Headingley Roofing

The brief

Paul Hughes spent fifteen years on Leeds roofs before going out under his own name. A new trading name has no reputation and no rankings, and the work starts arriving before either one is built, so the site had to be up before the first month of trading rather than after it.

The challenge

Ten working days is our standard, and ten working days would have put this site live after Paul's first month of trading rather than before it. A sole trader starting out has the opposite of a marketing budget and the opposite of time: the site is the first thing a customer checks and the last thing he has room to think about.

Our approach

The pages are generated from data rather than hand-built, so one service and one town are each a row rather than a file, which is what put 149 pages live in the time a ten-page site usually takes. The rest of the four days went where it shows: real content per page rather than a swapped town name, the Google Business Profile set up and verified alongside the site rather than afterwards, and the performance work done during the build instead of being left as a job for later.

Decisions, and what they beat

• D1

  Name the ten services the way a customer says them, trade terms as chips underneath

  Instead of

  The installer's own list (rejected: seven of thirteen return zero Leeds search volume)

• D2

  Cover 137 places, Leeds plus about fifteen miles

  Instead of

  A county-wide list (rejected: 94 such pages on a comparable site earn two visits a month)

• D3

  Generate every page from data, one service and one town per row

  Instead of

  Hand-built pages (rejected: 149 of them inside four days is not typing)

• D4

  Ship the schema with no address and no rating on it

  Instead of

  A plausible address and star rating (rejected: a fabricated citation has to be un-cited everywhere)

What we'd do differently

The 136 area pages are a coverage layer, and a coverage layer earns close to nothing on its own: on a comparable Leeds roofer, 94 of them across a year returned about two visits a month between them, because every search that matters returns a map pack above the organic results. We would spend that share of the build on the Google Business Profile and the first ten reviews instead, and add an area page only where there is a job on that street to put on it.