Signal Verifier
Special Award · Build It Anyway (Apify hackathon), Oct 2026

Buying signals, verified

Catch every real buying signal. Skip the false alarms.

Funding, new leaders, hiring, expansion. Signal Verifier treats each signal as a hypothesis, checks whether the company is really ready to buy what you sell, and alerts your team when the evidence holds up. Accounts that aren't ready yet stay on a watch list and are re-checked every week, so nothing slips through.

Case · Company E Checked Oct 9, 2026

Signal spotted

Reported $700M raise at a $10B valuation

Hypothesis: they are building a sales team and will buy sales software within about 3 months.

  • −2Only a signed term sheet, per an anonymous source
  • 0Nothing announced by the company itself
  • 0Recent posts are R&D, awards and partnerships
Score −2 · threshold 5 Suppressed

The problem

Signal tools flood reps with alerts. The real buyers get buried.

When every alert looks the same, reps either chase all of them or start ignoring them. Both ways, the account that was actually ready gets reached late, or by a competitor first. Three headlines from one week:

$700M

The round isn't closed

The biggest headline of the week was a term sheet reported by an anonymous source.

$220M

No one to sell to yet

A research lab with a waitlist and no sales team. Too early for a sales stack.

$153M

The money goes elsewhere

A round raised to scale clinical work, not to build a go-to-market team.

Each looked like a hot lead. Checking every alert by hand eats a rep's week; Signal Verifier checks a batch in minutes and keeps the real ones at the top.

How it works

Every signal is opened like a case file, and kept open until it's resolved.

Apify collects public data from LinkedIn and the web. Claude runs the investigation and classifies each piece of evidence. The points come from fixed rules, so every score can be checked.

  1. Filter

    Skip companies outside your target field before spending anything on them.

  2. Spot the signal

    Funding, a new leader in the team you sell to, relevant hiring, expansion or a product launch.

  3. Form a hypothesis

    For example: they are building a sales team and will buy within 3 months.

  4. Go back and check

    Ask what else would be true if they were really about to buy. Check careers pages, job descriptions, news and the people on the team.

  5. Score and decide

    Alert, watch or suppress. Alerts come with who to contact and a two-line opener that cites the evidence.

  6. Keep watching

    Watched accounts are re-checked every week. When new evidence appears, such as a first sales hire, they move up to an alert.

Scoring

Fixed rules, no made-up percentages.

The model only classifies evidence. A contradiction costs more than weak support adds, so an alert is rare and worth acting on. No source, no points.

EvidencePointsExamples
Strong support+2Funding announced with a go-to-market goal · 3+ sales roles open · sales leadership hired in the last 6 months
Weak support+1Funding with no stated use · 1–2 sales roles · conference booths or customer events this month
Contradiction−2Round only rumored · money going to R&D or infrastructure · layoffs · stale or missing sales roles

Alert 5 or more   Watch 2 to 4   Suppress 1 or less

Sample report

10 freshly funded companies, checked on Oct 9, 2026.

10companies checked
5skipped, outside the target field
1worth a call that week
2false alarms suppressed
~3 minper batch, under $1 in data costs
CompanySignalWhat the check foundScoreVerdict
Company A$255.5M Series BCRO and regional sales directors hired since leaving stealth; booth at an industry conference that week5Alert
Company B$223M Series B4 sales roles open, but the round is mainly for compute4Watch
Company C$42M Series BSales team in place; open sales roles not confirmed3Watch
Company D$220M roundResearch lab with a waitlist; no sales team found0Suppress
Company E$700M, reportedOnly a term sheet from an anonymous source−2Suppress

Company names are withheld on this page. The full report links every point to a public source and its date, and is delivered as a shared sheet with a "right / wrong" column for your team.

Who it's for

For anyone who sells to startups after they raise.

A new round opens a short buying window. The question is the same for every vendor: is this company ready for what I sell, right now? Each product gets its own hypothesis and its own evidence checklist.

GTM recruiting firms

Know which funded companies are about to hire sales reps, before every other recruiter calls them.

Fractional CROs and RevOps consultants

Find companies that just raised and have no revenue leader or sales process yet.

Compliance, payroll and finance vendors

Reach startups right as they move upmarket or set up finance and HR, not six months later.

Sales teams selling to startups

Point reps at accounts that are ready, and stop spending credits on the ones that aren't.

Buying momentWho sells into itEvidence the check looks forStatus
Building a sales teamRecruiters, fractional CROs, sales toolsFirst sales leader hired, sales roles opening, go-to-market goals in the announcementTested
Moving upmarketSecurity and compliance tools, auditorsEnterprise customers announced, security or compliance hires, trust pages addedOn request
Setting up finance and HRFinance, payroll, benefits and legal servicesFirst finance or people lead hired, headcount growth, new office locationsOn request
Scaling the productCloud, data and AI infrastructureEngineering hiring, infrastructure roles, launch and usage announcementsOn request

Tested so far on US B2B AI startups, Series A to C, 50 to 500 people. Other buying moments are set up with you, based on what you sell.

Pricing

Try it free for two weeks.

No card needed

First two weeks free

  • A weekly report on your target accounts
  • Rules tuned to what you sell
  • Tell us which alerts were right, and the rules get sharper for your team

After two weeks

$300 / month, standard rules
$800 / month, rules customized to your product
  • Monthly, cancel anytime
  • Delivered to your inbox or a shared sheet every Monday

Questions

What teams usually ask.

We already use Clay or Apollo.

This works on top of them. Those tools tell you something happened; this checks whether it's real. Send the accounts they flagged last week and compare.

How accurate is it?

Test it. Send 20 accounts where you already know the outcome, and see if it sorts them correctly.

Where does the data come from?

Public sources only: company pages, job posts, news and company announcements. No personal emails or phone numbers are collected.

What happens to our account list?

It is used only to produce your report. Your feedback is combined anonymously with other teams' to improve the rules, and your accounts are never shared.

Get a free check on 20 of your accounts.

Results in 3 business days. No call needed unless you want one.

  1. Reply to Xinran's email or LinkedIn message with 20 target accounts (company names or LinkedIn URLs).
  2. Include a few where you already know the outcome.
  3. Get back a scored sheet with sources for every point.
Founder Xinran Shi

UC Berkeley, Applied Mathematics. I run every report myself, so you work with me directly.

Message me on LinkedIn