Programmatic SEO means generating a set of pages from a template and a data source. It works when every page carries genuinely different, useful information. It fails, expensively, when the pages differ by one word. Most of my job here is working out which of those a proposal would actually be, before anything gets built.
If two pages differ only by a place name, they are not two pages, and search engines are good at spotting exactly that.
What programmatic SEO is
It’s building a page set from structured data: one page per location, per product type, per combination of the two. The template stays the same and the content changes, so a hundred pages can be produced from one build and one data source.
Done properly the reader can’t tell it was templated, because each page answers their specific question with specific information. Done badly it’s the same page with a place name swapped, which Google’s spam policies describe as scaled content abuse, and those policies apply however the content was produced.
Who it’s for
This suits businesses with genuine structured variation:
- service businesses covering many locations, with real differences between them
- shops with large catalogues where category combinations have real demand
- sites with a proprietary dataset worth publishing
- directories, listings and comparison sets built on real data
It doesn’t suit a business that wants more pages. That’s the request this service most often turns down.
The tests a page set has to pass
| Test | What it means |
|---|---|
| Real demand | People search this pattern, with volume per variant rather than for the pattern overall |
| Real data | Every page has genuinely different information behind it |
| A real answer | Someone landing on one page finds what they came for without the others |
| A sensible size | The set stops where the data or the demand stops |
| Somewhere to live | A hub, navigation and internal links exist or get built alongside |
| Maintainable | The data source stays current, or every page goes stale at once |
The second test is the one that kills most proposals, and it should. If two pages differ only by a place name, they aren’t two pages.
The problems it usually solves
A catalogue with more demand than pages. Where real searches exist for combinations your site doesn’t cover, and the data to answer them already exists in your product information.
A dataset nobody can see. Businesses often hold genuinely useful data in a spreadsheet or a system, and publishing it properly is the strongest form of this work.
Manual pages that can’t keep up. Where a team is writing near-identical pages by hand, a template with real data behind it is faster and more consistent.
What’s included
- A demand assessment per variant, not for the pattern as a whole
- A data audit, checking whether genuinely different content exists for each proposed page
- A recommended set size, which is almost always smaller than the number of possible combinations
- The template design, including what changes per page and what doesn’t
- Hub and internal linking structure, so the pages are reachable
- Indexing rules and a measurement plan from the start
- A review and pruning routine, because some pages will fail and should be removed
Where to stop
Sets grow because combinations multiply. Three product types across forty locations is a hundred and twenty pages, and typically ninety of them have neither demand nor distinct data.
I’d rather build thirty strong pages than three hundred thin ones. The thin version doesn’t just fail on its own terms, it can drag on the rest of the site.
Sample output
PROPOSAL: Service pages for 40 UK towns. DEMAND: Real for 11 towns. Negligible for the rest. DATA: Genuinely different content available for 11: local pricing, coverage and named work. The other 29 would differ by name only. VERDICT: Build 11. Do not build 29. STRUCTURE: Needs a coverage hub and menu links, which don’t exist yet.
Limitations
Data quality decides everything. A set built on an inconsistent source produces pages that are confidently wrong at scale, which is worse than not building them.
These sets also need maintaining. A page set nobody updates ages badly, and it ages all at once.
What this won’t solve
Programmatic pages won’t create demand that isn’t there, and they won’t fix a site with underlying technical or authority problems. Publishing a large set onto a weak site is the scenario where this work does actual harm.
Talk to the person doing the work
There is no account manager here and no team to be handed to. If you get in touch, it is me who reads it, me who looks at your site, and me who does the work if we go ahead.
Get in touch, or start with a free assessment.
Engagement and price
There are no list prices here, and that is deliberate. What a piece of work costs depends on the state of the site, how much of it there is and what you actually need, none of which I know before looking at it.
So it starts with a free analysis. I look at your site, your Search Console and who you are really competing with, then come back with what I would do first and what it would cost to do. No obligation attached, and if the honest answer is that you do not need me yet, I will say so.
How the work gets done
Proposals run through the programmatic planner, with demand data from the crawl and data integrations and structure checked through the technical SEO checks. This service sits under SEO, and the build side is web development.
FAQ
Is programmatic SEO allowed by Google?
In itself, yes. Google’s spam policies target scaled content created primarily to manipulate rankings rather than to help people, and they apply however it’s produced. A set where every page carries genuinely different, useful information isn’t what those policies describe.
How many pages should I build?
As many as have both real search demand and genuinely different data behind them, which is nearly always far fewer than the possible combinations. Thirty strong pages are worth more than three hundred thin ones, and they don’t drag on the rest of your site.
What makes a programmatic page thin?
Having nothing on it that couldn’t be produced by swapping one word in a template. If two pages in a set differ only by a place name or product name, they aren’t two pages, and search engines are good at spotting exactly that pattern.
What happens to pages that don’t perform?
They get measured individually and then merged or removed. Deciding that in advance, along with the indexing rules, is part of the plan. Publishing a large set and hoping is how sites end up with thousands of pages earning nothing.
Find out whether a page set is worth building
Get in touch, or start with a free assessment, and I’ll tell you honestly if the answer is no.