The only 5 metrics to check every morning

The morning metrics check fails in one of two ways: a dashboard with forty tiles that gets skimmed and retains nothing, or no ritual at all, with problems surfacing whenever someone happens to look. The fix for both is the same — a fixed list, short enough to finish with the first coffee, where every number has a baseline attached and a same-day action if it moves. Five covers it. The catch: the right five depend on what you run. So here they are as questions first, then the exact metrics for a product, a store, a sales team and an audience.

One rule before the list: every number is a comparison, never an absolute. 'Yesterday: 214 visitors' means nothing alone; '214 against a Tuesday average of 180' is a reading. Most sites breathe in a weekly rhythm, so compare against the same weekday — putting Saturday next to Wednesday manufactures false alarms.

Every version of the morning check answers the same five questions, in the same order:

• Reach — did anyone come yesterday? Read against the weekday baseline. A spike is something to do more of while it's warm; a dip that holds for two or three days means something upstream broke, and the source table is where you find out what.
• Conversion — did anyone act? The commitment number, read next to visitors: traffic up with conversions flat is an audience-quality question, while traffic flat with conversions down points at the page or the form — worth ruling out a technical break first.
• Follow-through — did yesterday's converts take the next step? Conversion is a promise; this checks it got kept. It catches breakage a day before the conversion count shows anything.
• The stakes — did the thing you can't afford to lose move? Most days this is a glance confirming nothing changed. It's on the list for the days it did, because that news ages badly over a weekend.
• The bet — is this month's experiment working? The only slot that rotates: whatever you're testing has one number that says so, and it lives here until the bet settles.

The classic version, and the one the rest of this library is built around:

• Reach — are enough people coming? Visitors against the weekday baseline, with returning visitors as the habit signal.
• Conversion — are visitors becoming users? Signups ÷ visitors, and signups by source. The biggest traffic source is rarely the best-converting one.
• Follow-through — are new users reaching value? Activation rate: the share of yesterday's signups who did the one action that predicts they'll stay. A broken onboarding shows up here while the signup count still looks healthy.
• The stakes — is the subscription base healthy? MRR movements: new, churned and failed payments read together. Total MRR can look flat while churn eats growth underneath.
• The bet — is the change working? One number per experiment: a redesigned signup flow → the form's conversion rate; a new pricing page → trials started.

Same questions, retail numbers (the shape works for physical retail too — swap visitors for footfall and read against the season as well as the weekday):

• Reach — are enough people coming in? Store visitors against the same weekday, remembering retail breathes with promotions and holidays.
• Conversion — are visitors becoming buyers? Orders ÷ visitors, average order value, and sales by source. The channel that sends the most traffic is rarely the one that sends the most revenue.
• Follow-through — where do buyers drop off? Checkout completion, cart and checkout abandonment, and payment failures.
• The stakes — is today's revenue actually healthy? Sales next to refunds, returns and disputes, plus margin when discounts are running. A good revenue day can hide a bad margin day.
• The bet — is the change working? One metric per experiment: free-shipping threshold → average order value; a new product page → its conversion rate; a new homepage layout → sales per visitor.

Sales-led businesses run the same check with the pipeline in the product's place:

• Reach — is the top of the pipeline filling? Visitors and new qualified leads, against the weekday baseline.
• Conversion — are visitors raising their hands? Demo requests or leads booked, with source.
• Follow-through — did booked turn into held? Meetings that actually happened, plus replies to yesterday's outreach. Booked-but-ghosted is the sales version of signed-up-but-never-activated.
• The stakes — is the pipeline moving? Deals advanced and deals gone quiet. A big deal going silent shouldn't wait for the Friday review.
• The bet — is the experiment working? A new outbound sequence → reply rate; a repriced tier → close rate.

For a blog, newsletter or community that isn't monetizing yet, membership takes the money slots:

• Reach — are people reading? Readers yesterday, against the weekday baseline.
• Conversion — are readers joining? New subscribers or members, with the post or channel that brought them.
• Follow-through — do new members come back? Welcome-email opens, or day-two returns.
• The stakes — is the audience growing or leaking? Net member change: joins minus unsubscribes. The audience is the asset until money arrives, so this is the row that can ruin a week.
• The bet — is the change working? A posting schedule → subscribers per post; a new format → replies or completion rate.

Ad spend isn't a sixth question — it's a price tag on the first two, whichever model above is yours. Once campaigns run, read reach next to yesterday's spend, and conversion next to cost per conversion, campaign by campaign. This is where the morning check earns hard cash: a broken tracking pixel, a fatigued creative or a runaway budget burns money silently, and all three are invisible in the visitor count alone. A campaign whose cost per order doubled overnight is stakes-grade news. Which channel deserves the budget is a different, slower question — answered by revenue per source, not by the traffic each one sends.

LTV, CAC, NRR, churn rate: real metrics, wrong cadence. They move monthly, and a daily reading of a monthly number is noise that trains you to ignore the whole list. One distinction worth keeping straight: blended CAC is a monthly number, but cost per conversion on a live campaign is a daily one — which is why it sits in the section above, not here. Pageview totals and follower counts fail differently — they can climb for weeks while nothing underneath improves. Competitor dashboards and rankings are morning reading, not morning metrics: nothing in them changes what you do before lunch.

The slow numbers belong in a weekly review, with time to actually think about them. The morning check has a narrower job: catch yesterday's breakage today.

Same five questions, same order, every morning: reach against baseline, conversion with source, follow-through, the stakes, and the bet. Three minutes when nothing happened — most mornings — and an immediate thread to pull when something did. Better still, let the check come to you: alerts on the thresholds that matter (revenue drops, failed payments, traffic collapse) turn the ritual into a safety net that works even on the mornings you skip it.