For December campaigns, the safest Smart Bidding setup is the one you change least: keep your existing bid strategy, widen your target rather than tightening it, and let budget — not bid targets — carry the seasonal increase. Google Ads' seasonality adjustments are the one December-specific control worth using, and only for short, sharp spikes you can predict within a few days. Smart Bidding is Google's set of automated bid strategies — Target CPA, Target ROAS, Maximize Conversions, Maximize Conversion Value, and Enhanced CPC — that set a bid for every auction using signals from the query, device, time, audience, and location. The reason December is awkward is structural: these models learn from a trailing window of conversion data, and December's conversion rates, competitive pressure, and shipping-deadline behaviour all shift faster than that window updates.
Table of Contents
- Why Smart Bidding struggles in December
- Which strategy to run, and when to change it
- Setting targets you can actually hit
- Seasonality adjustments versus data exclusions
- Budgets, pacing, and the post-Christmas cliff
- A conservative December checklist
- Frequently Asked Questions
Why Smart Bidding struggles in December
Smart Bidding predicts a conversion rate for each auction from recent history. In early December, that history is November — a month whose conversion rates on many accounts look nothing like the week before Christmas. The model is therefore systematically behind the curve on the way up, and systematically ahead of it on the way down after your shipping cutoff. The second problem is auction inflation.
More advertisers bid harder in December, so the cost of the same click rises. If your Target CPA stays fixed while CPCs climb, the strategy's only way to hit that target is to stop entering the auctions it can no longer win profitably — which shows up as a volume collapse rather than an efficiency win. The third is attribution lag. December shoppers research early and buy late, so conversions credited to a 10 December click may not land in your account for days. Judging performance on a three-day window during this period will make your campaigns look worse than they are, and tempt you into exactly the panic edits that hurt most.
Which strategy to run, and when to change it
If a campaign is already on Target ROAS or Target CPA with a healthy conversion volume, leave the strategy in place. Switching bid strategy resets the learning period, and a learning period that runs through your highest-value fortnight is an expensive thing to trigger. The general guidance from Google is to expect roughly a week of unstable performance after a strategy change — time you do not have in mid-December.
Changing strategy is defensible in two cases. First, if you are budget-constrained and your real December goal is volume at any reasonable price, Maximize Conversions (or Maximize Conversion Value) without a target removes the efficiency ceiling that is throttling you. Second, if you have been running manual CPC or Enhanced CPC and have enough conversion history to support a target, the move to full Smart Bidding is worth making in November — not December.
- Already on tROAS/tCPA with stable volume: no change; adjust targets only
- Hitting target but starved of volume: loosen target 15-20%, raise budget
- Very low conversion volume (single digits per month): target-based bidding will be noisy; consider Maximize Clicks or manual control
- Mid-December and unsure: do nothing structural until January
Setting targets you can actually hit
Treat your target as a statement about the December auction, not about your annual average. If your gross margin supports a £40 CPA and December CPCs run materially higher than November's, a £40 target quietly becomes an instruction to withdraw from the market. Raising the target — a higher tCPA, a lower tROAS — is how you tell the strategy that you accept a worse ratio in exchange for the volume that only exists this month. Move targets in small steps and give each one time to settle.
Shifts of more than about 20% at once are usually treated as a significant change and can re-enter learning; several such edits in a week make the cause of any performance change impossible to isolate. One adjustment, then several days of observation, is slower and far more legible. Decide the floor before you need it. Work out the CPA or ROAS at which a December sale stops being worth having — including returns, which spike in January for gift purchases — and write it down. A target set from that number holds up under pressure; a target set on 20 December while watching a live dashboard usually does not.
Seasonality adjustments versus data exclusions
These are the two December-specific tools in the bidding toolkit, and they do opposite things. A seasonality adjustment tells Smart Bidding to expect a conversion-rate change over a defined future window, so it bids into the spike before the data proves the spike exists. A data exclusion tells it to ignore a past window entirely, so a period of broken tracking or anomalous performance does not poison future predictions. Google's own guidance is that seasonality adjustments are for short events — on the order of one to seven days — where you expect a large, predictable conversion-rate change, such as a one-day flash sale or a free-shipping deadline. They are not for "December is busy".
Applying one across the whole month overlaps the model's own learning and tends to overbid the ordinary days between the peaks. You set an expected conversion-rate change as a percentage, and a wrong estimate is applied with full confidence, so err small. Data exclusions are the more commonly needed of the two. If your conversion tag broke for two days, or a checkout outage flattened sales, exclude that window rather than letting the model read it as genuine demand collapse. Both tools live under Bidding in the shared tools area of the account and can be applied at campaign or account scope.
- Use a seasonality adjustment: a 24-hour promotion, a shipping-cutoff deadline, a known one-day traffic surge
- Do not use one: the whole of December, an ongoing promotion of several weeks, a change you are guessing at
- Use a data exclusion: tracking outage, site downtime, payment failures, a period of accidental spend
- Do not use one: a day that simply performed badly
Budgets, pacing, and the post-Christmas cliff
A Smart Bidding strategy cannot spend money it does not have. If your campaigns are limited by budget, the strategy prioritises the auctions it predicts will convert and declines the rest, which caps December upside regardless of how generous your target is. Check the budget status column before touching any target — a "limited by budget" label means the budget is the binding constraint. Plan the exit as deliberately as the ramp.
Demand for gift-shipping queries falls off a cliff at your carrier cutoff, and a strategy still bidding to a loosened mid-December target will overspend into a thinner market for days. Diary the date you tighten targets back and the date any seasonality adjustment ends, and verify the adjustment's end date rather than assuming it expired. Late December also brings a reporting trap. If you run a seasonality adjustment or a loosened target through the peak and then compare January performance against it, January will look like a failure. Compare like with like — year on year for the same period, or against the pre-peak baseline — before concluding that something broke.
A conservative December checklist
Most of the work that makes December Smart Bidding behave happens in November, and most of the damage happens in the last ten days of the month when someone edits under pressure. The sequence below assumes an account that already has working conversion tracking and enough volume for target-based bidding. One thing worth checking explicitly: if you share a portfolio bid strategy across campaigns with very different December behaviour — a gift category and an evergreen one, say — the single target applies to all of them, and the seasonal loosening you want for gifts is also applied where it is not warranted.
- Before December: confirm conversion tracking fires correctly, including on mobile checkout, and that conversion values are accurate — Smart Bidding optimises to what you report, errors included
- Before December: make any bid strategy change now, so the learning period finishes before peak
- Early December: raise budgets on campaigns flagged as limited; loosen targets once if volume is being throttled
- Peak days only: apply a seasonality adjustment for specific one-to-seven-day events with a conservative conversion-rate estimate
- Throughout: wait at least a week between target changes, and judge performance on windows long enough to absorb conversion lag
Frequently Asked Questions
Should I switch from Target ROAS to Maximize Conversion Value for December?
Only if your real goal is volume and you accept a worse return. Dropping the target removes the efficiency ceiling, but the switch can trigger a learning period, so make it in November rather than mid-December.
How large should a seasonality adjustment be?
Base it on the conversion-rate lift you have measured for the same event previously, and round down. The adjustment is applied as stated, so an overestimate causes overbidding across the whole window.
Can I use manual bid adjustments alongside Smart Bidding?
Device, location, and schedule bid adjustments are largely ignored by Target CPA and Target ROAS, which set their own per-auction bids using those same signals. Device targeting exclusions still apply.
Why did my December CPA rise even though I never changed the target?
Rising competition raises CPCs, and the strategy keeps bidding toward your conversion goal within a more expensive auction. A target is an objective the system optimises toward, not a cap it guarantees.
