Key takeaways
- Google has delayed the automatic Dynamic Search Ads migration to AI Max from September 2026 to February 2027.
- The delay matters only if advertisers use it to run controlled DSA versus AI Max tests before forced migration.
- AI Max gives Google wider matching and automation freedom, so poor conversion tracking will make bad optimisation faster.
- UK lead generation accounts are at particular risk where low-quality form fills, short calls and sales-qualified leads are mixed together.
- Advertisers should export DSA baselines, clean conversion goals, segment page intent and build negative controls before January 2027.
The DSA AI Max delay is not Google being generous. It is Google recognising that too many advertisers were being pushed towards a forced migration before their accounts were ready. For UK advertisers, the extra time only matters if it is used to build a controlled test, clean up conversion data and stop AI Max inheriting bad signals from old Dynamic Search Ads campaigns.
DSA has always been useful when websites contain more commercial intent than keyword lists capture. AI Max changes that equation because it gives Google more freedom over query matching, creative generation and landing page selection. That matters in accounts already affected by automation pressure, especially where budget-limited bidding changes have already squeezed decision-making room.
Our view is simple: do not treat the delay as breathing space. Treat it as a deadline extension for evidence gathering. The advertisers who run clean tests now will move on their own terms. The advertisers who wait will inherit a February 2027 migration with weak benchmarks and messy reporting.
What’s actually changed in the DSA AI Max timetable
Google has pushed back the automatic migration of Dynamic Search Ads to AI Max for Search campaigns from September 2026 to February 2027. Advertisers also regain the ability to create new DSA campaigns from 15 June 2026, reversing the earlier move towards phasing them out.
The practical timeline is clear. New DSA creation returns in June 2026. Advertisers then have a testing and voluntary migration window through January 2027. New DSA creation ends in January 2027. Remaining DSA campaigns start moving automatically in February 2027.
Two related changes still matter. Automatically created assets and the campaign-level broad match setting continue moving towards AI Max in September 2026. Google has also made AI Max the default setting when creating new Search campaigns, which tells you where product strategy is heading even though the DSA deadline moved.
Why the DSA AI Max delay matters for advertisers
The money moves because DSA and AI Max do not find demand in the same way. DSA primarily uses your website content to match searches to relevant landing pages. AI Max sits inside Search and combines broader query expansion, asset automation and Smart Bidding logic. That shift changes what Google is allowed to test on your behalf.
In a well-structured account, that extra reach creates useful discovery. In a weak account, it turns existing problems into faster spend. If your website has mixed intent pages, thin service pages or unclear conversion paths, AI Max receives more permission to chase queries that look semantically relevant but have poor commercial value. The campaign does not fail loudly. It spends, learns from poor conversions, then optimises towards more of the same.
Lead generation advertisers face the sharpest risk. A DSA campaign for a solicitor, accountant or managed IT provider often works because the website contains detailed service pages that match long-tail demand. Move that same activity into AI Max without tight conversion definitions and Google starts valuing whichever enquiry is easiest to get. That usually means junior queries, research intent, support requests and price-led prospects.
This is where tracking quality decides whether automation helps or damages you. PPC Geeks tracking-health probe, Q1 2026 audit data, shows at least 56% of active UK accounts have a conversion-tracking fault serious enough to distort the numbers they optimise on. It is a floor, not a ceiling: the Google Ads API cannot see Consent Mode, web Enhanced Conversions or tag-firing errors, so the true rate is higher.
That statistic matters more for AI Max than for old DSA. DSA can still waste money, but its behaviour is easier to inspect through search terms, landing pages and obvious page matching. AI Max pushes more decisions into automation. If the imported conversion data is polluted, Smart Bidding treats pollution as truth. It raises bids for low-quality signals, reduces spend where real value sits and makes the account look sophisticated whilst the economics get worse.
Migration also changes your benchmark problem. If you wait until forced migration, you compare February 2027 AI Max performance against old DSA performance under different auction conditions, budgets and seasonality. That comparison is weak. A side-by-side test in 2026 gives you a cleaner read on CPA, lead quality, landing page behaviour and search term expansion. That is why this delay has value. It gives advertisers time to create evidence before the platform removes the old control.
PPC Geeks’ View
The specific problem advertisers will face is contaminated learning. AI Max will be asked to improve campaigns using historical conversion signals that were never clean enough for that level of automation. We see this most often in lead-gen accounts running broad match with Smart Bidding, especially where form fills, phone calls and offline sales outcomes are all treated as equal conversions.
Do not migrate DSA into AI Max until you know which conversions deserve bid weight. Automation does not fix bad measurement. It scales it.
Dan Trotter, Head of PPC, PPC Geeks
The practical pattern is consistent. A campaign shows strong conversion volume, but sales says the leads are poor. Dig into the account and the conversion set includes newsletter sign-ups, short calls, duplicate forms or quote requests from areas the business does not serve. DSA made that uncomfortable. AI Max makes it expensive because the bidding system reacts faster and with wider matching freedom.
This is exactly the type of issue we look for in a free Google Ads audit, especially where automation, tracking or campaign structure is affecting performance. Before any DSA AI Max migration, the first job is not choosing a new campaign type. The first job is deciding which conversion actions earn budget.
Advertisers also need to stop thinking about DSA as just an old campaign format. For many UK SMEs, DSA is a cheap discovery layer for overlooked commercial search demand. Losing it without a replacement plan creates a query coverage gap. AI Max can fill that gap, but only if landing pages, exclusions, goals and value signals are built deliberately. Otherwise the campaign expands into noise and the account manager spends the next quarter explaining why lead volume rose while sales quality fell.
If your reporting already struggles to connect paid search spend to real pipeline, use this migration to fix the measurement model. Our cross platform measurement framework sets out the same principle: paid media optimisation only works when the data has commercial meaning.
What advertisers should do next with DSA AI Max
1. Export your current DSA baseline this week. Pull the last 90 days of DSA performance by search term, landing page, device, location and conversion action. Add lead quality fields from your CRM if you have them. Mark which landing pages produced revenue or qualified pipeline, not just enquiries. This becomes the control group for every AI Max decision.
2. Split DSA coverage by page intent before testing. Do not test one broad DSA campaign against one broad AI Max campaign. Build segments around page purpose: high-margin services, local service pages, product categories, support content and blog content. Exclude non-commercial sections before the test starts. AI Max should not learn from pages that sales would never want traffic from.
3. Run a 30-day side-by-side test with budget caps. Keep your existing DSA campaign live and launch AI Max against a defined page set or equivalent keyword theme. Cap spend tightly enough that a bad learning phase cannot consume the month. Compare cost per qualified lead, not account-level CPA. If your sales cycle is longer than 30 days, import offline lead stages before judging the test.
4. Remove weak conversion actions from bidding. In Google Ads, open Goals, Conversions, Summary and check every action included in account-default goals. Take newsletter sign-ups, short calls, accidental page views and soft engagement events out of bidding. Keep them as secondary actions if reporting still needs them. AI Max should optimise for business outcomes, not dashboard activity.
5. Build negative controls before migration. Create shared negative lists for recruitment queries, free advice, jobs, complaints, support, existing customer logins and irrelevant locations. Add brand exclusions where AI Max risks inflating results with demand you would have captured anyway. This becomes more important as Google expands matching beyond the old DSA model.
6. Document your migration decision before January 2027. Write down which DSA campaigns will migrate voluntarily, which will be rebuilt as standard Search with broad match and Smart Bidding, and which should be retired. Assign each one an owner, a test date and a success metric. If you need outside help, our Google Ads agency team can interrogate the structure before the forced timeline arrives.
7. Use the confirmed timeline as a planning constraint. The Search Engine Land report confirms the February 2027 automatic migration date and the June 2026 return of DSA creation. Put those dates into your testing calendar now. A migration plan that starts in January is not a plan. It is a scramble.
What this means for your campaigns
The DSA AI Max delay gives advertisers one valuable asset: a proper comparison window. Use it badly and nothing changes. Use it well and you enter 2027 knowing which pages deserve automation, which conversions deserve bid weight and which parts of your account should never be handed more matching freedom.
Google has made its direction clear. Search is becoming more automated, more AI-led and more dependent on the quality of the signals you feed it. Fighting that direction is a waste of budget. Accepting it without controls is worse. The right response is disciplined testing, cleaner tracking and a migration plan that protects commercial intent.
The advertisers who win this transition will not be the ones who click upgrade first. They will be the ones who make AI Max prove itself against DSA before the old option disappears.
Want a no-nonsense view of what to change first? Start with a free Google Ads audit from our team.
Frequently asked questions
What is changing with Dynamic Search Ads and AI Max?
Google has delayed the automatic migration of Dynamic Search Ads to AI Max for Search campaigns until February 2027. Advertisers also regain the ability to create new DSA campaigns from 15 June 2026, with new DSA creation ending in January 2027.
Should advertisers stop using Dynamic Search Ads now?
No. Advertisers should keep useful DSA campaigns live while building controlled AI Max tests. DSA still provides a valuable benchmark for search demand, landing page coverage and long-tail query performance.
Why is conversion tracking so important for AI Max?
AI Max relies heavily on Smart Bidding and automated matching. If weak actions such as short calls, newsletter sign-ups or poor-quality forms are included in bidding, the system optimises towards those actions and spends more on lower-value traffic.
How should UK advertisers test AI Max against DSA?
Run a side-by-side test with clear page segments, capped budgets and qualified lead metrics. Compare cost per qualified lead or revenue, not just total conversions, and keep DSA running as the control group during the test.
When should advertisers finalise their migration plan?
Advertisers should have a documented plan before January 2027. That plan should name which DSA campaigns migrate voluntarily, which are rebuilt as standard Search campaigns, and which are retired.
