Indexing control is the set of rules that tells search engines which of your boiler pages to index, and the monthly habit of checking whether they listened. On a 40-page site you can skip most of it. On a catalogue of 2,000 model and part pages it is the difference between knowing what works and guessing.
Like everything we do, this is website and search work. We are not a heating contractor and don't install or service equipment.
Why page count changes the job
Small sites have small indexing problems; large ones have structural ones. When a site publishes hundreds of generated pages, three things happen.
- Google indexes some and not others, and the overall number hides the pattern.
- Weak rows dilute the strong ones, because the site as a whole starts to look like it publishes thin pages.
- Duplicate variants appear, such as the same part reachable through two category paths.
Google's spam policies describe scaled content abuse as a pattern across many pages, not a single weak URL. That's why holding weak rows back matters as much as publishing strong ones.
What's included
Sitemaps by page type
We split sitemaps so each page type has its own file: parts, models, fault codes, city hubs, guides. Google's sitemap documentation sets a limit of 50,000 URLs or 50MB uncompressed per sitemap, and notes that lastmod is used only when it is consistently accurate. We go further and split by type at any size, because a per-type file gives a per-type indexing report.
Noindex and publish rules
Rows that fail the pass mark from your page system are either not generated or published with noindex until the data improves. The rule is written down, so nobody switches pages on by hand to hit a target.
Canonicals and duplicates
Where the same item appears under more than one path, one URL is chosen as canonical and the rest point to it. Superseded parts redirect to their replacements rather than lingering as dead pages.
Monthly indexing review
Each month we read Search Console's indexing and performance reports by sitemap and write up what changed: which types gained indexed pages, which lost them, and the reasons Google gives for pages it crawled but didn't index.
Who it's for
- boiler parts and model catalogues past roughly 300 URLs
- installers who launched city or service pages at scale and now see many "crawled, not indexed" entries
- any site going through a scaled-content risk audit, where removals and redirects must be tracked
Timeline
Setup takes two to four weeks: crawl the site, map page types, build the sitemaps, apply publish rules and canonicals. After that it is a monthly cycle. Expect the first useful trend after two or three monthly reports, since Google revisits large sites gradually.
How results are measured
| Measure | What it tells you |
|---|---|
| Indexed share per sitemap | Whether Google values a page type |
| Crawled but not indexed, by type | Where pages are seen but judged weak |
| Pages with impressions in 90 days | How much of the indexed set does any work |
| Redirect and 404 counts | Whether removals are handled cleanly |
A worked example: if your part-page sitemap holds 1,200 URLs and 1,020 are indexed, that type sits at 85%. If your sizing-page sitemap holds 300 URLs and 60 are indexed, that 20% is Google telling you something about those pages, and the next step is a review, not more pages.
Mistakes to avoid
- One giant sitemap. It works technically and hides everything you need to know.
- Touching lastmod on every build. If every date changes nightly, the field stops meaning anything.
- Blocking pages in robots.txt to hide them. Blocked pages can't be crawled, so Google never sees the noindex tag. Use noindex for pages you don't want indexed.
- Deleting without redirects. A removed part page with links pointing to it should redirect to the replacement or the parent category.
- Chasing a total. A rising indexed count means little if the new pages earn no impressions.
For the arithmetic behind keeping or cutting a page type, see the build economics guide. Costs sit within the plans on our pricing page.
Frequently asked questions
Why not just submit every page and let Google decide?
Google will decide either way, but a site that submits thousands of weak URLs makes it harder to see which page types are working. Separate sitemaps and noindex rules turn indexing into a signal you can read.
Does noindex hurt the rest of the site?
No. Holding back pages that fail your data threshold is normal housekeeping. Readers can still reach those pages through internal links if they are useful to browse.
How often should sitemaps update?
Whenever pages are added, removed or meaningfully changed. The lastmod date should only change when the main content really changes, not on every rebuild.
Do you need access to our Search Console?
Yes, as a user you add and can remove at any time. The property stays in your name.