Seasonality Adjustments and Data Exclusions in Smart Bidding
When and how to use Google's seasonality adjustment and data exclusion tools so Smart Bidding doesn't overreact to short-term spikes.
Written by Mantas Jurgutis — Founder, Adsify — builds the Google & Meta automation merchants use daily
Editorially reviewed by Adsify Editorial on June 20, 2026 — Reviewed against Shopify, Google Ads and Meta official documentation.
Two distinct tools with different purposes
Google Ads gives Smart Bidding two related but distinct override tools, both found under Tools and Settings > Bid strategies: seasonality adjustments and data exclusions. Seasonality adjustments tell the system to expect a temporary conversion rate change (higher or lower) for a defined window, useful for planned events like a flash sale. Data exclusions tell the system to ignore a past date range entirely when training its models, useful for periods where conversion data was distorted by something abnormal, like a site outage or a tracking bug. Using the wrong one for the wrong situation can actively hurt performance.
When to use a seasonality adjustment
Seasonality adjustments are appropriate for short, planned windows where you expect conversion rate to shift meaningfully from normal — a 48-hour flash sale, a Black Friday/Cyber Monday weekend, or a planned site-wide discount. You specify a date range (Google recommends keeping these short, generally under 7 days per adjustment) and a conversion rate change percentage, which tells Smart Bidding to adjust its bidding for that window without waiting for enough live conversion data to detect the shift on its own — critical because by the time the algorithm would naturally detect a spike, a short sale window may already be over.
Estimating the right adjustment percentage
Google's own guidance suggests basing the percentage on your actual historical performance during comparable past events rather than guessing. If last year's Black Friday saw a 40% lift in conversion rate over a normal week, a seasonality adjustment of roughly +40% for the matching window this year is a reasonable starting point, refined by any changes in discount depth or promotional reach since then. For sales events without prior history, start conservative (10-20%) since an inflated adjustment can cause Smart Bidding to overspend chasing a conversion rate lift that doesn't materialize.
Applying adjustments at the right scope
Seasonality adjustments can be applied to specific campaigns rather than the entire account, which matters for a Shopify store running a sale on only part of its catalog. If a promotion applies only to a subset of products represented in one PMax asset group or one Shopping campaign, scope the adjustment to that specific campaign rather than applying it account-wide, where it would incorrectly tell Smart Bidding to expect a conversion rate lift on unrelated, non-discounted product lines too.
The danger of not removing an adjustment after the window
A seasonality adjustment is designed to auto-expire at the end of its specified date range, but it's worth confirming this happened correctly by checking Tools and Settings > Bid strategies > Seasonality adjustments after the promotional period ends. An adjustment that's still technically 'active' due to a date range error (a common mistake is entering the wrong year on a date picker) will continue distorting bidding expectations well after the actual promotional lift has ended, leading to overspend chasing a conversion rate that's since returned to normal.
When to use a data exclusion instead
Data exclusions are the right tool when a past period's conversion data was abnormal in a way that doesn't reflect real, repeatable customer behavior and shouldn't inform future bidding decisions — for example, a week where your Shopify store had a checkout bug suppressing real conversions, a period with broken conversion tracking that under- or over-reported, or an unusual one-off traffic spike from unrelated media coverage. Rather than adjusting expectations forward like a seasonality adjustment, a data exclusion tells Smart Bidding to disregard that historical window when training on past performance.
How to set up a data exclusion
Under the same Bid strategies section, choose Data exclusions and select the affected date range along with the specific campaigns impacted. Google recommends applying data exclusions as soon as you identify an anomaly rather than waiting, since Smart Bidding models continuously incorporate recent data — the longer a distorted period remains un-excluded, the more it can skew target CPA or target ROAS calculations calculated from that history.
Distinguishing a genuine anomaly from a real trend
The hardest part of using data exclusions correctly is telling the difference between a one-off distortion and an early sign of a genuine shift in demand or performance that Smart Bidding should actually be learning from. If conversion rate dropped because of a real, ongoing issue (rising competitor activity, a genuine product-market fit problem, a price increase), excluding that data would hide a real signal the algorithm needs. Reserve data exclusions specifically for verifiable technical or operational anomalies — not for periods that simply underperformed for reasons rooted in real market conditions.
Interaction with Target CPA and Target ROAS targets
Both seasonality adjustments and data exclusions affect the underlying model Smart Bidding uses, but they don't override your explicit Target CPA or Target ROAS values — those remain set by you. If a seasonality adjustment or data exclusion reveals that your target itself is now unrealistic (for example, after excluding a broken-tracking period, the true recent cost per conversion is meaningfully higher than your current Target CPA), update the target directly rather than relying on these tools to compensate for a stale target.
Documenting adjustments for future reference
Keep a simple internal log (a shared spreadsheet works fine) of every seasonality adjustment and data exclusion applied, including the date range, the reasoning, and the actual outcome once you can measure it. This becomes valuable historical reference for estimating next year's seasonality adjustment percentage more accurately, and prevents the common mistake of re-diagnosing the same recurring issue (like a known site slowdown during a specific promotional period) from scratch each time it recurs.
Seasonality patterns specific to ecommerce calendars
Beyond the obvious Black Friday/Cyber Monday window, ecommerce stores often have secondary seasonal patterns worth pre-planning seasonality adjustments for: a post-holiday returns and exchange bump in January, a pre-summer or back-to-school period depending on category, and any recurring in-house promotional cadence (monthly sale days, subscriber-exclusive events). Build a running calendar of these dates alongside expected conversion rate impact from prior years, and set seasonality adjustments a few days in advance of each one rather than reacting once the event has already started.
How Adsify's optimizer relates to these settings
Adsify's optimizer adjusts bidding signals roughly every six hours based on live profit and POAS data, which helps catch shorter-term underperformance faster than a manual review cycle would. It doesn't replace the specific seasonality adjustment and data exclusion tools described here, though — those remain campaign-level settings inside Google Ads itself, and a merchant running a planned sale event should still set the seasonality adjustment directly in Google Ads ahead of the promotion rather than relying solely on any third-party optimizer to react after the fact.