Seasonality adjustments tell Smart Bidding about a short future conversion-rate change, such as a major promotion. Data exclusions tell it to ignore affected conversion data after a tracking outage or bad measurement period. They solve opposite problems: real customer behavior versus unreliable data.

VISUAL LESSON

What you will learn

  1. 01Distinguish real behavior from bad measurement.
  2. 02Configure the narrow supported control.
  3. 03Verify bidding and reporting effects responsibly.
A future demand wave and a broken measurement segment follow different Smart Bidding control paths
Real future behavior calls for seasonality guidance; corrupted conversion data calls for exclusion.

ILLUSTRATIVE WORKED EXAMPLE

Triage an illustrative anomaly set

Anomalies reviewedIllustrative queue
20
Planned short promotionSeasonality candidate
5
Tracking outage or bad dataData-exclusion candidate
7
Normal volatility or other causeDiagnose, no control
8
Illustrative example—not a benchmark. Replace every sample value with your own campaign, market, and measurement data.

PRACTICAL INTERFACE MAP

Move from anomaly diagnosis to controlled change

Evidence01Classify the event

Compare promotion calendar, site and tag releases, conversion logs, CRM outcomes, reporting lag, consent state, and campaign history.

Control02Choose seasonality or exclusion

Set the supported campaigns, conversion actions, devices, click-date window, expected rate change, and duration shown in the current flow.

Verify03Monitor without rewriting the report

Check control status, bidding behavior, reporting, corrected measurement, downstream quality, expiry, and a named owner.

Conceptual walkthrough. Labels, controls, and availability can vary by account, region, plan, and interface version; verify the current screen before acting.

STEP-BY-STEP LESSON

Anomaly → behavior or measurement → narrow control → expiry review

DiagnoseReal demand or bad data
ConfigureScope and dates
VerifyBids, data, quality

THE LEAD ATLAS METHOD

Lead Atlas Data can research business contacts for a paid campaign's precise categories, locations, and market while the advertiser protects bidding inputs with evidence-based controls.See how custom list research works ↗
01

Diagnose the anomaly

Create a timeline of campaign settings, promotion and price, inventory, site and form changes, tag releases, consent, conversion logs, CRM outcomes, imports, attribution, reporting delay, bids, budget, policy, and external events.

Decide whether customers truly converted at a different rate, measurement failed, or ordinary volatility and other campaign changes explain the movement. Do not use a bidding control to hide weak performance.

02

Choose the matching control

Use a seasonality adjustment for a known future, short event expected to create a significant conversion-rate change. Google says Smart Bidding already handles regular seasonality and describes the control as best for brief events, ideally about one to seven days.

Use a data exclusion when faulty conversion data would mislead bidding, such as a tracking outage or incorrect values. Exclusions apply to clicks in the selected date range and do not remove the data from normal reporting.

03

Set narrow scope and dates

Select only the supported campaigns and conversion actions affected, then set dates, timezones, devices, and any expected conversion-rate adjustment using documented evidence. Account for click-to-conversion delay before finalizing the window.

Avoid broad account-wide or long-running controls without evidence. Record the unaffected campaigns and conversions, approver, creation time, expected expiry, and conditions for edit or removal.

04

Repair measurement separately

For data incidents, fix the tag, consent flow, import, deduplication, value, event, or CRM source; validate with test conversions and reconciliation. Do not treat data exclusion as a substitute for correcting the pipeline.

For promotions, verify inventory, pricing, landing page, creative, operations, and post-event return to normal. Seasonality guidance cannot compensate for an unavailable offer or broken customer path.

05

Review effects and preserve evidence

Monitor control status, bids, budget, conversion delay, raw reporting, valid conversions, qualified outcomes, values, downstream quality, and system diagnostics. Compare with an annotated baseline and avoid causal claims from an uncontrolled before-after view.

Deliverable: anomaly timeline, root-cause decision, chosen control and rationale, exact scope and click-date window, conversion actions, delay analysis, measurement repair, test evidence, expiry plan, monitoring report, rollback owner, and postmortem.

THE TAKEAWAY

Use seasonality adjustments only for a forecastable short conversion-rate shift, data exclusions only for faulty measurement, and keep date, scope, click timing, and rollback evidence narrow.

OFFICIAL REFERENCES

Check the platform’s current instructions.

Platform labels, eligibility, and workflows can change. These official help pages were used to validate this lesson.