Singapore
Bing AI Overviews Ads: What Changed, and What to Fix First

Bing AI Overviews Ads: What Changed, and What to Fix First

As of November 2026, ads can appear inside AI Overviews on Bing, and Microsoft confirmed the placement on 19 November 2026 in its monthly Microsoft Advertising product post. There is no new campaign type to build: the ads are served through the Performance Max and AI Max campaigns an advertiser already runs. For a Singapore team, that turns the announcement into a measurement question. If a new placement arrives by default in an account that cannot separate its traffic, a seo consultant has no number to argue from, and the next budget decision gets made on impression volume instead of cost per conversion.

Key takeaways:

  • Microsoft confirmed ads in Bing AI Overviews on 19 November 2026, available in all markets except China, so Singapore accounts are included.
  • Advertisers do not set it up separately. It runs through existing Performance Max and AI Max campaigns.
  • The commercial case Microsoft makes rests on its own data, including a 21% increase in purchases after a Copilot interaction.
  • The announcement describes no report that isolates the placement, so the first job is to make the account measurable enough to judge it.

SEO news in 2026: what changed on 19 November 2026

Search Engine Roundtable reported on 19 November 2026 that Microsoft had confirmed ads in Bing AI Overviews, roughly three weeks after readers began spotting them in results. The confirmation sits in the Microsoft Advertising blog post "Put AI to work engaging customers in more places: What's new in Microsoft Advertising this November", also dated 19 November 2026, which is the official source the trade coverage points to.

Ads can show next to AI-powered answers on Bing when the offer is relevant. Microsoft's own example is a traveller asking Bing about a trip to Snowdonia National Park: the AI Overview shows a sample itinerary and travel tips, and a hotel advertiser's ad can appear beside that answer. The answer is a summary built from across the web, so the ad sits next to a recommendation the advertiser did not write.

No separate campaign setup is required. The placement runs through campaigns already in the account, including Performance Max and AI Max, which Microsoft describes as the most effective way to appear in AI experiences.

Availability covers all markets except China. A Singapore advertiser is therefore inside the launch, not waiting for a regional rollout, and the question shifts from eligibility to whether the existing campaigns are worth showing there.

What "no separate set-up" does not mean

The placement inherits the structure underneath it. Which campaigns serve, which final URLs expansion may reach, which assets are eligible and how conversions are counted all belong to the account as it stands today. If AI Max's URL inclusion settings are loose, the ad can point at pages a buyer should not land on, which is reason to check those settings now rather than after the first month of data.

A marketer at a desk reviews campaign settings on a monitor while ticking landing-page checks on a printed list.

Why Microsoft is putting ads next to AI answers

Microsoft's case has two parts. The first is about search behaviour. Microsoft's framing is that search has moved from a list of links to a faster path to an answer, and that answers often reveal needs beyond the words someone typed, so each answer is a chance to offer a next step. The Snowdonia example makes it concrete: the prompt is about a trip, and the useful ad is a hotel.

The second is about shopping behaviour. Microsoft cites its Holiday Survey Insights research from July 2026, which found that most holiday shoppers plan to use conversational AI to help with gift ideas: 76% in the US, 74% in Australia and 70% in the UK. Those figures describe intent to use AI, not intent to click an ad, and none of them is a Singapore number. A Singapore reader should treat them as context for why Microsoft is investing in the surface, not as a forecast for a local account.

The purchase claim is the one most likely to be quoted in a pitch deck. Microsoft says its data shows a 21% increase in purchases after a Copilot interaction, drawn from internal data on global Edge users at the end of 2024. Read it carefully: it measures purchases after an interaction with Copilot, not the lift from an ad inside a Bing AI Overview. It is Microsoft's own measurement, published without an independent audit, and the strongest commercial argument in the post rather than proof that the placement performs. A seo marketing agency quoting it to a client should say whose data it is and what it measured.

Who this affects in a Singapore marketing team, and how

A marketing manager with a modest Bing budget finds the placement switched on by default, because it rides on campaigns that already exist, so spend can move into a surface nobody planned for. The decision this week is whether to watch the campaigns as they are or cap the budget until the first month of conversion data is readable.

A performance lead running Performance Max and AI Max gains another placement without another campaign. The work is in the controls: which final URLs expansion can reach, which assets are eligible, and whether conversion tracking is clean enough that a traffic shift shows up as a change in cost per conversion rather than in reported volume.

A content or SEO owner now shares a screen with the ad. The AI Overview answers from web pages and the ad sits beside that answer, so a page that states the answer plainly is still worth building: it is the material the answer draws on, and the landing page is judged by the same reader in the same moment.

An analyst writing the monthly report faces the hardest part. Microsoft's announcement describes no placement-level report, so the account's existing dimensions carry the work: campaign, asset group, landing page and whatever URL-level reporting already exists. Where the data cannot separate the placement, a seo consultant writing the report should say so rather than infer it from a total.

This advice does not suit every business. A company that has never run Bing, or that sells offline with no tracked conversion, cannot judge the placement from platform data. The same applies to an advertiser whose Bing account copies a Google account with no local conversion setup: the first job there is a tracked test, not a budget shift.

Four colleagues around a meeting table discuss performance data on a laptop while one takes notes.

What to do this week, in order

  1. List the campaigns that will carry the placement. Open the Performance Max and AI Max campaigns live on Bing and write down which final URLs their expansion is allowed to reach.
  2. Read those landing pages as a buyer would. Check load speed, layout and whether the page answers the question the AI Overview was summarising.
  3. Fix conversion tracking before reading any figure from the new surface. One primary conversion, consistent tracking across the account, and a documented attribution window.
  4. Separate what the account can separate. Report by campaign, asset group and landing page; if the placement cannot be isolated, say that in the report instead of inferring it from a total.
  5. Set the decision rule in advance. Decide what cost per conversion and what conversion quality would justify more budget, and what result would end the test.
  6. Keep a manual sample of the answers. Record which prompts your buyers ask, which answers appear, and whether an ad is shown. A seo marketing agency that skips this step ends up explaining a platform change with no evidence from the market it sells into. If the account is new to the channel, start from how Bing ads work with AI marketing and conversions before adding spend.

Where Topkee sits on a Bing test

Topkee runs Bing/Yahoo ads as a full-process engagement, and the parts that matter here are already in scope, from a full-funnel Bing Ads programme to the monthly report. Account structure, budget allocation and bidding, geo and dayparting, automated rules and keyword research are planned before spend goes live. Search, display and audience ad formats are covered, with AI-assisted copy and image production finished by designers, and landing pages are audited with Weber's site-scoring tool on load speed, layout and content. Conversion setup uses global tracking to shorten the learning period, and TTO attribution goes beyond UTM to record channel, medium, creative and custom attributes. Reporting is a KPI report plus the next period's plan, covering budget control, click-through rate, conversion volume and quality, and cost per conversion.

On the organic side, Topkee's SEO work aims commercial keywords at the organic top ten rather than at publishing volume: keyword clusters and a per-term tracking list, index and structure diagnosis, on-page fixes, content added or refreshed against demand, and rank tracking by the engine, region, language and device agreed in the contract. Quoting follows the site's baseline and workload, rankings are not guaranteed, and the first three months are a delivery rhythm, not a ranking deadline. The GEO service, which tracks how a brand appears in AI answers, runs on monthly tiers with a set number of tracked questions and reviewed knowledge pieces, with blunt limits: sampling runs through an API and does not equal what a consumer sees, and a brand that is not mentioned is not thereby disliked.

One boundary belongs here. The knowledge base holds no page on Bing AI Overview ad placements, Copilot ads or Microsoft Advertising policy, so these facts describe the surrounding service rather than the new placement, and Topkee makes no Microsoft partnership claim. Bing's ad formats and its AI surfaces are defined by Microsoft, and any statement about reach in Singapore should come from Microsoft directly.

What to watch next

Placement reporting comes first. Until Microsoft publishes a breakdown that separates AI Overview impressions from the rest, an advertiser's evidence stays partial, and any agency claiming to isolate the placement should be asked how. That question is worth putting to a seo consultant before a report is signed off, because the answer decides whether the placement has its own line or hides inside a campaign total.

The second is AI Max's controls. URL inclusions for final URL expansion are being rolled out, and the post does not say whether they govern what appears in an AI Overview. A campaign owner should watch the product notes rather than assume.

The third is creative disclosure. Microsoft's AI content transparency controls are generally available across product, multimedia, audience, display and video ads, with advertisers responsible for their own declarations and for local law. A team using AI-assisted creative should know which assets carry a label.

The fourth is the local listing. Bing business profiles accept photos, opening information and reviews and feed into Bing Maps; a listing kept consistent across Bing, Yahoo, Google, Facebook and LinkedIn holds the name, hours and contact details steady wherever an answer is assembled. A wrong opening time costs a click no bid adjustment recovers.

A pedestrian checks a phone map outside a shopfront, glancing between the screen and the shop entrance.

Sources

  1. Microsoft Ads In AI Overviews on Bing, Search Engine Roundtable, 19 November 2026. https://www.seroundtable.com/microsoft-ads-in-ai-overviews-on-bing-42578.html
  2. Put AI to work engaging customers in more places: What's new in Microsoft Advertising this November, Microsoft Advertising, 19 November 2026. https://about.ads.microsoft.com/en/blog/post/november-2026/put-ai-to-work-engaging-customers-in-more-places-whats-new-in-microsoft-advertising-this-november
Category:
Share to:
Date: 2026-11-20
Wing Yau

Article Author

Wing Yau

Marketing Manager