PQA Signal Detection
Find the exact usage signals that mean 'engage sales now' using users, volume, velocity, and behavior.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 90%
A Product Qualified Account is only useful if you know which signals actually predict enterprise readiness. Elena Verna identifies four signal families — number of users (the 'magic seven'), a volume threshold, velocity of change, and specific behavioral triggers — and stresses that behavioral signals like an admin switch or a visit to the terms-of-use page are among the strongest and are surprisingly universal across companies.
Origin
Derived from Elena Verna's hands-on PQA modeling at Miro, Amplitude, and MongoDB.
Core principles
- 01Number of users in a company using your product is the single biggest signal; ~7 users is a recurring magic-number threshold (echoing Facebook's 7 friends, LinkedIn's 7 connections).
- 02A volume threshold specific to your product (events sent for Amplitude, boards for Miro, revisions for Figma) signals deep adoption.
- 03Velocity — the rate of change in users/events/storage — is the trickiest to measure but one of the most powerful timing predictors.
- 04Behavioral signals tied to evaluation intent (admin transfer, terms-of-use page visits) are strong, universal predictors.
- 05Shallow signals like watching a webinar or landing on an enterprise page are weak — they trigger annoying premature outbound.
How to run it
- 1
Count users per company and set a threshold
Track how many users at a company use your product; treat crossing ~7 as a strong sign there is enough distributed value for an enterprise conversation.
Pro tip Engineer both pull and push adoption so users cluster into one account — herd mentality inside the company raises this count.
- 2
Set a product-specific volume threshold
Pick the usage-volume metric that shows the account is tackling the whole use case (Amplitude: events; Miro: boards; Figma: revisions) and set the level that indicates broad adoption.
- 3
Track velocity of change
Monitor sudden increases — e.g., going from adding 1 user/day to 15 users in a day, or a spike in events/storage — as a trigger for the right timing to engage.
Watch out Velocity is hard to measure in transactional data, but the change signal often beats absolute thresholds for timing.
- 4
Watch high-intent behavioral signals
Flag actions that only happen during a buying evaluation: an admin switch/transfer in the account, or anyone landing on your terms-of-use or privacy pages — then reach out.
Pro tip Nobody reads the terms-of-use page unless they're considering an enterprise deal; a visit means you likely have a buyer on the hook.
Watch out Do not treat generic marketing touches (webinar views, enterprise landing pages) as buying signals — they trigger spammy, ill-timed outreach.
In the wild
At Miro, an admin switch — a new admin being assigned or admin transfer — reliably meant some evaluation was happening in the account. Verna treated it as 'ding ding ding, all the bells ringing' to reach out and see what was going on.
→ This behavioral signal, later confirmed universal at MongoDB, let sales time outreach to real evaluation moments instead of guessing.
Verna found that visits to terms-of-use and privacy policy pages, which normal users ignore, spike only when someone is considering an enterprise deal.
→ Reaching out to anyone from an account who lands on those pages surfaces a likely buyer you can help increase perceived value.
Common mistakes
Treating shallow marketing touches as buying signals
Watching a webinar or accidentally landing on an enterprise landing page isn't real intent; acting on it spams users with irrelevant SDR sequences and erodes channel trust.
Ignoring velocity because it's hard to measure
Velocity is difficult to compute from transactional data, but skipping it means missing one of the strongest timing predictors of when to engage.
Is it for you?
Best for
Growth/analytics teams building a first PQA model for a usage-rich B2B product
Not ideal for
Products with too little usage data or where a single user already equals the buyer
From the transcript
“one of the biggest pqas is number of users in the company using your product by the way the magic number is usually seven”
“your terms of use Pages your privacy and policy nobody cares about it unless they're considering Enterprise deal if you see anybody from account land…”
“if there is an admin switch in the account some sort of evaluation is happening”
“the change in velocity either in number of users being added or events being sent or storage that is being utilized is usually a fantastic…”
From the episode
The ultimate guide to product-led sales
Elena Verna