Core metrics sheet

Five numbers a studio can read on Monday. Impressions are diagnostics. Conversations are the scoreboard.

One sheet, one tab, last 12 weeks. Update every Monday. If a number takes more than five minutes to find, the store is wrong, not the metric.

Do not add vanity columns because a tool offers them. Traffic without conversations is a hobby.

flowchart LR
    I["Inbound"] --> Q["Qualified"]
    Q --> P["Proposal"]
    P --> W["Won"]
    I --> T["Time to reply"]

Definitions (write these on the sheet header)

Metric	Counts as	Does not count
Qualified conversation	They match the who-line and can buy	Newsletter signups, spam, “just researching”
Proposal sent	A dated PDF or scoped email they can accept	A vibe call with no artifact
Close rate	Won / proposals sent, last 90 days	Won / all conversations (that hides a leaky qualify step)
Site contact completion	Form row that is a real person	Honeypot hits, empty tests, duplicate refreshes
Time to first reply	Hours from inbound timestamp to first human send	Auto-ack “we got it”

These five definitions pasted at the top of the sheet so next quarter’s you does not renegotiate them

Week-starting date, then the five counts, then source notes. That is enough.

Source column uses search / direct / referral / outbound. If you cannot name it, write “unknown” and fix tracking that week
Customer acquisition cost (CAC) and lifetime value (LTV) stay off this sheet until the counts above exist for a full quarter

Fill last week’s row from the intake sheet, not from memory
If median reply time missed the service level agreement (SLA), name the cause in Notes
One fix for the worst number. Not five new channels