How to Read a Google Ads Search Terms Report as a Roofer

The search terms report lists what people actually typed before your ad showed. With broad match, that list can wander far from roofing, and every click on it is billed the same.

A keyword is an instruction to Google. A search term is what a real person typed when your ad appeared. On broad match, one roofing keyword can show your ad for do-it-yourself questions, supply stores and other trades, and each click costs the same as a homeowner with a leak.

Google's search terms report is the list of those searches. It has one limit worth knowing: searches too rare to show are left out for privacy, so the visible list is a sample, and any total built from it is a floor.

Sort each visible search into could become a roofing job, never will, or another roofer's name. Block the junk with negatives built from those terms, and decide the competitor pile on purpose.

Reading the report comes down to sorting. Every visible search goes into one of three piles.

The junk pile became a list of 31 negative phrases in September, every one taken from the account's own report rather than a generic download, and the keywords moved from broad to phrase match at the same time. A phrase match negative blocks any search containing those words in that order, which is why the list is short.

After the negatives went in, spend on the blocked searches went to zero and total clicks fell with it. That is the point. The measure that matters is leads and booked inspections, and Search had not booked one yet.

Ten days later I read the account back to check. Spend on the blocked searches had gone to zero, and total clicks and spend fell with it. That is the intended result, and on a dashboard it looks like the campaign shrank.

The honest part: paid Search had still not produced a booked job. The page those clicks land on only got its estimate form onto the first screen of a phone on September 27, and a paid call still showed up in A&J's own reports as someone who found them on Google. Clicks are cheap to count; the question is which ones became inspections, and that link is the next thing being built.

Once the junk was gone, a new pattern showed: about half of what we could see went to people typing another roofer's name, with no leads so far. Blocking them is easy. Whether to is a business call, some owners want those clicks and some do not, so it went to the owner as a decision instead of a quiet change.

When is none of this worth doing? If your campaign already runs on tight exact match keywords and a short list of towns, the report will be mostly clean. Read it once a month and spend the time on the follow-up instead.