Buying Signal Tracking Template
Two documents and three sheets that log every signal's source, then roll up the response rate by type with the sample size sitting beside it.
Free download · No account needed
Response Rate by Signal · last 90 days
The highest reply rate sits on the thinnest sample. The rate worth trusting is the one worth cutting.
| Signal type | Acted on | Rate | Sample | Verdict |
|---|---|---|---|---|
| Funding news | 9 | 44.4% | Thin | Keep watching |
| Leadership change | 15 | 33.3% | Thin | Keep watching |
| Job posting | 42 | 14.3% | Measured | Keep |
| Site or pricing change | 61 | 4.9% | Measured | Drop or scale back |
Ranked by raw rate, funding news wins by thirty points. Ranked by what the sample actually supports, it is the least trustworthy row on the sheet.
Search this term and page one hands you the same scoring rubric five times. Rate every job posting, funding round, leadership change and site change on intent, fit and timing, add the points, and route the highest scorers to immediate outreach. The rubric is a reasonable guess at what matters before anything is tried. What almost nothing checks afterward is whether the guess was right, so a signal type that scores well and a signal type that actually earns a reply keep getting treated as the same thing quarter after quarter.
The fix is to log what happened and roll it up by type. A response rate computed on a small sample carries a wide margin of error, so a type acted on nine times can show a rate thirty points off its true one by chance alone. Across a fictional forecasting vendor's quarter, funding news shows the highest raw rate, 44.4%, on only nine instances, too thin to trust. Site and pricing changes show the lowest rate, 4.9%, but on sixty-one instances, the most measured number on the sheet, and the one to confidently cut.
Built for the SDR or account executive running outreach off a watch list, and for the manager deciding where a team's signal-checking hours actually belong. Source every account first with the account research brief. Once a raw signal turns into a real conversation, the outbound sequence pack and the territory account plan take it from there, tracking the reply and the relationship rather than the trigger that started it.
What's in the pack
A Rate Needs Its Sample Before It Earns Trust
The method itself: no rate gets acted on below thirty instances, no matter how high or low it looks.
Signal Definitions doc
Exactly what counts as a job posting, funding news, a leadership change or a site change, and the real source each one has to come from.
Outreach Angle per Signal doc
One angle per signal type, each tied to what the signal actually implies rather than a template with the name swapped in.
Signal Register sheet
Every signal detected, with its date, its source, whether it was acted on, and what happened.
Account Watch List sheet
The accounts being watched and which of the four signal types actually apply to each one.
Response Rate by Signal sheet
The rate per signal type, the sample size and confidence interval beside it, and a verdict of keep, keep watching, or drop.
How it works
- 1
Bring the watch list
The accounts you're tracking and which of the four signal types actually apply to each one.
- 2
Log every signal
Date, source and outcome for everything acted on, whether it earned a response or not.
- 3
River rolls up the rate
By signal type, with the sample size and a measured-or-thin label sitting beside every number.
- 4
Act on what the sample supports
Keep, keep watching, or drop, decided by the confidence interval rather than the raw rate alone.
Frequently asked questions
Is this template free?
Yes. Two Word documents and three CSV sheets, downloaded as a zip, no signup. River is the optional half: it logs every signal with its source, rolls up the response rate by type, labels each rate measured or thin, and writes the verdict on what to keep, keep watching, or drop.
Why would a lower response rate ever be the one to trust?
Because a rate on a handful of instances can move by thirty points the next reply either way, and a rate on sixty instances cannot. In the worked example, funding news shows the highest raw rate at 44.4%, on nine instances. Site changes show 4.9%, on sixty-one. The second number is the one with enough sample behind it to act on with confidence.
What counts as a real source for a signal?
One that is checkable later: a job board the employer runs itself, a company's own funding announcement, or for a leadership change, an SEC Form 8-K filed under Item 5.02 where the company files one at all. A personal profile update with no independent source behind it does not qualify.
Why thirty instances specifically?
Thirty is not arbitrary; it is roughly where the normal approximation behind a simple confidence interval on a proportion starts to hold up, so the interval it produces is trustworthy rather than decorative. Below thirty instances acted on, a rate is labelled thin regardless of what it shows, because the interval around it is too wide to act on.
Does this run the monitoring automatically?
Not yet on its own. You bring what you or an integration already surfaced, job postings, funding news, leadership changes, site changes, and River logs the source, rolls up the rate by type, and writes the verdict. The finding is the math on top of what you feed it, not a background crawler running unattended.
How is this different from the outbound sequence pack?
Different stage of the same job. This decides which signal types earn outreach at all and logs the source and the rate behind that decision. The outbound sequence pack takes over once outreach is actually sent, coding what every reply said rather than just whether one arrived.
Find out which of your signal types is actually worth the hours it's getting
Send your watch list and what you've already acted on. The rate, the sample size and the verdict come out of that alone.
Edit with AI