How VC Firms Find Startups Before Everyone Else: The Pre-Announcement Stack
How VC firms find startups weeks before rounds are announced: the leading indicators (GitHub acceleration, hiring, product telemetry), the tools, and the workflow that turns early signals into first meetings.
Key Takeaway
Every fund says it wants to meet founders before the round is competitive, but only some run a stack that actually does it. This post maps the pre-announcement toolkit: the categories of leading indicators (engineering acceleration, hiring, product and release telemetry, community activity), how each one leads the funding databases by weeks, how funds verify a signal before reaching out, and the weekly workflow that converts early sight into early meetings.
The most valuable question in venture is not "is this a good company?" It is "how early did you meet them?" Meeting a company the week it announces is a pricing exercise; meeting it a month earlier is a sourcing advantage. This post is about the second thing: the specific stack funds use to find startups before the market sees them.
The timing gap is real and measurable. Funding databases like Crunchbase and PitchBook record rounds when they are announced or discovered, typically at or after the wire. But the behaviors that predict a raise start earlier: teams staff up, code accelerates, products ship, and domains get shuffled weeks before any announcement. Funds that instrument those behaviors systematically get a head start everyone else cannot buy, because the underlying data is public and mostly free.
The five leading indicators that matter#
- Engineering acceleration. Commits, contributors, and repositories rising together against the company's own baseline. Across the 350+ organization panel tracked on this site, sustained acceleration has preceded public announcements by a median of three to six weeks. The full method is on the methodology page.
- Hiring patterns. Job postings for roles that precede raises: founding sales hires, finance leads, senior infrastructure engineers. A burst of senior postings usually means the round is already closed and being deployed, so treat hiring as a confirming signal more than a leading one.
- Release and product telemetry. Version numbers, changelogs, app store updates, and landing page iterations. A company shipping weekly is a company with engineering capacity to spare, which itself correlates with recent or imminent capital.
- Community and open-source activity. New public repos, maintainer activity, developer adoption curves. Open-source companies telegraph momentum continuously; the investing in open source startups guide covers how to read it.
- Registry and corporate exhaust. Trademarks, domain registrations, SEC Form D equivalents in other jurisdictions. Useful for confirming, weak for discovering.
The order matters. Engineering acceleration and community activity lead by weeks; hiring and corporate exhaust confirm. Funds that confuse the two categories reach out either too early (nothing to say) or too late (round already competitive).
How verification works before any outreach#
A leading indicator is a reason to look, never a reason to send a calendar link. The workflow that respects both the fund's time and the founder's:
- Confirm the signal is real, not a hack week or a one-off burst. Baseline-relative checks (is this acceleration against the company's own normal?) filter most noise; the signal vs noise guide covers the heuristics.
- Check the round context. Last announced round, its date, and typical interval to the next round for that stage. The pre-seed vs seed vs Series A explainer gives the stage math.
- Read the engineering work itself, not just the counts. What got built tells you what the raise will fund. The GitHub due diligence checklist is a 20-minute pass for exactly this.
- Find the warm path. A shared connection converts 3-5x better than cold outreach; the warm introductions guide covers the ask.
The tools, honestly assessed#
The commercial stack (PitchBook, Crunchbase, Dealroom, Affinity, and the newer AI sourcing platforms) is good at breadth and relationships, weak on lead time: they organize the market as it already exists. The edge lives in sources they underweight. Public GitHub data is the clearest case: free, weekly, and leading. This site runs an open panel on exactly that principle, with a free tools directory covering the rest of the zero-budget stack. For how the commercial platforms compare on sourcing specifically, the Harmonic vs PitchBook comparison and the best deal flow tools list are the honest maps.
The weekly pre-announcement workflow#
Knowledge without cadence decays. The minimal system that works:
- Monday: refresh the watchlist from your leading indicators (for this site's users, the weekly startups to watch list is exactly this artifact).
- Tuesday: verify one to three candidates using the checklist above.
- Wednesday: outreach, warm path first.
- Friday: log outcomes against sources, so next Monday's list gets smarter.
That loop, run for a quarter, produces what no database sells: a proprietary map of which companies were accelerating before anyone wrote about them. The weekly sourcing workflow and how to track startups before they announce expand the loop into full playbooks.
The honest limits#
Lead time is an advantage, not a guarantee. Roughly one in eight acceleration signals in the panel never resolves to an announced round (extensions, quiet SAFEs, or launch-driven bursts), and early sight of a bad company is still a bad deal. The stack gets you the meeting earlier; judgment still closes it. Use the head start to do diligence others skip, not to skip diligence others do.