Product Qualified Lead (PQL)

What the signal is

Demonstrated product behaviour rather than stated interest. Because using a product costs the user time and effort, the signal is generally harder to fake and more closely related to value received than engagement with marketing material.

Fitting the threshold

The common failure is choosing a threshold that sounds meaningful — five users invited, ten reports created — without checking whether it separates buyers from non-buyers.

1. Assemble accounts with known outcomes over at least two quarters, with their usage history.

2. For each candidate signal, compute conversion rate above and below several thresholds. Keep signals where the gap is large and the population above the threshold is big enough to matter.

3. Prefer signals that indicate value received over signals that indicate effort spent. Reaching a limit is a stronger indicator than logging in frequently.

4. Combine account-level and user-level signals. One enthusiastic user in an otherwise inactive account is a different situation from broad adoption.

5. Validate on a holdout and re-fit quarterly, since product changes move the meaning of every threshold.

Where the concept breaks

• No free tier or trial means there is no behaviour to observe, and the model has nothing to run on.

• Enterprise buying groups include people who never touch the product, so usage signals miss the economic buyer entirely.

• Heavy usage by a user without budget authority generates a PQL that no sales motion can convert.

• Signals fitted before a product change silently stop meaning what they meant, which is why re-fitting is a maintenance requirement rather than an improvement.