Case study / Venture-backed / Series A
A Series A AI infrastructure company: three TechCrunch beats on launch day.
A small, newly funded AI infrastructure company briefed three TechCrunch beat reporters before announce day. Result: a feature, two follow-ups, a vertical trade cite, a Hacker News front-page surge, and 14,000 GitHub stars in 30 days.
The challenge
The company had just closed its Series A and was preparing to open-source its agent framework. The CEO had been pitching reporters in a quiet way for six months. Opens were happening. Replies were not. The team thought the technology would speak for itself. The technology had been speaking for six months and TechCrunch had not heard it.
When they reached out, we had three weeks to launch. Their first instinct was to send the press release to a wire service. Our first move was to pull it back.
"We had been pitching reporters for six months and getting opens, no replies. Digitize spent two weeks on the angle and then everything moved. Two TechCrunch beat reporters had already opened the conversation by the time we sent the embargo."CEOSeries A AI infrastructure company
What we did
A six-week sprint, in six steps.
Angle workshop. The framework was not the story.
"We open-sourced our agent framework" is a press release. "Open vs closed agent ecosystems: which side is winning the developer mind?" is a story. We landed on the second framing in week one and structured every pitch around it.
Reporter list. Not a database export.
Daniel pulled 18 months of TechCrunch AI infrastructure coverage and ranked reporters by depth, not frequency. We picked three. Each got a different angle that fit a thread they had already been writing about.
Embargo strategy. Three reporters. Three angles.
We briefed each reporter individually a week before announce day. The TechCrunch lead reporter got the exclusive feature angle. The two follow-up reporters got the secondary angles for pieces that would land in the week after the launch.
Hacker News orchestration.
The CEO posted to Hacker News two hours after the TechCrunch feature went live. We pre-staged the HN title with a verb and a number, primed the team to be available for technical questions in comments, and the post hit page one within four hours.
Founder LinkedIn rollout.
Three ghostwritten posts staged across launch week, each tied to a different TechCrunch placement. The founder's LinkedIn audience grew roughly 4x in 30 days.
The vertical follow-throughs.
The company's first customer was a vertical software platform. We placed a cite in that vertical's leading trade publication naming the product as a tool the industry was watching, and a byline by the CEO on agent framework architecture in a developer-focused publication. Both moved real demand.
The results
Six weeks. Counted in coverage and stars.
Coverage placed: TechCrunch (3 pieces, including a long-form feature), a developer-focused publication (1 byline), a vertical industry trade publication (1 cite), Hacker News (front page, hundreds of comments).
What we would do differently
Honesty section.
We underestimated the inbound that would land after launch. The CEO and three engineers spent two weeks of post-launch handling demo requests when they should have been shipping. Next time we would pre-build a sales triage process and brief one team member specifically to handle press follow-up so the engineering team could ignore the inbox.
Common questions
About this case.
How long did the launch campaign take?
Six weeks from kickoff to launch day. Three weeks of angle work and reporter briefing. One week of embargo coordination. Two weeks of post-launch follow-through and second-wave pitches.
What did this engagement cost?
Engagement fees are confidential, and this engagement predates our current pricing. A comparable program today is a Tier-One Launch, from $25,000 per project, followed by an Authority Retainer for ongoing thought leadership and follow-on coverage, from $15,000 a month with a six-month minimum.
Do you take on seed-stage companies?
Rarely. Most of our venture-backed clients are Series A and later, preparing a launch, a raise, an expansion, or an exit narrative. We make exceptions for well-funded seed companies with a shipped product and a clear category story. Pre-product startups are not a fit.
Why don't you name the company?
Client confidentiality. The launch and results are real. We withhold the company name, founders, and product details. A reference call is available once a proposal is on the table and a one-page NDA is signed.
How did you choose the TechCrunch reporters to brief?
Daniel pulled the last 18 months of TechCrunch coverage in AI infrastructure, ranked the reporters by frequency and depth of coverage, and we picked three who had each written a piece about agent frameworks specifically. We sent each a different angle that fit their prior coverage thread.
What surprised you about this launch?
How quickly the GitHub stars compounded after the first TechCrunch piece dropped. We had modeled 3,000 to 5,000 stars in the first 30 days. Actual was 14,000. The Hacker News surge after the second TechCrunch piece was the turning point.
Related
MULTI-LOCATION
A multi-location home services operator: from page two to AI Overview
How a multi-location regional brand won the Map Pack and the AI Overview in six months.
FOR VENTURE-BACKED COMPANIES
PR for launches, raises, and expansions
What we do for Series A and later companies, how engagements are structured, and what to expect.