Dynamic Search Ads are ending: plan the move to AI Max

In many store accounts, Dynamic Search Ads do a quiet job: they catch searches that nobody created a keyword for and build the headline and landing page from the website content. This campaign format is being retired. Starting in February 2027, Google will automatically upgrade remaining campaigns that use Dynamic Search Ads to AI Max (Google Ads Help). For campaigns using automatically created assets and the campaign-level broad match setting, the automatic upgrade already began in September 2026 (Google Ads & Commerce Blog). This guide explains what changes and how online stores can manage the switch themselves before the deadline: take stock of campaigns, prepare landing pages and page feeds, test in an experiment. The technical basics are on our page about the Google Ads integration.

The timeline: announced in April, postponed in June

Google announced the switch on April 15, 2026 (Google Ads & Commerce Blog). Originally, campaigns using Dynamic Search Ads, automatically created assets and the campaign-level broad match setting were all to move to AI Max starting in September 2026 (Google Ads & Commerce Blog). On June 11, 2026, Google updated the post and changed the timeline for Dynamic Search Ads (Google Ads & Commerce Blog). According to the post published the same day on Google’s developer blog, the automatic upgrade was postponed from September 2026 to February 2027 (Google Ads Developer Blog), and since June 15, 2026, new Dynamic Search Ads can be created again (Google Ads Developer Blog).

We are extending the timeline for Dynamic Search Ads sunset and auto-upgrade, which will begin in February 2027.

Google Ads & Commerce Blog, We’re upgrading Dynamic Search Ads to AI Max, update from June 11, 2026

For planning, the more detailed calendar Google published on its developer blog in August is what counts. According to it, Google shows in-account banners in affected accounts from September 2026 that encourage a voluntary upgrade (Google Ads Developer Blog). On January 15, 2027, a reminder is to go to all affected accounts (Google Ads Developer Blog). Google names February 1 to February 28, 2027, as the window in which the automatic upgrade begins (Google Ads Developer Blog); during that window, the ability to create new Dynamic Search Ads ad groups is permanently removed (Google Ads Developer Blog). Campaigns with dynamic ad groups become standard ad groups with AI Max (Google Ads & Commerce Blog).

Google does not say how many merchants are affected. Official statistics provide a frame: in 2024, 32.6 percent of enterprises in the EU used paid internet advertising (Eurostat), and 76.8 percent of these advertising enterprises used contextual advertising based on page content or search terms (Eurostat). The figures cover enterprises with at least ten employees; the release gives no separate value for Germany.

What AI Max is and how it differs from Dynamic Search Ads

AI Max is not a new campaign type but an optimization layer within existing Search campaigns (Google Ads Help). It bundles three features (Google Ads & Commerce Blog) and adds new controls. Dynamic Search Ads selected searches based on website content and generated the headline mainly from the HTML title of the page (Google Ads Help). AI Max goes further: it combines advertiser input such as ads and website content with additional signals to find more relevant searches (Google Ads & Commerce Blog).

What happens to your campaigns during the switch

Google wants the switch to keep performance as stable as possible and therefore carries over the existing settings. Campaigns with dynamic ad groups become standard ad groups, and settings and data move into AI Max (Google Ads & Commerce Blog). All three features, namely search term matching, text customization and final URL expansion, are turned on, and the existing URL controls are preserved (Google Ads & Commerce Blog). If you switch earlier yourself with the upgrade tool in the account, the confirmation dialog shows what changes:

• Ad groups: dynamic ad groups become standard ad groups (Google Ads Help).
• Ads: Dynamic Search Ads become responsive search ads; Google generates the minimum required assets through text customization (Google Ads Help).
• Targeting: targeting types that AI Max does not support become read-only. This applies, for example, to URL rules based on page title or page content: afterwards they can only be removed, not created or edited (Google Ads Help).
• Transition: the old ads keep running until the new ones have passed policy review; then Google pauses them (Google Ads Help).

For campaigns with automatically created assets or the campaign-level broad match setting, the switch began in September. For campaigns with the broad match setting, Google turns on search term matching by default; for campaigns with automatically created assets, search term matching and text customization (Google Ads & Commerce Blog).

Why the switch is mainly a data question for online stores

AI Max writes ads from what is on the website. Text customization combines snippets from page titles, descriptions and meta tags with generative AI grounded in the landing page content (Google Ads Help). Google reviews generated assets at least every 48 hours and refreshes them as needed (Google Ads Help); if a landing page changes, the text is refreshed within 48 hours (Google Ads Help). With changing prices and promotions, an ad may therefore rely on the previous version of the page for a while.

Responsibility for the content stays with the advertiser. Google explicitly points out that website content must be accurate, not misleading and compliant with the law, and recommends monitoring generated text in the asset report (Google Ads Help). Under German law, the standard is unfair competition law:

A person acts unfairly if they engage in a misleading commercial practice that is likely to cause the consumer or other market participant to take a transactional decision that they would not otherwise have taken.

Section 5(1) UWG, German Act against Unfair Competition (own translation)

Under Section 5(2) UWG, a commercial practice is misleading, among other things, if it contains untrue or deceptive information about the availability of goods, about a specific price advantage or about the price. A generated headline such as "Clearance, heavily reduced" from a promotion that ended long ago is therefore a legal risk. Prevention works mainly through current source data and through text guidelines, which are described further below.

For AI Max to select a page, Google has to be able to fetch it. The ads crawler AdsBot does not follow the global group User-agent: * in robots.txt; rules for it have to name it explicitly (Google crawler documentation). If robots.txt blocks AdsBot, Google may disapprove ads because the landing page is not accessible; landing pages should respond with status 200 (Google Ads Help). Which pages the crawler actually fetches is shown by a log file analysis for online stores.

• Check page titles and meta descriptions, because they are raw material for generated headlines
• Tie promotional text to run times so that it disappears from the page when the promotion ends
• Clearly mark sold-out products or redirect them to a suitable category
• Exclude filter and sorting URLs, cart, customer account and legal pages as landing pages
• Remove redirect chains and error pages from campaigns and page feeds; if addresses change, a redirect plan helps
• Check tracking templates for the {lpurl} placeholder

On the last point: Google warns that dynamic landing pages can end on an error page with tracking templates that are not set up compatibly, and recommends templates that start with {lpurl} (Google Ads Help). How to control filter URLs is covered in our post on faceted navigation. Address changes in Shopware belong before the switch, see SEO URL templates in Shopware.

Page feeds from the store: the control that remains

Many stores control Dynamic Search Ads through page feeds: a list of URLs, each with one or more custom labels. This work is not lost. Page feeds can also be used with AI Max. With final URL expansion turned on, labels control URL exclusions at the campaign level and inclusions at the ad group level. If you want to target only the feed URLs, use labels as inclusions in the ad group and turn final URL expansion off (Google Ads Help).

Fixed limits apply. An account can have no more than 100 page feeds (Google Ads Help), and each URL can carry up to 20 labels (Google Ads Help). It can take an average of 2 to 14 days for Google to crawl a new or edited feed, depending on the size of the feed (Google Ads Help). URLs belong in the feed without tracking parameters, duplicate URLs cause an error on upload, and cross-domain redirects are disapproved (Google Ads Help). A page feed in the format of the Google template looks like this:

Page URL,Custom label
https://shop.example.com/hiking-boots/,CATEGORY;STOCK_GOOD
https://shop.example.com/hiking-boots/model-a/,PRODUCT;MARGIN_HIGH;STOCK_GOOD
https://shop.example.com/rain-jackets/,CATEGORY;SEASON_WINTER
https://shop.example.com/promotion/,PROMOTION_RUNNING

An online store has an advantage here: the data for such labels already exists. Stock levels, margin, season and running promotions are stored in the store or the ERP system. A feed that is regenerated from them every day keeps the labels current without anyone maintaining spreadsheets. How similar product data can be used for Google is shown in our post on optimizing the Shopping feed. If you are working on the feed anyway, you can add the new conversational attributes in the product feed in the same pass.

Taking stock: which campaigns are affected

It starts with a list of all active campaigns and ad groups with Dynamic Search Ads. With the Google Ads API, a query like this one is enough. It returns all active Search campaigns that have a domain set for Dynamic Search Ads:

SELECT
  campaign.id,
  campaign.name,
  campaign.status,
  campaign.dynamic_search_ads_setting.domain_name
FROM campaign
WHERE campaign.status = 'ENABLED'
  AND campaign.dynamic_search_ads_setting.domain_name IS NOT NULL

For campaigns with automatically created assets or the broad match setting, version 25 of the API and later let you query when Google upgraded them, using the fields aca_migration_date_time and broad_match_migration_date_time (Google Ads API). For each campaign, these points belong in the inventory:

• Targeting per dynamic ad group: categories, URL rules, page feeds, rules based on page title or content
• Exclusions and negative keywords that have to move along
• Bid strategy: manual CPC or automated bid strategy with a conversion goal
• Budget: is the campaign already limited by budget today?
• Shared budgets, portfolio bid strategies and running experiments
• Tracking templates at all levels
• Performance of recent months per ad group as a baseline

Two points on the list decide the way forward. Search term matching does not work with manual CPC, because AI Max relies on signals from automated bid strategies (Google Ads Help). And according to Google, AI Max won’t be effective if campaigns are limited by budget (Google Ads Help). Reliable conversion measurement therefore belongs before the switch; how to achieve it despite consent banners is described in our post on server-side tracking for Google Ads.

Test instead of waiting: the switch as an experiment

Google explicitly recommends not waiting for the automatic upgrade and provides tools for it (Google Ads & Commerce Blog). To compare an existing campaign with Dynamic Search Ads against AI Max, Google names two methods (Google Ads Help):

1. Custom experiment: in the draft of the test campaign, select the campaign upgrade in the Dynamic Search Ads settings. If the trial arm wins, apply the result to the original campaign or convert the trial arm into a new campaign (Google Ads Help).
2. AI Max experiment template: tests within the existing campaign instead of a copy, but is not available for campaigns made up entirely of dynamic ad groups (Google Ads Help). In mixed campaigns, the trial arm upgrades only the standard ad groups to AI Max, and the dynamic ad groups stay dynamic (Google Ads Help).

The experiment template has further limits. It cannot be created if the campaign already uses text customization, targets the Display Network, belongs to a portfolio bid strategy, uses a shared budget, uses bidding exploration or already has an active experiment. Except for the Display Network, these limits also apply to custom experiments (Google Ads Help). Eligible accounts may find a one-click experiment on the Recommendations page that runs the base campaign in a 50/50 split test against an AI Max arm (Google Ads Help).

How much a switch brings can only be measured in your own account. Google cites averages: advertisers that activate AI Max typically see 14 percent more conversions or conversion value at a similar CPA or ROAS (Google Ads Help). With all three features, the average is 7 percent more than with search term matching alone (Google Ads & Commerce Blog). Both figures come from Google internal data from 2025 and 2026 and explicitly apply to non-retail advertisers. For an online store, they are therefore not a forecast but one more reason to measure yourself.

Control after the switch: brands, locations, URLs and text

After the switch, the work shifts from maintaining dynamic targeting to maintaining guardrails. Negative keywords stay in effect even when AI Max is turned on (Google Ads Help). On top of that, there are four controls merchants should know:

Text guidelines are the direct lever against the risks from the section on data. Words like "free" are useful term exclusions when shipping costs apply; discount and delivery promises can be ruled out through messaging restrictions. Before you turn the feature on, a preview shows up to 10 example assets for a landing page, even without active text customization (Google Ads Help).

If Performance Max runs in the same account, both campaign types compete for the same searches. If a search can trigger both a Search campaign with AI Max and Performance Max, the higher Ad Rank decides (Google Ads Help). More on this in our guide to Performance Max.

Roadmap to February 2027 with the holiday season in mind

The deadline falls in a quieter period, the road to it does not: Black Week and the holiday season come first, and Black Friday is on November 27, 2026. Rebuilding high-revenue campaigns during these weeks is risky, because learning phases and promotional effects cannot be separated. A roadmap that takes this into account:

1. By mid-November: complete the inventory, check landing pages and page titles, prepare a page feed with labels.
2. Black Week and December: do not switch high-revenue campaigns, save search terms reports as a baseline, test text guidelines in the preview.
3. Early January: start experiments for high-revenue campaigns, switch low-revenue campaigns directly with the upgrade tool.
4. By the end of January: evaluate and apply experiments. Do not plan any new structures with Dynamic Search Ads, because according to the June developer post, creation ends in January (Google Ads Developer Blog).
5. February: Google automatically upgrades the remaining campaigns (Google Ads Help). Afterwards, check settings and reports.

In February 2027, Google also changes Display Expansion in Search campaigns: campaigns with Display Expansion enabled automatically migrate to channel prioritization and budget allocation (Google Ads Help). If you are working on your Search campaigns anyway, plan both changes in one pass; the details are in our post on channel prioritization in Google Ads.

Adapting APIs and custom tools

If you create campaigns through the Google Ads API, scripts or custom tools, there is extra work. Since August 3, 2026, campaign-level broad match settings and automatically created assets can no longer be newly created in the interface, in Editor or through any version of the API (Google Ads Developer Blog). Older API versions still allow editing such legacy setups until their scheduled sunset, which Google expects around September 2027 (Google Ads Developer Blog). Scripts and tools that still create such legacy setups therefore need an audit (Google Ads Developer Blog).

One difference between interface and API deserves attention. In the interface, text customization and final URL expansion are enabled by default when AI Max is turned on (Google Ads Help). Through the API, however, ai_max_setting.enable_ai_max only turns on search term matching; both asset features initially stay at OPTED_OUT there (Google Ads API). If you create campaigns through software, specify the features explicitly; such adjustments are part of our custom programming.

How we support the switch to AI Max

The switch is a small data migration. We take on the parts that involve technology and store data, together with you or your campaign team.

If you are also working on product data and feeds, we combine the work with the connection to Google Merchant Center. For a first assessment, a short message through our contact form is enough.