AI ad spend review for small businesses

How the AI ad spend review works

Illustration with sample data.

Three AI helpers review a sample ad account for a made-up plumbing business against example rules. The owner picks a goal first: more calls, or a lower cost per call. The helpers suggest searches to cut, a new budget split and new ad drafts, and the owner approves the plan. Nothing in any ad account is changed.

Example rules

• Job seekers: employment-related searches are cut.
• Outside the area: a search that names a place outside Long Beach, Lakewood, Signal Hill or Seal Beach is cut.
• Too little data: a search with fewer than 20 clicks is held for the owner.
• No calls: a search with $40 or more spent and no calls in the period is cut.
• Costly calls, when the goal is a lower cost per call: a search with 20 clicks or more whose cost per call is more than twice the account’s is cut.
• Weak ad: an ad shown 1,000 times or more with less than half the clicks per view of the best ad beside it is replaced with a draft.
• Freed money goes to the campaign that runs out of budget early and fits the goal: the most calls, or the lowest cost per call.
• Drafts use only the owner’s facts. A line that needs anything else is held for the owner.
• The total budget stays the same.

What it does not do

• It does not change any ad account, budget or ad.
• It does not predict calls, savings or results.
• It does not write an ad line that needs a fact the owner did not give.

Who does what

• Joe, double-checker: Reads the search report and cuts searches that break a rule. Holds a search with too little data for the owner.
• Tyson, fast writer: Replaces the weakest ad with a draft that uses only the owner’s facts, and holds any line that does not.
• Maya, the boss: Moves the freed money to the campaign that fits the goal and keeps the total budget the same.