How often to review Google Ads search terms

The cadence question is more important than most operators give it credit for. Wrong cadence is the difference between catching leakage early and finding it after eight months of compounding waste.

There is no single right answer; the right cadence depends on spend volume, campaign age, and how recently the account changed. This piece is the working version of that decision.

The default cadence by spend level

Three numbers, three cadences. These are starting points; the cadence can tighten if signals demand.

Under $500/month in spend. Monthly is fine. The volume is not high enough to surface much new pattern between weeks, and the leakage cost of waiting two extra weeks is small in absolute dollars.

$500 to $5,000/month. Every two weeks. This is where most small-business accounts live, and it is also where the leakage compounds fastest. Bi-weekly is the cadence that keeps the obvious blocks fresh without becoming a weekly chore.

Above $5,000/month. Weekly. The pattern shifts faster, the leakage in absolute dollars is bigger, and the marginal hour of work pays back inside a week.

These numbers assume a settled, mature campaign that has been running for at least three months. Newer campaigns need a tighter cadence regardless of spend.

The first thirty days, why they matter

A new campaign produces more findable junk per week than the same campaign will at month four. Three reasons.

Match types are still expanding. Google's matcher takes a few weeks to settle into a stable distribution of close-variant matches. Early on, the matcher is testing wider variants; later it narrows toward what is converting.

Quality Score is still being learned. Until your ads have run against enough impressions to establish a Quality Score baseline, your CPCs are noisier and the queries you serve against shift more.

The negative list is mostly empty. A campaign in week one has no negatives. Every junk query comes through. By week eight, the obvious blocks are in place and the remaining junk is finer-grained.

The combined effect: a weekly sweep in the first thirty days catches a higher fraction of total-account leakage than weekly sweeps in months two through twelve combined. The compounding savings start day one and last for years, so the time investment in the first month repays itself faster than any other operator action on the account.

What changes the cadence: four signals

The default cadence is fine until something specific changes. Four signals tell you to either tighten or loosen.

Signal one: ad copy changes. New ad copy can shift which queries get served. After a copy change, do an extra sweep one week later regardless of normal cadence.

Signal two: budget shift. Doubling a campaign's budget often pulls in a different distribution of queries because the matcher serves against a wider set when budget allows. After a budget change of more than 25%, treat the campaign as new for a few weeks.

Signal three: a competitor change. When a major competitor enters or exits your market, the queries you serve against shift because the competitive bidding landscape has changed. Sweep within a week of any visible competitor shift.

Signal four: seasonality. For seasonal businesses, the start of the season shifts query patterns. The first two weeks of the busy season warrant weekly sweeps regardless of normal cadence.

What the cadence is for

The right cadence is not "review once a unit of time." It is "look at the report often enough to catch new junk before it costs more than the review takes."

A simpler framing: the cost of a sweep should be less than the leakage it would have caught. If a typical sweep finds three negatives that save $40 a month combined, and the sweep takes thirty minutes at any reasonable hourly rate, the sweep paid back. If the sweep finds nothing useful and takes forty minutes, the cadence is too tight.

This produces a self-correcting dynamic. The first sweeps on a new account find a lot of junk because the list is empty. The cadence is justified. By month four or five, sweeps find less per session because the obvious junk is gone, so the cadence can drop. By month eight, monthly sweeps on most accounts are producing only one or two new findings, which is the right time to consider quarterly.

The signs that the cadence is too tight

Two signs.

The sweep produces zero new findings two times in a row. If you have done two consecutive weekly sweeps and found nothing, the weekly cadence is now over-investment. Drop to bi-weekly.

You catch yourself dreading the sweep. Operator psychology matters. A cadence you dread is a cadence you skip. Better to do a thirty-minute monthly sweep reliably than a weekly sweep that gets postponed every other week.

The cure for both: drop the cadence by one step (weekly to bi-weekly, bi-weekly to monthly) and see if the new cadence still produces findings. If it does, settle there. If it does not, drop again.

The signs that the cadence is too loose

Three signs.

A specific high-cost junk query has been visible in the report for multiple sweeps without being blocked. That means the previous sweep missed it, or the cadence is so loose that the query was visible in only one sweep before being noticed. Either way, the leakage compounded longer than it should have.

The "(other search terms)" share of spend has grown month over month. If "(other)" is taking a larger and larger share, the visible-query leakage may also be growing in ways the report is not surfacing in time. Tighten cadence.

The campaign's cost-per-acquisition is drifting up. Negatives are not the only lever, but they are an inexpensive one to check first. If CPA is rising and the last sweep was a month ago, an extra sweep before deeper diagnostic work is cheap insurance.

A working schedule for a multi-account operator

For PPC managers with three to ten accounts, the cadence question becomes a scheduling question. The simplest workable schedule is a weekly cycle where each account gets visited at a frequency matching its spend.

This schedule sounds like a lot when written out. In practice it adds up to about three to four hours of search-terms work per week across an entire portfolio of accounts, which is about right for a portfolio of seven to ten accounts in the $500-to-$5,000-per-month spend range.

When monthly stops being enough

Eventually most growing accounts cross the threshold where monthly is no longer the right cadence. The signals are usually clear: rising spend, drifting CPA, an audit finding three or four obvious negatives that should have been caught earlier.

When that happens, the right move is not to switch to weekly immediately. Switch to bi-weekly first and observe for a month. If bi-weekly is keeping up, settle there. If it is not, then go to weekly.

The reason to step rather than jump: weekly is the most expensive cadence in operator time, and most growing accounts find that bi-weekly is sufficient until they are well past $5,000 a month. Jumping straight from monthly to weekly often over-corrects and produces sweep fatigue without proportional benefit.