Microsoft Ads upgrades bidding: what retailers should test

Microsoft Advertising announced an automated-bidding upgrade on 18 November 2026, with faster reactions to target changes and more even budget pacing. For retailers, the useful next step is to check…

SSophie Harrington1 min read

Microsoft Advertising announced an automated-bidding upgrade on 18 November 2026, with faster reactions to target changes and more even budget pacing. For retailers, the useful next step is to check conversion data before changing bid strategies ahead of Black Friday.

What Microsoft measured

In its announcement, Microsoft reports three separate results from internal global data:

  • 10% more conversion value per click at similar ROAS for Maximize Conversion Value or Target ROAS, using August 2026 Search and Audience data.
  • 8% more conversion value at similar ROAS when Search campaigns switched from Target CPA to Target ROAS, using August–September data.
  • 18% more conversions at similar CPA when campaigns switched from enhanced CPC to Maximize Conversions with a CPA target, using June–July data.

These are different comparisons, not gains to add together. Microsoft says further improvements will reach more accounts automatically into 2027; the announcement does not establish an activation date for every account.

Choose a test your store can actually interpret

First, reconcile purchase events and order values with your store records. Duplicate purchase events or missing values can make a revenue-based bidding experiment hard to trust. Record the strategy, target, budget and measurement settings before making a change.

Choose one campaign with enough useful history, rather than changing an entire account during a sale. Decide beforehand which outcome matters: profitable orders, revenue, or acquisition cost. A lower CPA is not automatically better if the orders are smaller or less profitable.

Microsoft recommends allowing a full conversion cycle between target adjustments. A simple week-before/week-after comparison around Black Friday can mix the bidding change with discounts, demand shifts and delayed purchases. Keep those effects visible in reporting; do not assign every improvement to the new model.

The vendor's averages support a controlled test, not an automatic budget increase. Search Engine Roundtable's report was the research handoff for this announcement.

Related reading: our Shopify BFCM preparation checklist.

ℹ️
The reported gains are vendor test averages from separate cohorts. They are not a forecast for your store.
Filed underMarketing

More from Marketing

See all →