Reading Your SEO Report: What the Numbers Actually Mean

TL;DR: Most SEO reports lead with the metrics that always look good and bury the ones that decide whether you’re winning. Four numbers matter, three are decoration, and five questions turn a report from a document you skim into a conversation that changes the plan. If you’re evaluating reporting tools rather than reading a report, see our comparison of real-time versus periodic reporting.

Table of Contents

• The four numbers that matter
• The three that are mostly decoration
• Impressions up, clicks flat: the most common pattern
• Rankings without context
• Five questions to ask about any report
• What a good report looks like

The Four Numbers That Matter

1. Qualified leads by channel. The only number that connects to revenue. Not contacts — qualified: in your service area, wanting something you sell, with a real timeline. If your report doesn’t have this, it’s measuring the internet rather than your business.

2. Clicks from organic search. Actual visits earned. Reported in Search Console, hard to spin, and directly comparable month over month and year over year.

3. Average position on your named target keywords. Not “keywords in the top 100.” The specific set you agreed to pursue. Movement from 18 to 9 on a term that matters is worth more than fifty new rankings at position 60.

4. Cost per booked job by channel. Requires your CRM data, which is why most reports omit it. It’s also the number that should drive every budget decision you make.

Everything else supports these four or distracts from them.

The Three That Are Mostly Decoration

Impressions. Useful as a diagnostic, misleading as a headline. Impressions rise when you rank at position 40 for more things. That’s not progress you can bank. A report leading with impression growth is often a report avoiding click data.

Total keywords ranking. Grows almost automatically as you publish. A site can add three hundred ranking keywords in a quarter and earn nothing, because all three hundred sit on page four.

Domain authority or similar third-party scores. Vendor-invented metrics, not Google signals. They correlate loosely with performance and are easy to move without improving anything. Fine as a rough sanity check, meaningless as a goal.

None of these are lies. They’re just answers to questions you didn’t ask.

Impressions Up, Clicks Flat: The Most Common Pattern

This deserves its own section because it’s the single most frequent shape in local SEO reporting, and it gets interpreted wrongly in both directions.

The pattern: impressions climbing steadily, clicks close to flat.

What it usually means. You’re ranking for more queries, but at positions where nobody clicks. Positions 11 through 30 generate impressions and almost no traffic. You’re visible and not selected.

What it can also mean in 2026. AI-generated answers are satisfying the query above your result. You get counted as an impression, the searcher gets their answer, nobody clicks. This is genuinely happening and it’s worth distinguishing from the position problem — the fix is different.

How to tell which. Look at average position for the affected queries. If you’re at 15 to 25, it’s a position problem: the answer is better content and stronger internal links, not a different strategy. If you’re at 3 to 8 with impressions up and clicks down, that’s answer-capture, and the response is different — earn the citation, restructure for direct answers, and accept that some informational queries no longer send traffic.

Our own site shows the pattern clearly: a group of queries pulling roughly 13,000 impressions in a 28-day window at positions 16 to 26, generating essentially no clicks. In our case the cause was three of our own pages competing for the same query family — a consolidation problem, not a content shortage. Which is a good illustration of why the diagnosis matters before the prescription.

Rankings Without Context

Ranking reports are the easiest section to make look good, using three techniques worth recognizing.

Selective keyword sets. Reporting on 15 keywords when 200 are tracked. Ask for the full tracked list, every month, including the ones that fell.

Keywords nobody searches. Position one for “emergency plumber in [tiny neighborhood] same day affordable” is trivial to achieve and worth nothing. Ask for search volume alongside position.

Position without a click column. Position matters only insofar as it produces clicks. A ranking table with no click data next to it is a table you can’t evaluate.

The fix for all three is the same request: rankings and clicks and volume in one table, for the full tracked set.

Five Questions to Ask About Any Report

Ask these on the monthly call. They’re more useful than any additional metric.

1. “What changed this month, and what did we do about it?” You’re testing whether anyone read the report before sending it. A report nobody read is a PDF, not a service.

2. “Which target keyword moved the most, and why?” You want a mechanism — a page rewritten, a link earned, a consolidation. “Google updates” is not a mechanism.

3. “What isn’t working?” Every month has something. A report with no negatives is being curated for you rather than written for you.

4. “What are we doing next month, and how is that different from this month?” If the answer is identical activity at identical volume, nobody is deciding anything.

5. “How many qualified leads did this produce?” If the answer is “we don’t have that data,” offer CRM access and ask for it next month. If the answer stays no, that’s informative.

What a Good Report Looks Like

Short. One page of numbers and one page of narrative beats twenty pages of charts.

The numbers: clicks and impressions trend, target keywords with position and volume and clicks, qualified leads by channel, work completed this month, work planned next month.

The narrative: what changed, why we think it changed, what we’re doing about it, and what we got wrong last month.

That last item is the tell. A provider willing to write down what they got wrong is a provider actually running experiments. A report that has never contained a mistake in twelve months is marketing collateral.

Quick Recap

• Four numbers matter: qualified leads by channel, organic clicks, position on named targets, cost per booked job.
• Impressions, total ranking keywords, and third-party authority scores are diagnostics, not results.
• Impressions up with clicks flat usually means positions 11–30 — check average position before changing strategy.
• Demand rankings with volume and clicks in the same table, for the full tracked set.
• Ask what changed, what moved and why, what isn’t working, what’s different next month, and how many qualified leads resulted.

Frequently Asked Questions

How often should I get a report?
Monthly is right for most local businesses. Weekly reporting on organic search mostly measures noise, and quarterly is too slow to correct course.

Should I be looking at Google Analytics myself?
Worth having access and checking occasionally, but don’t try to run the analysis yourself. Search Console is more useful for a business owner — clicks, queries, and positions are readable without training.

My report is 30 pages. Is that good?
Usually the opposite. Length often substitutes for judgment. One page of numbers plus a page of reasoning is a better sign.

What if my rankings improved but leads didn’t?
Check three things: whether the improved rankings are on commercial or informational queries, whether the pages ranking have working calls to action, and whether your intake changed. We find broken forms more often than we find broken campaigns.

Is it reasonable to ask for the raw Search Console data?
Yes, and it’s your data. You should have owner access to your own Search Console property regardless of who does the work.