He built it with
Claude Code. He still didn't know what he was missing.
Ben is a developer. He built HourHound, a happy hour directory covering thousands of venues across ten cities, using Claude Code. He could implement anything he was handed. The open question was whether anyone could tell him something he didn't already know.
A capable builder is a harder client, not an easier one.
Most people who hire an AI consultant are hiring capability they don't have. Ben already had it. He'd shipped a working product on his own, with AI tooling, at a scale that would take a small team months.
That changes what an engagement has to be worth. Handing him a list of problems would have been close to useless, because finding problems was never his constraint. He could read a report, agree with it, and still be exactly where he started: with a to-do list and no more capability than he had the week before.
The most valuable page in every city he covered was showing search engines almost nothing.
HourHound's whole value sits on its city pages. Somebody looking for a drink in a new neighborhood lands there, and that page is the product. Those were the pages a search engine could see the least of.
That finding is the part of the engagement that reads like a normal audit, and it's the part Ben already had the skills to act on once he knew. It's also the smaller half of what got delivered.
The deliverable was a teaching document.
Every fix arrived as a prompt he could paste into Claude Code and run himself. That's a deliberate structure. A fix you're handed is worth one fix. A prompt you can read, run, adapt, and reuse is worth every future instance of that problem, and Ben is the kind of client who will do exactly that.
Six runnable prompts
One per fix, written to paste straight into Claude Code. One of them teaches him to have Claude Code write a script that calls Google's Indexing API, so new venue pages get submitted automatically instead of by hand, forever.
A primer on what actually applied
What SEO, GEO and GBP each are, and which of the three matters for a directory like his. Two of those acronyms get sold hard to people they don't apply to.
A sequencing call
Ship all three page fixes before submitting the sitemap. A search engine's first crawl impression is sticky, and inviting it to look before the pages are ready spends something you can't get back.
Search Console, step by step
Set up, verified, and explained, so the measurement exists before the changes do. Otherwise there's no way to know later whether anything worked.
Three analytics options, one recommendation
With the reasoning shown rather than a single tool asserted, because the right answer changes if his traffic or his budget changes.
A monetization map he never asked for
Five revenue paths, each tagged by whether it can work before traffic exists or only after. It opens: you didn't ask me for a business plan, but you built something with real assets, so here are some ideas.
I built HourHound with Claude Code, so I figured I already knew what these tools could do. Annette sent back an audit where every fix came as a prompt I could paste and run myself, including a couple of things Claude Code hadn't told me. She also caught that the most important page in every city I cover was showing Google almost nothing. Then she added a whole section on how to actually make money from what I'd built, which I never asked for. Learning to do this stuff myself was worth more than the fixes.
Google has started showing HourHound. We're not claiming credit for it yet.
Checked on August 8, 2026: HourHound ranks for 932 keywords in Google, and 33 of those sit on page one. None are in the top three. That split matters more than the headline number, because a keyword ranking at position 47 sends nobody. The pages doing the ranking are venue pages, the individual bar and restaurant listings, across Denver, Tampa, Phoenix, Chicago, Naples and Fort Lauderdale.
Two honest caveats on those figures. They're third-party estimates rather than his own Search Console, and a new site accumulates rankings on its own as Google works through it, so arriving after an engagement isn't the same as being caused by one. Attributing this to the audit would need his data and his account of which fixes he shipped and when.
The AI search picture is different and worth separating out, because they're measured differently and they move at different speeds. HourHound's visibility there has been tracked weekly since July 13, 2026, across ChatGPT, Claude, Gemini and Perplexity, at eighty queries per run. The most recent run, August 3, came back with zero organic mentions, same as every run before it.
The fear this answers is dependency.
The question that stops most people from hiring a consultant is what happens afterward. You pay someone, a thing gets built, and then it breaks or needs changing and you're back on the phone to the only person who understands it.
Team enablement is one of the five things Verity Agentic does, and this is what it looks like in practice: the client ends the engagement more capable than he started, holding prompts he can run again, on a problem he can now recognize on his own. Ben's closing line is the whole argument, and he made it better than we could. Learning to do this stuff himself was worth more than the fixes.
What would you want to be able to do yourself?
Tell us where the work gets stuck. We'll tell you whether the right answer is building it for you, teaching you to build it, or leaving it alone.
GET THE FREE CONSULT →Already know where the problem is? The Bottleneck Audit maps it and hands you a prioritized plan of what to automate first, from $500, credited toward a build if you go ahead.