Chapter 12 · Stage 04
Measuring ABM SEO by Segment
Search Console knows pages and your CRM knows leads, but neither knows segments. Join them with one sheet and report leads and SQLs per segment every month.
- Search Console knows pages, the CRM knows leads. The join between them is one sheet that gives every page a segment.
- Four fields on every lead record carry the segment from the first visit to the SQL.
- Check attribution before you trust a number: at one client, 10 to 15 organic leads a month were counted as direct.
- Read the funnel per stage. A falling SQL rate can sit entirely after sales accepts the lead.
The forecast gives you leads per segment before you publish. This chapter is how you get the actuals to compare it with. Without them, the quarterly review has nothing to show and the forecast that won the budget quietly disappears.
The problem is a missing join. Search Console reports pages and queries. The CRM reports leads and deals. Neither knows which segment a page or a lead belongs to.
The join: a page-to-segment sheet
Section titled “The join: a page-to-segment sheet”Every page in a cluster belongs to one segment. Every lead has a first page it landed on. Put those two facts in one sheet and you can report anything per segment. An illustration, with made-up paths:
| Path | Segment | Stage | Live since |
|---|---|---|---|
| /construction/invoice-financing/ | Construction | Decision | 2026-03 |
| /construction/borrowing-without-annual-accounts/ | Construction | Consideration | 2026-05 |
| /compare/us-vs-competitor-a/ | All segments | Decision | 2025-10 |
Keep the path without the domain, the query string or the trailing variations, so it matches what the CRM stores. Pages that serve every segment, such as comparison pages, get “All segments” and their own line in the report. Don’t force them into one.
Done when every URL in every cluster has a row. This sheet is the whole infrastructure. The rest reads from it.
Search Console: one filter per segment
Section titled “Search Console: one filter per segment”In the Performance report, add a page filter, choose Custom (regex), and match the segment’s paths. If the segment name is in the URL, the filter is short. An illustration:
/construction/|contractorIf it isn’t, paste the paths from the sheet as one pattern separated by |. Save the comparison or pull it through the API every month, so the numbers come from the same filter each time.
Split brand from non-brand queries before you read a segment. Brand searches land on the homepage and a few product pages, and they hide whether the segment pages are earning anything on their own.
The CRM: four fields on every lead
Section titled “The CRM: four fields on every lead”- First page seen. The path of the first page the visitor landed on, matching the sheet. HubSpot fills this by default.
- Original source. Organic search, paid, direct, referral. Also a default in HubSpot.
- Segment. A custom field, filled from the sheet by a workflow or at import. Not chosen by a sales rep from a dropdown.
- Stage dates. The dates the record became a lead, a sales-accepted lead and an SQL. Without dates you can’t tell which month’s content produced which month’s SQLs.
At the business lender I work with, the leads export from the CRM goes into a sheet with one row per month and landing page: month, path, organic leads. That shape is enough. It rolls up to segments through the page sheet, and it is the same shape the forecast uses.
Check attribution before you trust a number
Section titled “Check attribution before you trust a number”Two things went wrong at that client, and both would have made the segment numbers look worse than they were. The figures in this chapter come from that engagement, anonymised, and illustrate the method; they are not a benchmark for your site.
Organic leads counted as direct. I estimated that 10 to 15 organic leads a month were recorded as direct. Check it by comparing the first page seen with the source: a lead whose first page is a cluster article and whose source is direct is almost always organic.
An empty value field. The pipeline value field in the CRM was empty, at that client and at a second one. Until it is filled, report leads and SQLs per segment, not euros. A revenue number built on an empty field is worse than no number, because people trust it.
The monthly readout
Section titled “The monthly readout”One table, one row per segment:
| Segment | Clicks | Leads | SQLs | SQL rate | Forecast leads | Gap |
|---|
Then ask the question the quarterly review asks: is the gap in traffic or in conversion? Traffic gaps point to rankings, links and authority. Conversion gaps point to the page, the offer and the intent.
Read the funnel one stage at a time before you blame the content. At the lender, the organic SQL rate fell from 23.7% in the last quarter of 2025 to 15.0% in the third quarter of 2026. Over the same period, lead to sales-accepted held at 45 to 47%, and sales-accepted to SQL fell from 51% to 36%. The leads were as good as before; the drop sat after sales had accepted them. The first step is a conversation with sales about lost reasons and credit criteria, not a content rewrite.
Before you rewrite a page for a falling SQL rate, find the stage where the rate fell.
The segment view also shows which pages to build next. At the same client, a 37-page sector set brought 6 leads and no SQLs from June to mid-September, while one comparison page brought 16 leads and 6 SQLs from June to late October. Without the page sheet, both would have been “organic leads”.
Small numbers
Section titled “Small numbers”Most segments produce single-digit leads a month for the first quarters. Treat 0 to 3 conversions as a direction, not a result. A page with real traffic and a consistent gap is a signal; one lead more or less is noise. Report three-month rolling totals per segment next to the monthly figure.
What this doesn’t do
Section titled “What this doesn’t do”It doesn’t measure accounts that never fill in a form. Visits from target accounts who read and leave are invisible here. Account-identification tools show some of them; this setup only counts what reaches the CRM.
It doesn’t fix the CRM. If sales doesn’t record stage dates or lost reasons, the SQL rate can’t be read per stage. That is a sales process fix, and it needs the sales lead in the room.
It doesn’t replace the forecast. Actuals without a forecast tell you what happened, not whether it was enough. Build the forecast spreadsheet next to it.
Where to start
Section titled “Where to start”Export every URL in your clusters and give each one a segment in a sheet. Then filter your CRM’s last three months of leads by first page seen and count them per segment. If that count is far from what your team believes, start with attribution.
Take this away: 3 asks for your CRM owner, your developer and your AI
Ask your CRM owner
You need: the four fields on every lead.
Ask your developer
You need: the first page stored on every form submission (only if your CRM doesn't do it already).
Your AI
Your AI can find the misfiled leads in a minute.
See the prompt
Get notified when new chapters publish
You're on the SEO lead path. Step 5 of 5.
0 of 5 read
- Next: Content Clusters
- Not read: Internal Linking
- Not read: The Content Quality Gate
- Not read: The Forecast Spreadsheet
- Reading now: Measuring by Segment
Switch path: CMO or Head of Growth