How to run a weekly signal review without drowning

A weekly signal review that does not cap the list is theater with a spreadsheet.

Ehrenberg-Bass and the LinkedIn B2B Institute put about 95% of B2B buyers out of market. Salesforce still puts about 27% of reps at quota. If your review dumps hundreds of "hot" accounts into Convert every Monday, you are feeding the 95% problem into the 73% miss rate.

Cap the weekly list and score buying windows. Vanity volume is not Signals. Meetings booked from the review are.

What is a weekly signal review?

A weekly signal review is one fixed meeting (or async checklist) where you score which accounts show a real buying window this week, then hand a capped list to Convert.

It is not a product tour of every tool in the stack. It is not scrolling LinkedIn for vibes. It is a short operator loop: collect the same three signal classes, score them, kill vanity, ship work that can book meetings.

If the review ends without a numbered list, owners, and a send plan, you held a status meeting. That is not Signals.

Which signals belong in the review?

Three classes. Same every week. Do not invent a fourth because a vendor shipped a feature.

1. Website de-anon. Tools like Vector or RB2B show which companies hit pricing, docs, or high-intent pages. A visit alone is not a window. Repeat visits on commercial pages in a short window matter more than a bounce on the blog.
2. G2 category and competitor traffic. Someone researching your category or a named rival on G2 is closer to a purchase conversation than a cold firmographic match. Treat that as input for scoring buying windows, not as a reason to spray the whole TAM.
3. Competitor LinkedIn engagement. Comments, follows, and shares on rival posts from ICP titles. Job change is a signal. It is not the motion. Engagement on a competitor's offer is closer to intent than "new VP started."

That is the same spine as signal-based outbound and website de-anonymization. The review is where you force the stack to produce a short list instead of a dump.

How do you run the review without drowning?

Sixty minutes. Same agenda. No demos of dashboards inside the hour.

Before the hour: export candidates from de-anon, G2, and competitor LinkedIn. Deduplicate by account. Enrich only what you need for a first touch (title, email path, the signal itself). Leave Clay theater for later.

In the hour:

1. Sort by signal strength, not logo size.
2. Score each account: window open, window maybe, window closed. Write one plain reason.
3. Cap the ship list. I use a hard ceiling: enough that Convert can write signal-matched notes this week without spam. If the ceiling feels too small, your notes are too long or your list is still vanity.
4. Assign owner and channel. One primary person. One primary first touch. Multi-thread later in Convert, not as a Monday free-for-all.
5. Kill anything that is only "they look like ICP." Firmographic fit without a window goes to nurture or nowhere. It does not go to this week's send queue.

After the hour: Convert gets the capped list with the signal quoted in the brief. No "research and figure it out." The signal is the brief.

What vanity do you kill on sight?

Anything that inflates the pile without raising meeting odds.

• Page views with no commercial page path
• "Liked a competitor post once six weeks ago"
• Job changes with no follow-on behavior
• Open rates and reply fluff as success metrics for the review itself
• TAM spreadsheets that restart every Monday because nobody capped last week

Meetings, not theater is the law. If the review cannot name which meetings it is trying to create, stop the meeting and rewrite the agenda.

What metrics prove the review worked?

Track the review as a system, not a feeling.

• Accounts reviewed vs accounts shipped (cap ratio)
• Meetings booked from that week's shipped list (primary)
• Meetings held (not just booked)
• Notes sent that could only have been written from that week's signal (spot check five)
• Domain health and complaint rate if email is in the mix

Would you rather ship 12 accounts with real windows and book 3 meetings, or ship 200 "fits" and burn the domain for polite replies?

Adapt or fail. Same three signals. One cap. One score. Ship the window, not the vanity.

Start Signals, Convert, Grow

FAQ

How long should a weekly signal review take?

About an hour if prep is done. If it runs two hours, your export is too wide or you are debating logos instead of scoring windows.

Can one founder run this alone?

Yes. Solo means the prep and the review are the same person. Cap harder. Ship fewer. Quality of the note beats volume of the list. See Signals, Convert, Grow for a team of one if that is your reality.

Do I need Vector, RB2B, and G2 all at once?

No. Start with the one signal class you can already trust. Add the second when the first produces meetings. Tools are examples, not a shopping list.

What if G2 traffic is sparse for my category?

Weight de-anon and competitor LinkedIn harder. Sparse G2 is not a reason to invent vanity signals. It is a reason to tighten the score on what you do have. More on G2 category pages as buying signals.

How does this fit Signals vs Convert?

Signals owns the weekly score and the cap. Convert owns the note, the multi-thread, and the meeting path. Grow owns scale after the loop books meetings without theater.