Deciding which market gets more budget is only half the job. The other half is deciding how often you make that call. LSA budget reallocation cadence—the rhythm at which you move money between markets—sits between two failure modes. Reallocate too rarely and misallocated budget piles up unnoticed for weeks. Reallocate too often and you make decisions on noise, resetting learning faster than the data can stabilize. The right cadence is the one that catches drift while each decision still has real signal behind it.
The two ways cadence goes wrong
Both errors are common, and they pull in opposite directions:
- Too slow (the quarterly trap). LSA markets drift far faster than a quarter. A spam pattern appears, a competitor changes behavior, demand shifts with the season, a review dip softens rank. Review budgets only every few months and you let all of that compound—months of budget stranded in markets that stopped earning.
- Too fast (thrash). Change a market's budget every day or two and you never let a change propagate. A budget adjustment typically needs a couple of days to show its true effect; judge it the next morning and you are reading noise. Worse, constant up-down-up changes reset learning and guarantee the data stays noisy—so you make the next bad decision on the mess the last one created.
The goal is a cadence fast enough to catch real drift and slow enough that every move is backed by data that had time to settle.
Cadence should scale with volume
There is no single correct interval, because the right cadence depends on how fast a market produces trustworthy signal—and that depends on lead volume. A high-volume metro accumulates enough leads to read cost per booked job quickly, so it can be adjusted more often. A thin, rural market produces so few leads that any single week is dominated by luck; it needs a longer window before its numbers mean anything.
| Market volume | Signal speed | Suggested reallocation cadence |
|---|---|---|
| High (busy metro) | Fast—stable weekly | More frequent; weekly is readable |
| Medium (suburban) | Moderate | Every week or two |
| Low (rural/thin) | Slow—weeks to trust | Longer window; judge over multiple weeks |
Applying one cadence to every market is a mistake in both directions: it reallocates thin markets on noise and neglects busy markets between reviews. Match the cadence to the market's signal speed.
Separate the review loop from the emergency loop
A useful mental model splits cadence into two layers. The first is the routine reallocation review—the scheduled loop where you rank markets on cost per booked job and shift budget in measured steps toward the efficient ones. The second is exception response—the fast reaction to a protective signal that shouldn't wait for the next scheduled review.
Some conditions warrant acting immediately regardless of cadence: a budget on track to blow past its cap, a sudden spike in invalid or credited leads, or a market that has run dry midweek and gone dark. Those are protective moves, and protecting spend always comes before the routine growth decision. The scheduled cadence governs optimization; the exception loop governs emergencies. Confusing the two—either sitting on an emergency until the weekly review, or treating every wiggle as an emergency—is how cadence discipline breaks down.
Move in steps, and let each step breathe
Cadence and step size work together. Whatever your interval, reallocate in measured steps rather than swings—add budget to a strong market incrementally, and pull from a weak one toward a floor rather than to zero, since fully pausing an LSA campaign risks weeks of ranking recovery. Then let each step propagate before the next review reads it. A cadence that is otherwise sound gets ruined by oversized moves, because the big swing overshoots the sweet spot and floods the next cycle with noise.
What "enough data" actually means
The honest test for whether it is time to reallocate is not the calendar—it is whether you have enough booked-job data to trust the decision. A weekly cadence is a convenient default, but a week with three leads in a thin market is not a week of signal; it is a coin flip. Before moving money, ask: has this market produced enough booked jobs since the last change for its cost-per-booked-job number to be real? If not, wait, even if the calendar says review. Reallocating on a fixed schedule regardless of data quality is just thrash with a timestamp.
The cadence that compounds
- Run a routine loop that ranks markets on cost per booked job and moves budget in steps toward the efficient ones.
- Scale the interval to volume—faster for high-volume markets, slower for thin ones.
- Keep a separate exception loop for protective signals that can't wait for the schedule.
- Judge on settled data, not the morning after a change, and never on a single low-volume week.
Done consistently, this quietly compounds. Each cycle nudges a little more budget toward the markets that convert it and a little less toward the ones that don't—and the drift that would otherwise pile up over a quarter gets caught and corrected while it's still small.
Frequently asked questions
How often should I reallocate LSA budget between markets?
Often enough to catch drift, rarely enough that each decision has real data behind it—commonly weekly to monthly depending on lead volume. High-volume markets can be adjusted more frequently because they accumulate signal fast; thin markets need a longer window before their numbers are trustworthy.
Can I change LSA budgets too often?
Yes. Changing budgets every day or two—up, then down, then up—resets learning and forces decisions on noise instead of signal. A budget change needs time to propagate, usually a couple of days, before its effect is readable. Thrashing guarantees noisy data and worse decisions.
Why not just review LSA budgets once a quarter?
Because LSA markets drift faster than a quarter—spam patterns appear, competitors shift, demand moves with the season, and reviews rise or fall. A quarterly-only cadence lets months of misallocated budget accumulate before anyone looks. A regular loop catches drift while it's still small and cheap to fix.