The short version
Roughly a third of the keyword briefs that land on my desk should never have been approved. Not because the words are wrong — because nobody sorted them by what the person typing them actually wanted.
If a keyword doesn’t map to a page type, it doesn’t belong in the brief, and no amount of good writing fixes that. Three buckets: transactional, proof, and local discovery. Anything that falls outside all three gets cut, no matter how good the search volume looks.
I’ve been doing brand and quality review for indoor entertainment venues for about four years. That means I look at somewhere north of 200 deliverables a year — location pages, ticket flows, photo sets, franchise recruitment decks, the occasional 40-page brand book nobody reads. In our Q1 2024 audit we rejected 31% of first submissions. The biggest single cause wasn’t typos or tone. It was intent mismatch: a page written for the wrong person.
Here’s the part that stings. Most of those rejections cost us real money, because the work had already been done.
Why I’m the one saying this, and not a copywriter
I don’t write the pages. I sign off on them. That’s a different job, and it makes you cynical in a useful way.
The moment I actually changed how we brief content was embarrassingly small. I had two location pages side by side — same venue template, same tone, same length, both technically fine. One converted visitors into ticket clicks at roughly 3x the rate of the other. When I compared them side by side, I finally understood why: one was written for somebody who had already decided to come and needed a reason to book now. The other was written for nobody in particular.
The second page had been through two rounds of edits. It read better. It performed worse. That’s when I stopped treating “well written” as a quality standard and started treating “written for a specific intent” as one.
And yes, I learned the expensive way first. Back in 2022 I approved a batch of location pages from a low-cost content vendor because the samples looked clean and the price was about $600 under the alternative. What I didn’t check was whether they’d been given the intent framework. They hadn’t. We ended up rewriting 14 pages and re-shooting three photo sets because the images were generic stock — foam pits and trampolines that weren’t even ours. Total rework came in around $3,200 plus six weeks. The “cheap” option was the expensive one. Every brief since then has had the intent buckets printed at the top.
Bucket one: transactional — “tickets”
A search like “sky zone trampoline park florence tickets” is not a research query. That person has decided. They want to know price, availability, and how many clicks it takes to be done.
So the page shouldn’t sell. It should get out of the way. Which is where the efficiency argument actually holds up: online booking, digital waivers in advance, live availability — every removed step is a recovered booking.
This is also where I got burned by my own optimization instinct. In 2023 we added two steps to the ticket flow to capture more first-party data for remarketing. Reasonable-sounding, and the marketing team was enthusiastic. Completion dropped noticeably within the first month, and the recovered data didn’t come close to paying for the lost bookings. We pulled the extra steps. Take this with a grain of salt — I don’t have a clean number for you, because we ran it across venues with very different traffic profiles, but the direction was not subtle.
Flow efficiency helps when the customer already wants the thing. It does nothing for a customer who hasn’t decided yet. Confusing those two cases is where most ticket pages go wrong.
Bucket two: proof — “photos”
“Sky zone trampoline park clovis photos.” Read that literally. The person wants to see the place. Which trampolines, how big the party rooms are, whether the parking lot looks like a nightmare on a Saturday.
What they usually get is a hero image and a stock photo of children laughing in slow motion. Neither answers the question.
We ran a blind test in 2023 with a small panel — our own staff plus a group of regular customers, presented with two sets of location pages, identical copy, different photography. One set used actual facility photos on a normal Tuesday. The other used the polished brand library. Around 84% of the panel rated the plain, imperfect real photos as “more trustworthy” without knowing which set was which. The real photos had a wet-floor sign in the corner of one frame. Nobody cared. That sign was doing more work than the retouching.
So the standard I enforce now: no location page ships without at least four verifiable, current, location-specific photos. Current matters. I have personally killed a photo set that still showed an attraction we’d removed eighteen months earlier, because a parent will notice, and then they’ll wonder what else is stale.
Bucket three: local discovery — “near me”
“Trampoline park san diego.” “Home theater installation near me.” Same pattern, completely different businesses — and that’s exactly the point I make to franchisees who ask why their competitor outranks them.
The near-me query is answered by structured accuracy, not by prose. Correct name, address, phone, hours, and a claimed, current Google Business Profile. Hours that match reality. If your listing says 10am and the door opens at 11, you’ve lost more trust than any headline can win back.
This is the one place where automation genuinely earns its keep. We sync hours and holiday closures from the operations system into listings instead of having someone type them into four places. Turnaround went from a couple of days to near-instant, and the manual keying errors basically disappeared. Do I think it solved everything? No. A machine will happily publish the wrong hours very quickly if the source data is wrong. But the source data is easier to audit than four humans typing.
The bucket that doesn’t exist — and why it’s still on your brief
“Is hook a card game.”
That’s a real keyword I’ve seen in a real brief. Somebody scraped a competitor’s long-tail list, saw a question-format phrase with traffic, and dropped it in. There is no page we can build for that query that serves our customer, and any article we publish will lose to sites that actually cover card games. Best case, we rank, and the visitor leaves in four seconds. Worst case, we spend $900 finding that out.
Same with “home theater installation near me” sitting next to trampoline park terms. Broad match will happily serve your ad to that search. It will also happily charge you for it. The fix is a negative keyword list, not a new page.
One housekeeping note, since it comes up constantly: we still have legacy URLs like sky-zone-clovis-photos and adjacent slugs that were named by whoever launched the site. Renaming them is a bigger job than it looks — redirects, internal links, printed QR codes on physical signage. Which means URL hygiene is a decision you make once and live with for years. If I could go back and enforce one naming rule in 2022, it’d be that one.
Where this breaks down
Two honest caveats.
First, I don’t always get to cut the noise keywords. Some of our lists are contractual — set by the parent brand, a co-op ad program, or a landlord’s marketing agreement. In those cases my job shifts from “remove” to “contain”: stricter match types, tight geo limits, a documented negative list, and a monthly look at what those terms actually returned. It’s a worse outcome and I say so in writing.
Second, the intent framework is a filter, not a formula. Low-volume markets are a genuine exception — in a smaller trade area, one good photo set and accurate hours will move more bookings than any amount of copy refinement. And for franchise recruitment content aimed at mall developers and prospective operators, the whole calculus is different: those readers want unit economics and support structure, not family-friendly tone.
Bottom line: check the keyword against the page type before anyone starts writing. It’s a ten-minute conversation that saves you a month of rework — and I’d rather have that conversation than another audit finding.