Quick answer: SEO query analysis means tagging each Search Console query as research, comparison, or ready to book, then checking which landing page and which result (call, form, or nothing) it produced. One blended conversion rate hides the only number that matters.
Rankings are a weak report. A query that brings a same-day call and a query that brings a student writing a paper should not share a row. New Jersey makes this obvious: a shore town in July and a B2B supplier in Newark do not share a calendar.
Analytics posts love attribution models with six touches. A local business usually needs two numbers: calls from the profile, and forms from the site, split by intent. This is that report.

Tag the query before you celebrate the rank
| Tag | Words you will see | Fair result |
|---|---|---|
| Research | what is, how long, cost of | A pageview, maybe a return visit |
| Comparison | best, vs, reviews | A longer visit, a profile view |
| Ready | near me, open now, emergency, book | A call or a form the same day |
Export 16 months from Search Console. Tag in a sheet. Then look at the landing page. A ready query hitting a 2,000-word essay, or a research query hitting the contact form, is a mismatch you can fix without a new article.
Do not average the emergency with the essay
An emergency query should show up as a call today. A research query may take weeks. Averaging them into one conversion rate makes the local pages look like a failure and the blog look like a success, or the reverse.
Track calls and forms by landing page. Keep the intent tag on the query. That is enough to see whether the town pages or the guides are paying for themselves.
One account measured on this site returned about 10x. That is one account, not a forecast. The homepage also shows a 367% increase in calls from forms and the phone, and a 220% increase from the profile plus the site. Those cards do not name the client. Treat them as the shape of the report, not as a promise.
Season before panic
Pull two years before you call a dip a problem. HVAC, landscaping, tax, schools, and shore rentals in New Jersey move on different months. Publish the seasonal page before the spike. By the time the spike is obvious, the week is half gone.
Lakeside's map position moved from 12.89 to 1.29, and Metro Storage's average position moved from 13.4 to 1.6 with a 700% rise in profile interactions. Those are position and interaction numbers. They are not revenue. Put revenue next to them in your own sheet or do not claim it.
Local calls that never hit the site
People search, read a review, tap the map, and call. If you only count form fills, the profile looks useless. Use a tracking number on the profile and a different one on the site if you can.
Advanced Integrated Health's six-month result was leads per month, from zero to more than 100, not a rank screenshot. Leads are the column. Position is context.
What to add, and what to leave
A query that was nothing in March and keeps growing in April deserves a section on an existing page. It does not deserve a new site. Add a page when the query has impressions and the current URL answers a different question.
Monthly checklist
- Queries tagged research, comparison, or ready
- Landing page checked for the top 50 queries
- Calls and forms split by page
- Profile calls counted separately from site calls
- Two-year view checked before a "drop" email
- New queries either added to a page or ignored on purpose
Five columns, one sheet
Make a sheet. One row per query that has impressions. Five columns:
| Column | What you type |
|---|---|
| Query | The exact Search Console string |
| Intent | Research, comparison, or ready |
| Landing page | The URL that got the click |
| Right page? | Yes, or the URL that should have it |
| Result | Call, form, or none, from that page that month |
This is a method, not a client report. A row might read: query "emergency plumber near me," intent ready, landing page a blog post from 2022, right page no, result none. The fix is a link and a title change, not another article.
Sort by impressions and work the top 30. The long tail can wait. Review the sheet once a month, and once before a seasonal spike you already know about, such as shore traffic in summer or tax season for an accountant.
The figures on the home page (calls and form leads, not a blended rate) are the kind of result this sheet is trying to explain.
FAQ
Is average position a KPI?
It is context. Metro Storage's move from 13.4 to 1.6 mattered because interactions rose 700% with it. Position without a call or a form is a chart.
How often should I pull Search Console?
Monthly for the tagged sheet. Weekly only during a launch or a bug. Daily pulls create fake emergencies.
What if I cannot track calls?
You can still tag queries and landing pages. You cannot claim revenue. Say so in the report.
Which page should a ready query hit?
The service or town page with the phone number in the first screen, not the guide. Guides are for research queries. See voice and near-me setup for the page itself.
Export the queries and the call log if you have one. We will tag a month of it on a call and show which URLs are earning their keep. The review side of the same numbers is Google Business Profile reviews.