Clearly
Chapter 5 of 6

What to watch in your first year

The two or three metrics per platform that give you real signal in year one, and what a problem looks like before it becomes expensive to fix.

2 min read · Chapter 5 of 6

Each platform surfaces dozens of metrics. Most of them are noise in year one. The goal in your first year is to build a mental model of what normal looks like for your business, so that when something changes, you notice it. Here are the numbers that give you that baseline.

Image placeholder

A week-over-week sparkline dashboard

A simple visual showing 8 weeks of three key metrics per platform arranged in a grid, each with a small sparkline trend and a current week total. Illustrates what "watching the right metrics" looks like in practice versus opening five separate dashboards.

Platform Watch this Problem signal
GA4 Sessions, Engagement rate, Key events Sessions drop more than 20% week-on-week with no obvious cause. Engagement rate below 40%. Zero key events recording.
Search Console Clicks, Average position for top 10 queries Clicks drop without a corresponding drop in impressions (your rank dropped). A page that was ranking on page 1 falls to page 2 or beyond.
Google Ads Cost per conversion, Conversion rate Cost per conversion rising week-on-week without a change in bids. Conversion rate dropping while click volume holds steady (a landing page problem, not an ads problem).
Meta Ads Cost per result, Frequency Frequency above 3 with declining results (audience fatigue: the same people are seeing the same ad too many times). Cost per result increasing with no change to the bid.
Merchant Center Active products, Disapproved products Disapproved products increasing. Active product count dropping without products being intentionally removed (a feed update broke something).

Checking in weekly, not daily

The temptation in the first few months is to check your metrics every day. Daily data is noisy: weekends look different from weekdays, a single large order can skew a day's conversion rate, and one slow Tuesday does not mean your site is broken. A better rhythm is a weekly check (Monday morning, looking at the prior week versus the week before) and a monthly review where you look at trends. Daily checks are useful only when you have just made a change and want to see its early effect.

When to act versus when to wait

Not every dip requires a response. If sessions fall 10% for one week and recover the next, it is probably noise. If sessions fall 20% for three weeks in a row, something has changed and it is worth investigating. The same principle applies to paid ads: let a new campaign run for at least two weeks before drawing conclusions, because Google's smart bidding is still learning in the first week.

The hardest skill in year one is not reading the dashboards. It is knowing which changes require action and which ones you should wait out.

A simple weekly check routine

Template placeholder

Downloadable: weekly metrics log template

A simple one-page spreadsheet template with columns for date, platform, metric name, this week's value, last week's value, percentage change, and an observation note. Designed for a 10-minute weekly review. Available as a Google Sheets template.

You now have the foundation: the right platforms, the right connections, a setup order, and the metrics to watch. In the final chapter, we look at the mistakes that come up most often when small businesses first get serious about their data.

PreviousThe right setup order NextCommon mistakes