The tip: pull your Search Console page-state counts before you buy anything with "AI" in the name
If someone is pitching you an AI-powered content system for your firm, here is the one-question audit that tells you whether you're ready for it: ask your team to open Search Console and read out four numbers — indexed, discovered, crawled, and excluded.
Not traffic. Not rankings. Not a keyword report. Page states. How many of your pages Google has actually put in the index, how many it has found but not visited, how many it has visited but not indexed, and how many it has looked at and deliberately set aside.
That single readout tells you more about whether automation will help or hurt you than any vendor demo. Because AI readiness is less about buying an AI label and more about clean data, clear rules, and human oversight. If your workflow cannot explain its source data, its owner, and its exception path, AI will automate confusion rather than intelligence.
Why those four numbers matter more than they look
Most firm owners have never seen this report, because agencies tend to present the flattering slice: sessions up, keywords up, a chart moving to the right. Page-state counts are less flattering and far more useful, because they describe the mechanical relationship between your site and the search engine.
Discovered but not crawled
Google knows the URL exists and has not bothered to fetch it. At small volume this is noise. At the scale of a real local coverage strategy, it is a signal that your internal linking and site architecture aren't giving crawlers a reason or a route to those pages.
Crawled but not indexed
Google fetched the page, evaluated it, and declined. This is the number that exposes thin, templated, near-duplicate output. It is the failure mode every serious operator worries about when they hear the word "programmatic."
Excluded
A catch-all bucket that quietly hides canonical conflicts, redirect chains, parameter sprawl, and pages you didn't know you were publishing. Firms are often surprised to find that a meaningful share of their site is in here.
Indexed
The only number that can earn anything. Everything else is inventory sitting in a warehouse with no door.
Read together, these four numbers are a diagnostic of your data hygiene and your publishing rules. That's exactly what any AI layer will inherit.
If the workflow cannot explain its source data, its owner, and its exception path, AI will automate confusion rather than intelligence.
What "AI-ready" actually means in a programmatic context
Bosseo describes Programmatic SEO as a system designed to publish a genuinely useful page for every service-city combination in the market you actually serve. That definition does a lot of work. Read it slowly:
- Genuinely useful — the page has to justify its own existence to a human reader, not just to a crawler.
- Every service-city combination — coverage is the point. Not one flagship page, but the full matrix.
- In the market you actually serve — you don't manufacture pages for cities where you can't take the case.
Automation is what makes that matrix physically possible. A firm with a dozen practice areas across a few hundred cities is looking at page inventory it will never build by hand — not in a year, not with a full-time writer, not with an agency retainer aimed at four blog posts a month. The scale Bosseo works at is roughly 10,000 service-and-city pages on the firm's own domain, with schema, internal linking, and AI-search optimisation built in.
But scale amplifies whatever your process already is. If your underlying data is clean — real service definitions, real city lists, real jurisdictional facts, real intake routing — automation multiplies something valuable. If your data is a spreadsheet somebody inherited, with no owner and no rule for what happens when a row is wrong, automation multiplies that instead. Ten thousand times.
The three questions that make a workflow AI-ready
1. What is the source data, and where does it live?
Every page in a programmatic system is assembled from inputs: practice area, city, jurisdiction detail, relevant local facts, firm-specific proof, calls to action. Somebody has to be able to point at the canonical source for each of those. "It's in the template" is not an answer. If the source of truth is ambiguous, every downstream page inherits the ambiguity — and you will find out about it in the crawled-but-not-indexed bucket.
2. Who owns it?
One named human. Not "marketing." Not "the agency." Someone whose job it is to approve the city list, retire pages for markets you exited, and sign off on legal language before it publishes across an entire state's worth of URLs. Human oversight isn't a compliance box — it's the thing that keeps a template from confidently repeating an error at volume.
3. What is the exception path?
This is the question that separates real systems from demos. What happens when a city genuinely doesn't fit the template? When a practice area has different rules in one jurisdiction? When a page needs to be pulled? A mature workflow has a defined route for exceptions. An immature one either ignores them or lets one person quietly hand-edit pages until nobody knows which version is authoritative.
The honest framing: this is infrastructure, not a magic ranking trick
Programmatic SEO at this scale is geographic coverage infrastructure. It isn't "better SEO," and it isn't a replacement for the agency already handling your brand, your money pages, and your paid campaigns. It is the page inventory nobody is going to build by hand — the reason a firm ranking well in one city can start showing up in the forty around it.
The firms that get the most out of it are usually established, already ranking somewhere, and want to own more cities and more practice areas than their current content velocity allows. That profile matters, because coverage infrastructure compounds on top of existing authority. It does not manufacture authority from nothing.
Investment typically runs $1,750–$5,997/mo depending on state and city count, often with a build fee. That's a real line item, and it deserves the same diligence you'd apply to a new hire — which brings us back to the audit.
Your next step
Do this before your next marketing meeting:
- Ask for the four numbers. Indexed, discovered, crawled, excluded — straight from Search Console, not a dashboard summary.
- Ask who owns the city list. A name, not a department.
- Ask what happens when a page is wrong. If the answer is a shrug, you have found your gap.
If your team can answer all three cleanly, you're ready to scale — and automation will multiply real intelligence. If they can't, fix the workflow first. That fix is cheaper than ten thousand pages of automated confusion.
See how Bosseo approaches this at bosseo.com/programmatic-seo.
Next step
See how Bosseo closes this gap
Book a short call and we’ll show you exactly where the leak is.