Structuring an account that runs four months a year

Google states that a bid strategy re-enters a learning period when it is created, and when campaigns, ad groups or keywords are added to or removed from it, with calibration taking up to three weeks or one to two conversion cycles.

A seasonal account triggers that on purpose, once a year, at the worst possible moment. If your season is sixteen weeks and three of them are recalibration, you have spent nineteen percent of your entire advertising year learning.

The arithmetic gets worse if the conversion cycle is long. A considered purchase that closes in six weeks means one to two conversion cycles is six to twelve weeks, which for a four month season is most of it. That single sentence in Google's documentation is the reason seasonal structure is a real design problem and not a scheduling one.

Row two is the one to stop doing. Rebuilding a seasonal campaign from scratch every year, which happens more often than you would think because the previous year's build looks messy, throws away the only asset that carried over. Paused campaigns count toward Google's published ceiling of 10,000 campaigns per account including active and paused, which is not a constraint any real business will hit, so there is no volume reason to delete them.

Row three is the one people try and abandon. Running all year at a token budget against genuinely absent demand buys expensive clicks from the wrong people, and it teaches the bid strategy about a population that will not be your season's population.

Row four is the underrated option, and the only one that solves the actual problem. If the business has an off-season offer that shares a landing page and a conversion action, keeping it running keeps the conversion data recent and the structure warm. A pool business that sells covers and closings in autumn is not padding, it is a different real product with real conversions.

They come up in every conversation about seasonal accounts, and they solve a much smaller problem than the name suggests.

The Google Ads API documentation on creating seasonality adjustments states plainly that they are an advanced tool for Smart Bidding to predict changes in conversion rates for upcoming short events of one to seven days, and that they are less effective for longer durations of more than fourteen days at a time.

The mechanics are equally specific. An adjustment carries a conversion_rate_modifier between 0.1 and 10.0, requires start and end date and time, and has a scope of either CAMPAIGN, applied to named campaigns with a maximum of 2,000 campaigns per adjustment, or CHANNEL, applied to campaigns of a given channel type. It can optionally be restricted to specific device types.

So the tool is built for a weekend sale, not a season. If your peak lasts four months, a seasonality adjustment is the wrong instrument by Google's own description of it, and using one for the whole season is a good way to distort bidding for sixteen weeks.

There is a budget type built for exactly this shape. Google documents campaign total budgets for campaigns with a fixed start and end date, where you set an amount for the run rather than a daily average.

That removes a recurring seasonal argument. With an average daily budget, individual days run above the number and Google applies a monthly limit, which is described in Google's page on overdelivery and your average daily budget. Across a season that spans partial months at both ends, reconciling that to a fixed seasonal commitment is manual work every year.

A total budget makes the commitment the unit, which is how the client already thinks about it. The mechanics of the daily figure, and why a client sees a day above their number, are covered in budget pacing and overdelivery.

The most common seasonal mistake is starting on the day demand starts. That guarantees the recalibration period overlaps the earliest and often most profitable part of the season.

Start early enough that the learning period lands before demand does, using Google's own figure as the planning input: up to three weeks, or one to two conversion cycles, whichever is longer for your business. For a six week conversion cycle that means switching on roughly a quarter before you expect volume.

Accept that the ramp period looks bad and budget for it explicitly. A client shown week one of a ramp as a performance report will ask to stop, and stopping restarts the clock. Naming the ramp as a line item in the plan is the difference between a known cost and an argument.

Do the same in reverse at the end. Winding down mid-season to protect a monthly number is a structural edit that can restart the clock for whatever is left of the season.

Keep the seasonal work in as few campaigns as the accountability allows. Every extra campaign divides a compressed run of conversions further, and a seasonal account has less total data than an always-on one to begin with.

That argues against the tidy per-product or per-region splits that feel natural when you are building a season from scratch. The density test is the same one in how many campaigns is too many: divide expected seasonal conversions by the campaign count and ask whether each campaign has enough to be readable within the season, not within the year.

Where thin data is unavoidable, pooling is the tool. A portfolio strategy across the seasonal campaigns lets them learn from a common pool while staying separate for reporting and budget, with the trade-offs set out in shared budgets and portfolio strategies.

Do the structural work now. Any restructure, renaming, tracking correction or conversion action cleanup costs a learning period, and the off-season is the only time that cost is free. Doing it in week three of the season is how a good change gets blamed for a bad month.

Verify the conversion actions while nothing depends on them. Count settings, duplication, and whether a location or product parameter resolves on every event. That verification is the substance of a conversion attribution audit, and it is much cheaper to run against last season's data than against live spend.

Write down what you changed and when. A seasonal account is judged year over year, which means the person comparing this season to last is comparing across a gap where nobody was watching. A dated change log is what makes that comparison mean anything, for the same reason change history forensics works: the record of what changed is what turns a difference into an explanation.