For six months I wrote search proposals for other people while my own site had no prices on it, no dated figures, and a contact form that asked for everything at once. The contradiction only became obvious when a client asked me to demonstrate the audit I was proposing, and I opened my own domain to use as the example. Half of what I found had been sitting there since the first year.
The proposal that read like a mirror
A proper proposal has to state what will be done, in what order, by whom, and what it costs. Writing that section for a two-van repair business took a weekend, and every sentence I typed described something my own site was missing. Clear service pages, a price range, one page per subject, numbers with the month they were checked.
I had been selling the discipline from a page that did not practise it. The reader who lands on a site about careful routes deserves the same transparency I demand of a bus timetable, and mine offered neither prices nor dates. That gap did not show up in any ranking report, which is exactly why it survived so long.
Proposals also force a sequence. You cannot write step three until step two is decided, and putting the work in order on paper showed how much of my own site had been built in whatever order the trips happened to allow. A calendar shaped by ferry timetables is a poor editorial plan, and the document made that impossible to ignore.
The audit that took one hour
The check took an hour with a notebook: eleven pages that answered overlapping questions, four of them with numbers nobody could verify, and a contact form with nine fields that produced maybe one message a month. Two pages said the same thing in different words, so they split their own relevance instead of adding to it.
Nothing in the list was technical. There was no plugin to install, no schema to fix, no migration to survive. Every item was a decision I had avoided because writing one more post felt more productive than merging two old ones.
The embarrassing part was that I had billed a client for precisely this list the month before. Charging for advice and ignoring it at home is the fastest way to stop trusting your own recommendations, and I did not enjoy the six weeks that followed.
Reading your own site as a stranger also changes the order of things. I had assumed the homepage was fine because I visit it daily, yet the first screen explained the route notes and never mentioned the caption lists that half the search traffic came for. A stranger gets one glance to decide, and that glance was aimed at the wrong half of the site.
The fixes, in the order I did them
Merging came first: eleven pages became seven, with the stronger URL kept and the other redirected. Then the numbers, each one stamped with the month checked and a line saying where it came from. Prices went on the services page as ranges, with a note that they depend on scope rather than a promise of a call.
The contact form lost six fields and kept two: what the project is, and how to reply. Messages went from about one a month to four in the first three weeks, and none of them needed a follow-up to find out the budget.
Finally the pages that overlapped got one job each, so a reader who arrives from a search about captions does not land on a paragraph about analytics. It took two evenings, and the site has been easier to plan since because the calendar now fills gaps instead of repeating subjects.
The homepage got one honest paragraph about what is published here and when it was last checked, and the navigation lost four links nobody clicked. Small edits, but they are the ones a visitor reads before deciding whether anything else on the page is worth their time.
I now run the same document on this domain once a quarter, in a single sitting with a notebook. It takes an hour, it costs nothing, and it has caught every problem that later turned into a week of apologising to readers by email.
What I tell clients now
I show the same hour-long audit I run on my own domain, including the parts that are still weak, and the conversation gets shorter because the work is visible before anyone signs anything. Clients ask better questions once they can see a real list rather than a promise of one.
The lesson is not that my site was badly built, it is that a practice only counts once the practitioner is inside it. My contact form now asks two questions, the prices are where a reader can see them without sending a message first, and the audit I sell is the same one I run here every quarter.