Demand generation for devtools and AI startups

What demand generation actually means for a devtool

Demand gen is the machine that connects everything upstream of a closed deal. Outbound, lead gen, email and lifecycle, funnel design, and the plumbing that keeps sales and marketing chasing the same targets. For a product-led, self-serve tool it also owns the path from first signup to activated user to paying customer. If a program does not eventually show up as pipeline or revenue, it is not demand gen. It is decoration.

Most early devtools do not have a demand problem so much as a conversion and follow-through problem. Engineers are already finding you in the docs, in a Slack community, on Hacker News at midnight. What is usually missing is the system that catches that interest, qualifies it, and moves it forward on purpose instead of by accident. That system is what I build.

Why it is different when your buyer is an engineer

Marketing to engineers means selling to the person who actually uses the tool, not a committee that will never open a terminal. Engineers do not want to be marketed to. They want to try the thing, read the docs, and decide for themselves. So the job is not persuasion. It is removing friction and showing up with something genuinely useful before they ever talk to a human.

Attribution also breaks for this buyer, and you have to design around that reality instead of pretending it away. This is the dark funnel. Someone reads a Hacker News thread, lurks in a Slack, skims your docs at 1am, mentions you to a teammate, and then signs up directly two weeks later with no traceable path. Your analytics will happily credit that to direct traffic and teach you nothing. If your reporting only trusts what it can perfectly track, you will defund the exact channels that are actually working.

The practical consequence is that I do not optimize for what is easy to measure. I optimize for what moves the business, and I use a mix of self-reported attribution, signup surveys, and directional signals to stay honest about where demand really comes from.

The programs I actually run

Outbound is the fastest lever when you need pipeline you can predict. I build the whole engine from scratch, targeting, list building, messaging that sounds like a human wrote it, sequencing, and the inbox infrastructure that keeps it deliverable. For one client that outbound engine drove roughly 34x growth in LinkedIn visibility plus dozens of customer meetings, starting from nothing. Outbound to engineers only works when the message respects their time and leads with something real, so that is how I write it.

Email and lifecycle is where most devtools quietly leak revenue. A signup that never activates is a lost customer, and a nurture sequence that reads like a newsletter is worse than nothing. I build the onboarding and lifecycle flows that get a new user to their first real win, then keep pulling active users toward the paid moment. Every message earns its send.

Funnel design is the connective work that ties it together. I map the actual path a buyer takes, find the stages where people fall out, and fix the specific step that is broken instead of dumping more volume on top of a leaky funnel. Then I make sure sales and marketing are aligned on the same definitions and the same numbers, because a lead marketing calls qualified and a lead sales calls garbage is a very expensive disagreement.

How I measure it without lying to you

Vanity metrics do not survive a conversation with me. Raw sessions, impressions, and follower counts tell you almost nothing about whether the business is growing. I report on signups, activation rate, sourced pipeline, and booked meetings, because those are the numbers a founder can actually take to a board or a bank.

Because the dark funnel means perfect attribution is a fantasy, I combine hard tracking with self-reported data. A short how did you hear about us on signup will teach you more than a month of staring at a multi-touch attribution model that quietly rounds everything to direct. I would rather be directionally right about what is working than precisely wrong.

You will always know what I am spending your time and money on and what it is returning. If a program is not producing, I kill it and say so. No dashboard theater, no burying a flat quarter under a chart of rising impressions.

How an engagement runs

I work on a monthly retainer, starting at 3,500 USD, Monday through Thursday. I am American, based in Spain, and I cover EMEA and North America, so overlap with both sides of the Atlantic is built in. Retainers mean I am in the weeds with you, not parachuting in for a strategy deck and vanishing.

The first few weeks are diagnosis and quick wins. I learn the product well enough to talk to your buyer, audit what you already have, find the one or two things that will move a number fastest, and start there. From there we build the durable programs, outbound, lifecycle, funnel fixes, and the reporting that keeps everyone honest.

I have run GTM inside more than 30 companies over 12 plus years, including IBM, Zerve, Grepr, Steadybit, OpenMeter, DBmarlin, Instana, Shoreline, and Prodly. That range means I have seen what works for a scrappy seed-stage devtool and what works at enterprise scale, and I know which lessons transfer and which ones will sink you if you copy them at the wrong stage.

What this covers

“Summer was amazing to help us get our initial outbound engine going, and her work has helped us increase our visibility around 34x on LinkedIn as well as land dozens of meetings with customers that have been essential in our growth.”

Questions on this service