The negative keyword checklist I run before and after every launch

The cheapest money I ever save a client comes from a list of words we never want to show for. It costs nothing to add, and it protects every dollar spent after.

Negative keywords are the part of the account where the leak is visible and the fix is free. Here is the checklist I run before launch, and the one I run every week after.

The three leak classes

Almost every wasted click in a local service account falls into one of three buckets.

1. DIY and how-to. “how to unclog a drain,” “fix a garage door spring yourself,” “parts diagram for a water heater.” These searchers want instructions, not a technician. They click, they read, they leave, and the click is billed.
2. Parts and supply. “where to buy a lock cylinder,” “hvac filter near me,” “pressure washer for sale.” These are shoppers looking for a product, not a job. The volume looks real in the report. The booked-job rate is near zero.
3. Job seekers. “plumber jobs hiring,” “electrician salary,” “cleaning company careers.” Resumes and job alerts quietly feed on your keyword list, and every one of those clicks is a click you paid for and can never deposit.

These terms masquerade as high-volume queries in the platform report. The report shows them as interest. The ledger shows them as a leak.

Before launch: the seed list

Before a dollar spends, I add the negative list built from those three classes, tailored to the trade. For most local services that is forty to sixty terms at minimum, and it is built from three sources:

• The obvious DIY and parts terms for the specific service.
• The job-seeking terms for the trade and the area.
• Anything the keyword planner shows as high volume that clearly cannot become a booked job.

This is free insurance. It costs no spend to add, and it stops the most predictable bleed on day one. The full per-trade seed list lives in the negative keyword guide.

After launch: the weekly sweep

The seed list is a start, not a finish. New junk arrives every week as the platform matches your ads to searches you never named. The recurring checklist is short:

1. Pull the search terms report for the last seven days.
2. Sort by spend, not clicks. The expensive mismatches are the ones that matter.
3. Mark every term that cannot become a booked job and add it as a negative.
4. Check the match type. A phrase negative on “how to” catches more than a dozen exact-match additions.
5. Note what you added so the list compounds instead of resetting.

Ten minutes a week. The list grows, the bleed shrinks, and the budget concentrates on searches that can actually book work.

The test that tells you it is working

The number to watch is not impressions. It is the share of spend landing on searches that can become a booked job. If more than a few percent of weekly spend is going to terms you then have to negate, the list is behind the leak, and the sweep needs to run more often until it catches up.

A clean search-term report is not a cosmetic win. It is the difference between a budget that buys jobs and a budget that funds other people’s research.

I count conservatively. Every untracked job is your upside, not my credit.