The tools got cheap. The judgment did not.

Patrick Stox, Ben Wills and Ben Senescu joined Jeremy Rivera for the first panel episode of the Unscripted SEO Podcast. Between them they cover two decades inside Ahrefs and IBM, a quarter century of SEO plus a detour through embedded engineering, and a tool built from scratch in seven months by one person. Nothing was pre-briefed.

Read as a builders’ conversation, it is a good listen. Read as a buyer of SEO software or a manager of an SEO team, it is a warning about where your leverage actually sits now.

The build constraint moved, and the bottleneck moved with it

Patrick Stox described himself as the one non-engineer on the call, said he could not have built most of his tooling before, and now has roughly a hundred and four free tools live on his own domain. Ben Senescu, running OpenSEO alone, said a feature that works ninety-five percent of the time is now easy. Ben Wills, who writes C for a living, had to run a full database refactor because the models made architectural decisions he had explicitly told them not to make.

“AI can produce stuff, but it doesn’t have the taste to decide, do we use a B-tree, or do we use binary search on a sorted list. That curation is your taste. That is the wisdom you’re embedding into the application.”Ben Wills, OppAlerts

The word he used was taste, and it is the same argument this practice makes about SEO programmes. The production step is no longer the expensive one. Deciding what should exist, and what should be pulled back out, is the expensive one — and it is the step that does not get cheaper when the tooling does.

Patrick made the commercial version of it from the other side of the table:

“If Ahrefs had listened to customers, they would be a low-cost link tool, something like Majestic still.”Patrick Stox, ex-Ahrefs

Features shipped there that were genuinely good, did not get used enough to justify their own maintenance, and were killed. The test was never quality. It was whether the thing would earn its existence.

What this changes about your stack

Three practical consequences for a team choosing tools this year.

• Cheap capable tools change who can do the work, not what good work is. Senescu’s customer base turned out to be entrepreneurs doing SEO for the first time, reachable at a ten dollar price point rather than a hundred and ten. That widens the field of people who can run a process. It does not widen the field of people who can judge one.
• A tool that does everything is a tool nobody has curated. Senescu’s constant question is what the small next thing should be, against a queue of community pull requests that would let him claim the tool does everything. Apply the same scepticism to a vendor roadmap that keeps growing.
• Budget for judgment, not for seats. If output has collapsed in price and curation has not, the line item that protects your results is the one that decides what to stop doing.

Two surfaces, two timelines

The most useful exchange came when Senescu asked the other two how to get OpenSEO recommended by ChatGPT and Claude. Patrick split the problem in a way that is directly applicable to any brand that is not yet a household name in its category.

“You’ve got a while before you’re really in the training data as a known entity. For now, it’s the retrieved pages where you’ll want to be found.”Patrick Stox

Being in the model is slow and largely passive. Being on the pages a model fetches at answer time is work you can scope this quarter: identify the roundups and comparison pages that get retrieved for the question your buyer actually asks, and make sure your one genuine differentiator is stated on them. Ben Wills added the outreach shape — category lists, curated directories, publishers who already maintain the list you want to be on.

For teams used to reporting on rankings, note what this reframes. There is no single position to hold. Ben Wills’s own testing found that changing one word in a prompt — naming a Corolla instead of a BMW when asking about a trip — changes which businesses come back, and that ChatGPT’s memory can carry that context into an unrelated conversation days later. Any visibility number you are shown is one person’s result, in one context, on one day. Treat it accordingly.

The risk that should reach your risk register

Patrick’s closing warning is the one worth taking to a leadership conversation.

“Any tactic that is abused will be shut down… When Penguin came along it hurt for a few years, and then Google just said, all those bad links you built, we just forgive you. That may not be the case in all these new systems. They might hold a grudge for the next 10, 20, 30 years.”Patrick Stox

The historical pattern was: abuse a tactic, absorb a penalty, get forgiven. He is not confident that forgiveness is on offer this time. That turns a short-term tactical decision into a long-horizon liability, which is exactly the sort of trade an in-house team should not be making without an adult in the room.

Senescu named the two offers currently landing in inboxes — a couple of thousand a month to spam every relevant forum, or a hundred a month for an automated link exchange. His read is that they may work briefly and are a distraction from learning to market the business. Ours is blunter: they are a decade-long bet placed by someone who will not be around to settle it.

The apprenticeship still exists

Ben Wills trained agency analysts with a kitchen rule: if you are going to be a sushi chef, the first year you just make rice. Keyword research and nothing else, across ten, twenty, fifty sites, until the relationship between a query and an intent stops being a slide and becomes instinct.

He thinks prompting deserves the same apprenticeship now — sit with it, work out how the answer gets built, work out what it takes to be in it. That is not a course you can buy for your team. It is hours you have to protect for them.

Which is the whole argument in one line: the tools got cheap, the judgment did not, and the judgment is now the only part worth paying for.