users succeed with the product, unlike something like cold email, which purely serves growth and nothing else.
How does the reputation of a devtool itself factor into all this?
This is the same Stack Overflow point — a company like DigitalOcean, with thousands of tutorials on setting up Linux servers, can get away with things a company that's contributed less to the community can't, whether that's driving more traffic and trial signups or building goodwill that generates referrals. A good reputation makes it easier to get people talking about your tools, whether through meetups, conference talks, or just unaffiliated engineers advocating for you — Kubernetes is a good example of a tool spun out of Google where people who had zero affiliation with Google were talking about it in its early days. I used to give talks on Docker when it was a private company, unpaid, simply because I thought it was cool and wanted to help other developers learn about it. Reputation has a multiplicative effect: every marketing dollar doesn't just turn into one new user, it turns into one new advocate who becomes a hundred new users. The risk with short-term, fundraising-driven thinking is losing track of that long-term compounding, and getting stuck operating out of fear about raising the next round instead of thinking about becoming a billion-dollar company.
You've written a promotion checklist arguing that writing content is only 20% of the work and promotion should be the other 80%. How do you actually think about promotion, and is there anything startups can do to shorten the payback period on content?
If you've spent four to eight hours writing a piece, posted it to Twitter and Facebook, and walked away, I'd encourage spending another two hours on promotion, because it makes a big difference. Share it with relevant newsletters — JavaScript Weekly and countless niche equivalents are all sourcing content every week, and it's a lot of work for maintainers, so they genuinely appreciate brands sending them good content. Same with relevant subreddits — people are afraid of criticism on Reddit, but you can't predict what gets voted up or down, and a single downvote doesn't hurt your brand reputation as long as you're not breaking rules. Slice key points into a Twitter thread or LinkedIn posts, make a video summary for short-form platforms like Instagram or TikTok — developer marketing is a bit behind on this because companies tend to target senior engineers who aren't on those platforms, ignoring the huge population of junior developers ten or fifteen years younger who go to YouTube or TikTok first for dev content. Syndicate to Dev.to and Medium too. Promotion is genuinely underdone in developer marketing, but it's critical if you want a piece of content to consistently get noticed and function as a sustainable growth channel.
Do you have any experience with developer content on TikTok specifically?
Not a ton personally — it's not a platform I'm as native on as the generation of developers younger than me. But I hear about it a lot from our writers, many of whom are junior or mid-level developers just starting their careers, and YouTube comes up constantly too. Personally, if I want to learn something I'll find a written tutorial and copy-paste code, but a lot of our writers rely almost solely on YouTube videos to solve problems. If you're not using the channels where your potential audience actually is, you're missing big portions of that community. It's also a competitive opportunity — if you enter a crowded space where every competitor is doing written content, doing video instead lets you zig where everyone else zags and set yourself apart, which matters if you want a bet to pay off exponentially.
So should everyone be investing in TikTok right now?
Who knows — it's still early to tell. A few years ago Snapchat was all the rage, and it's not gone, but TikTok is the one people talk about now, and it's genuinely hard to know which platforms stick around long enough to become uncool versus fading fast. What matters is meeting your audience where they actually are. On a related note, people also miss how many developers are international — Latin America is going to be huge in the developer space, and we have tons of writers and developers we work with in South Africa, Kenya, Nigeria, and across Eastern and Western Africa. If you build content only for a Western audience and ignore the roughly 75% of the world outside it, you're putting a hard limit on how many people you can reach, especially as remote and distributed teams become the norm rather than everyone returning to the same five cities.
What can teams practically do to make content more accessible to an international audience?
A big part of it is having people on your team who actually know those audiences. We have 16 full-time employees and a couple hundred writers, and working with an engineering manager based in South Africa taught me things people in the US and Europe just don't realize — like latency, since every web request from South Africa has to route up through Europe before reaching the broader world, without the same distribution and nodes we have access to in the US or London. Understanding how to build devtools for low-latency-constrained areas matters. Replit is a great example of a tool being genuinely forward-thinking about accessibility for developers everywhere, even though those users aren't as big a purchasing market as the US right now — but that market will only grow over the next ten or twenty years. When everyone builds for the same ideal customer profile — fast internet, MacBook Pro, based in a major Western city — that market gets saturated, but almost nobody builds for the hundred thousand developers in Nigeria working off tablets, old Windows laptops, slow internet, and intermittent power outages, even as that population becomes more educated and gets more access to coding education and Western employment. There's no such thing as a fully saturated, finished market, because the world keeps changing around it and forces every market within it to change too — that's true whether you're a startup finding your niche or a big company staying competitive once you've hit most of your normal target users.