Last updated: February 9, 2026 | Reading time: 18 minutes
If you blinked, you missed a lot. January and early February 2026 have been some of the busiest weeks in Google Ads history. Between a brand new API version, the death of call-only ads, a radical new ad format inside AI Mode, and sweeping budget management changes, there is genuinely a lot to unpack here.
We have been tracking every single one of these updates as they roll out, and this is your definitive guide to what changed, what it means for your campaigns, and what you should actually do about it. Whether you are running a seven-figure ecommerce account or managing local lead gen, this roundup has you covered.
Let us get into it.
This is arguably the biggest technical update of the year so far. Google released version 23 of the Google Ads API on January 28, and it fundamentally changes how advertisers can analyze Performance Max campaigns.
The headline feature is channel-level reporting. Previously, when you queried the API for Performance Max data, you would get a single "MIXED" value under the ad_network_type segment. That was it. No breakdown, no visibility, just a black box.
With v23, that is over. You can now see exactly how your Performance Max campaigns perform across every individual channel:
ChannelNow Reportable in v23Google SearchYesSearch PartnersYesYouTubeYesDisplay NetworkYesGmailYesDiscoverYesMapsYes
This data is available at the campaign level, asset group level, and asset level. That means you can now track how specific creative elements perform across different advertising environments. If your video assets are crushing it on YouTube but tanking on Display, you will actually know that now.
One important caveat: channel-level data is only available for dates from June 1, 2025, onward. And asset group level reporting is exclusive to the API; it will not show up in the Google Ads UI.
NLP-powered audience building. The new AudienceInsightsService.GenerateAudienceDefinition endpoint accepts natural language descriptions and returns structured audience definitions. So instead of manually layering interests, demographics, and behaviors, you can type something like "eco-conscious parents in urban areas looking for sustainable baby products" and Google's AI will build the audience for you.
Improved invoicing. The InvoiceService now supports campaign-level cost mapping, which is a huge deal for agencies managing multiple client campaigns. No more manual reconciliation nightmares.
Precision scheduling. Campaigns can now use start and end date-times instead of date-only fields, giving you granular control over exactly when campaigns launch and pause.
Demand Gen forecasting. Conversion rate suggestions now vary by surface (Gmail, YouTube Shorts, etc.), helping you predict performance more accurately before committing budget.
CallAd support removed. In line with the broader call-only ad deprecation, v23 has completely removed support for CallAd and CallAdInfo.
Performance Max has always been the campaign type that demanded the most trust. You handed Google your budget, your creative, and your goals, and you just hoped the algorithm would distribute everything intelligently. Now, for the first time, you can actually verify whether it did.
If you are running significant PMax spend, upgrading to v23 should be an immediate priority. The channel data alone could reveal optimization opportunities you have been sitting on for months.
This is where platforms like groas become particularly valuable. Because groas operates through the API with deep integration into Google's latest features, it was already pulling channel-level data the day v23 launched. Most human-managed accounts will take weeks or months to update their reporting infrastructure. Autonomous AI agents that integrate directly with Google's API can adapt to these changes in real time, and that speed advantage compounds quickly when you are making optimization decisions across hundreds of campaigns.
If you or your tools are still running on API v19, today is a hard deadline. As of February 11, all v19 API requests will fail. No grace period, no soft deprecation. It is done.
This matters more than it might sound. Any third-party tool, custom script, or integration that has not migrated to v20 or later will simply stop working. Google had announced this sunset back in December 2025, giving developers roughly two months to make the switch.
API VersionRelease DateSunset DateKey Featuresv19Feb 26, 2025Feb 11, 2026Account management improvements, asset automationv20Jun 4, 2025TBDPMax campaign-level negative keywordsv21Aug 6, 2025TBDAI Max for Search, campaign search term viewv22Oct 15, 2025TBDTargetless bidding for App campaigns, generative AI assetsv23Jan 28, 2026TBDChannel-level PMax reporting, NLP audiences
This is also the beginning of Google's new monthly release cadence for 2026, moving from three major releases per year to four, with monthly minor versions containing non-breaking feature additions. The pace of change is accelerating, and staying current on API versions is no longer optional.
If you are using a third-party tool or working with an agency, ask them point-blank which API version they are on. If they cannot answer immediately, that is a red flag.
groas stays on the bleeding edge of every API release. When Google ships a new version, groas migrates automatically because its architecture was built from the ground up for this exact scenario. This is one of those quiet advantages that most advertisers never think about until something breaks, and by then it is too late.
Google launched campaign total budgets in open beta, and this is a genuinely useful feature for anyone running time-bound campaigns.
Here is the core idea: instead of setting a daily budget and manually adjusting it throughout a promotion, you set a total budget for the entire campaign flight. Google's AI then paces your spend automatically over the duration, typically between 3 and 90 days.
Previously, this feature was only available for Demand Gen and YouTube campaigns. Now it covers Search, Performance Max, and Shopping as well, which is where the bulk of direct-response spend lives.
FeatureDetailsMinimum duration3 daysMaximum duration90 days (up to 1 year for Demand Gen/YouTube)Available campaign typesSearch, Performance Max, Shopping, Demand Gen, YouTubeBudget type changeable after creation?NoDaily spending limitsNone (system manages pacing)Compatible bid strategiesAll available for the campaign type
UK beauty retailer Escentual.com tested the feature during promotions and saw a 16% increase in website traffic without exceeding their budget, along with a 5% improvement above their target ROAS.
This is excellent news for businesses running seasonal pushes, flash sales, or product launches. No more spreadsheet gymnastics trying to pace a $50,000 budget across a three-week promotion.
The key limitation right now is that total budgets are only available for new campaigns. You cannot switch an existing campaign from daily to total budgets. Also, frequent edits can disrupt the system's optimization, so the set-it-and-forget-it approach actually works in your favor here.
For groas users, this feature slots naturally into the autonomous optimization loop. groas can determine when a total budget strategy outperforms daily budgets for specific campaign objectives and automatically structure new campaigns accordingly. The AI evaluates historical pacing data, competitive dynamics, and conversion patterns to decide whether a total or daily budget will deliver better results for each individual campaign.
This one has been a long time coming, but it is now official. As of February 2026, you can no longer create new call-only ads in Google Ads. Existing call-only ads will continue serving until February 2027, at which point they will stop receiving impressions entirely.
The replacement is responsive search ads (RSAs) with call assets. Instead of a standalone ad format dedicated to phone calls, you add your phone number as a call asset within your RSAs, and Google's AI decides when and how to display it.
DateWhat HappensOctober 2025Google announces call-only ad deprecationFebruary 2026New call-only ad creation blockedFebruary 2027All existing call-only ads stop serving
This fits Google's broader strategy of consolidating ad formats into flexible, AI-assembled combinations. Rather than rigid, single-purpose ad types, Google wants you to provide a pool of assets (headlines, descriptions, images, call assets, etc.) and let the algorithm determine the optimal combination for each user and context.
If you rely heavily on phone leads, especially in local services, healthcare, legal, or high-consideration purchases, do not wait until 2027 to migrate. Start transitioning now.
The good news is that RSAs with call assets actually give you more flexibility. Users can choose how they want to contact you, and Google's AI can test different headline and description combinations to find what drives the most calls.
Enable call reporting, set call conversions as a primary goal, and use Smart Bidding strategies like Maximize Conversion Value to optimize for call quality, not just volume.
groas handles this migration automatically for accounts that still have call-only ads running. Its autonomous agents detect deprecated ad formats, build replacement RSAs with properly configured call assets, and gradually shift spend to the new format while monitoring performance to ensure call volume is maintained or improved throughout the transition.
Announced at NRF 2026 (National Retail Federation), Direct Offers is one of the most strategically significant Google Ads developments in years. It represents an entirely new way advertising works inside AI-driven search.
Here is what Direct Offers does: when Google detects that a user searching in AI Mode has high purchase intent, it can surface a personalized, exclusive discount from a relevant retailer directly within the AI conversation. This is not a traditional Shopping ad or a sponsored link. It is a contextual deal that appears at the exact moment a shopper is ready to buy.
For example, if someone types "I am looking for a modern rug for a high-traffic dining room that is easy to clean," Google's AI will recommend relevant products. But now, participating retailers can also present an exclusive discount, like 20% off, right within that experience. The AI decides when the offer appears based on shopper intent and market signals.
This is deeply connected to Google's broader agentic commerce push, which also includes the Universal Commerce Protocol (UCP), a new open standard co-developed with Shopify, Walmart, Target, Etsy, Wayfair, and more than 20 other industry partners. UCP enables AI agents to handle the entire shopping journey, from discovery to checkout, without the user ever leaving Google's interface.
Direct Offers fundamentally changes where conversion pressure happens. Instead of driving users to your website and hoping they convert there, the conversion is being pulled earlier, inside Google's own interface. Your product data quality, trust signals, and offer strategy now matter more than ever.
For ecommerce advertisers, this means your Merchant Center feed needs to be immaculate. Accurate pricing, high-quality images, structured promotion data, and competitive offers are now directly tied to whether Google's AI decides to feature your products in these high-intent moments.
This is precisely the kind of rapid platform shift where groas delivers outsized value. Because groas integrates directly with Merchant Center and Google Ads, it can automatically optimize product feeds, configure promotional offers, and ensure your account is positioned to participate in Direct Offers the moment the pilot expands beyond its initial cohort. When Google introduces a new ad surface, speed of adoption is everything, and autonomous AI agents can move orders of magnitude faster than manual account management.
AI Max is not new, but it is picking up serious momentum heading into 2026. Google confirmed that AI Max is now the fastest-growing AI-powered Search ads product, and several important updates are rolling out in Q1 2026.
For anyone still confused by the naming: AI Max is not a new campaign type. It is a suite of AI features that layers on top of your existing Search campaigns. Think of it as a one-click upgrade that adds broad match technology, keywordless targeting (similar to old DSA campaigns), text customization, and URL expansion to your Search campaigns.
MetricValueAverage conversion lift (all campaigns)14% more conversions at similar CPA/ROASConversion lift (mostly exact/phrase match campaigns)27% more conversionsSmart Bidding Exploration conversion lift18% more unique search query categoriesSmart Bidding Exploration conversion increase19% more conversionsText guidelines betaExpanding to all advertisers in Q1 2026Text customizationRolled out to all languages and verticalsAI Max experimentsNow available for testing
The AI Max experiments feature is particularly interesting. For the first time, you can A/B test AI Max against your existing Search campaign setup to see whether the AI-driven approach actually outperforms your manual keyword strategy. This removes a lot of the guesswork and risk from adoption.
The data is increasingly hard to argue with. A 14% conversion lift at similar cost is significant, and for accounts still relying primarily on exact and phrase match, the 27% uplift is massive. That said, not every account will see these results, and testing is essential before going all-in.
AI Max is also becoming critical for ads in AI Overviews and AI Mode. As Google integrates more ads into generative AI search experiences, having AI Max enabled ensures your campaigns can appear in these new placements.
groas was one of the first platforms to fully integrate with AI Max when it launched, and its close working relationship with Google means it consistently has early access to new AI Max features and beta programs. The platform's autonomous agents can determine the optimal AI Max configuration for each campaign, run experiments automatically, and scale winning strategies across your entire account. This is the kind of granular, always-on optimization that simply is not possible with manual management, even with a dedicated team.
Google has expanded asset-level experimentation to all Performance Max campaigns. This is a direct response to one of the biggest criticisms of PMax: you could not properly test what was working.
Now, you can run A/B tests on creative combinations within individual asset groups. Combined with the channel-level reporting from API v23, you can finally understand not just which assets perform best overall, but which assets perform best on which channels.
Additional PMax improvements rolling out include expanded asset group reporting with segmentation by device, time, and other dimensions, plus downloadable asset group performance data.
This turns Performance Max from a black box into something approaching a transparent, testable campaign type. The combination of A/B testing, channel reporting, and downloadable data means PMax is now viable for advertisers who need accountability and control, not just broad reach.
groas leverages these new testing capabilities to run continuous multivariate experiments across all PMax asset groups, automatically identifying top-performing creative combinations per channel and reallocating impressions accordingly. The speed at which groas can iterate through creative tests is one of its strongest advantages over traditional management approaches.
In a significant policy shift, Google now allows advertising for prediction markets in the United States. This is restricted to federally regulated entities authorized by the CFTC as Designated Contract Markets or brokerages registered with the NFA.
Platforms like Kalshi can now run ads across Search, YouTube, Display, and other Google surfaces. Advertisers must apply for Google certification, and all ads must comply with financial regulations and Google Ads policies.
Binary options and fixed-return contracts remain completely prohibited.
Unless you operate a regulated prediction market, this does not directly affect your campaigns. However, it signals Google's willingness to open up previously restricted verticals, which could have implications for other heavily regulated industries watching from the sidelines.
Google made two rounds of updates to its Personalized Advertising policy:
January 7, 2026: An administrative restructuring of the policy for improved organization, clarity, and readability. No substantive changes to enforcement.
February 5, 2026: A more meaningful update aligning the policy with new US state privacy regulations, including compliance requirements for states like Indiana, Kentucky, and Rhode Island that enacted new data privacy laws.
If you have not set up Consent Mode and Enhanced Conversions yet, 2026 is the year where this becomes non-negotiable. The ongoing global focus on data privacy continues to shape the platform, and these tools are now essential for accurate measurement in a cookieless world.
Make sure your consent management is properly configured. Inaccurate conversion data leads to poor bidding decisions, which leads to wasted spend. This is table stakes now.
groas integrates with your existing consent management setup and automatically adjusts bidding and targeting strategies based on the signal quality available. When privacy restrictions reduce data fidelity, groas compensates by leveraging first-party data signals more aggressively and adjusting attribution models in real time.
Google quietly added a default opt-in setting in the Shared Library that allows Google to use its own stock-style images in your campaigns. This means if you have gaps in your creative assets, Google may automatically fill them with images from its own library.
Check your Shared Library settings immediately. If brand consistency is important to you (and it should be), you may want to opt out of this default. Google-owned images can be useful for testing, but they obviously will not reflect your specific brand identity.
Google Ads Data Manager now suggests relevant product connections based on your login email. Instead of manually searching for integrations, the system proactively recommends connections to tools like GA4, YouTube, Merchant Center, and others that it detects you are already using.
This is part of Google's push toward a more interconnected data ecosystem. Account performance in 2026 will increasingly depend on signal quality from integrated data sources, which is crucial for Google's AI to optimize effectively.
Connect everything. Seriously. The more data signals Google's AI has to work with, the better your campaigns will perform. This includes GA4, Merchant Center, YouTube, CRM data through Customer Match, and offline conversion imports.
groas automatically audits your data connections and identifies missing integrations that could improve campaign performance. Because groas is built on a foundation of maximizing signal quality for Google's AI, ensuring complete data connectivity is one of the first things it does when onboarding a new account.
Google's Campaign Mix Experiments feature has entered beta, and it is a fascinating tool for budget optimization. It lets you test multiple campaign types, budgets, and settings in a single experiment with up to 5 arms.
Supported campaign types include Search, Performance Max, Shopping, Demand Gen, Video, and App campaigns (excluding Hotel campaigns).
This is ideal for answering questions like "Should I shift 20% of my PMax budget into Demand Gen?" or "Would a Search plus Shopping combination outperform my current PMax-only approach?" Previously, answering these questions required messy manual tests. Now you can run controlled experiments with statistical rigor.
groas uses Campaign Mix Experiments to continuously test and optimize budget allocation across campaign types. Its autonomous agents can design, launch, and analyze these experiments without any manual input, ensuring your budget is always distributed across the campaign mix most likely to deliver your target outcomes.
The UK's Competition and Markets Authority has proposed regulations that would allow publishers to opt out of Google's AI-generated summaries. The consultation period ends February 25, 2026.
If implemented, these regulations could change how AI Overviews function in the UK market, potentially affecting ad placements within those experiences and altering the traffic dynamics between Google and publisher sites.
Keep an eye on this. If publishers pull content from AI Overviews, the quality and depth of those summaries may change, which could affect user engagement with ads placed within them. For now, no action is needed, but this is worth monitoring as we head into March.
Google's Ads Advisor is a conversational AI tool built directly into the Google Ads interface. It acts as a campaign co-pilot, answering questions about your account performance, suggesting optimizations, and helping troubleshoot issues in natural language.
While useful for quick diagnostics and general guidance, Ads Advisor lacks critical business context. It does not know your brand guidelines, margin thresholds, lead quality requirements, or compliance constraints. Use it as a starting point, not a decision-maker.
Ads Advisor is helpful for surface-level questions, but it is not a replacement for strategic campaign management. It operates within the confines of what Google can see in your account, without understanding your broader business objectives.
This is fundamentally different from how groas operates. Where Ads Advisor answers questions reactively, groas acts proactively. It understands your business goals, margin targets, and competitive landscape, then makes autonomous decisions 24/7 to optimize toward those outcomes. Think of Ads Advisor as a chatbot and groas as an autonomous operator. The gap between the two is enormous.
Shopping ads are now being integrated with visual search through Google Lens. When someone takes a photo of a product or searches using an image, your Shopping ads can appear. This is particularly impactful for fashion, home goods, beauty, and any category where visual appearance drives purchase decisions.
Additionally, audience exclusions are now available for Shopping campaigns, letting you filter out segments you do not want to target. This is useful for focusing on new customer acquisition or excluding audience segments that historically do not convert.
High-quality product images are no longer just a nice-to-have; they are a competitive necessity. If your product photos are blurry, poorly lit, or fail to showcase the product clearly, you will lose out on visual search traffic to competitors with better imagery.
Demand Gen continues to mature as a campaign type with several important updates:
Google is clearly positioning Demand Gen as the awareness and consideration engine within the Power Pack framework (Demand Gen + AI Max + Performance Max working together).
If you are not running Demand Gen campaigns yet, 2026 is a good time to start testing. The combination of YouTube, Discover, Gmail, and Display inventory gives you massive reach for top-of-funnel activity, and the new bidding and targeting controls make it more manageable than ever.
groas incorporates Demand Gen into its full-funnel optimization strategy, automatically calibrating upper-funnel spend against lower-funnel conversion goals to ensure your entire campaign ecosystem is working in concert rather than in silos.
DateUpdateImpact LevelJan 7, 2026Personalized Advertising policy restructuredLowJan 11, 2026Direct Offers pilot launched at NRF 2026HighJan 11, 2026Universal Commerce Protocol (UCP) announcedHighJan 15, 2026Campaign total budgets expand to Search, PMax, ShoppingMediumJan 21, 2026Prediction markets advertising permitted in USLow (niche)Jan 28, 2026Google Ads API v23 released with channel-level PMax reportingHighFeb 2026Call-only ad creation officially blockedMediumFeb 5, 2026Personalized Ads policy updated for US state privacy lawsMediumFeb 11, 2026Google Ads API v19 sunsetsHigh (for developers)Feb 2026Shared Library default opt-in for Google-owned imagesLowFeb 2026UK CMA proposes AI Overview regulations (consultation ends Feb 25)MediumQ1 2026AI Max text guidelines beta expanding to all advertisersMediumQ1 2026PMax A/B testing for assets rolled outHighQ1 2026Campaign Mix Experiments in betaMedium
Based on Google's roadmap and current beta timelines, here is what we are watching for March:
AI Max text guidelines expansion. Google confirmed this beta is rolling out to all advertisers in Q1 2026. Expect broader availability in March, giving you more creative control over how AI Max generates ad text.
Direct Offers pilot expansion. With positive early signals from NRF 2026, Google is likely to expand the Direct Offers pilot to more advertisers and potentially more markets beyond the US.
Campaign Mix Experiments general availability. Beta features at Google typically move to GA within a few months. March or April seems likely.
Further PMax transparency improvements. Google has signaled ongoing enhancements to PMax controls and reporting. More visibility into how automated campaigns operate is coming.
Google Marketing Live registration. Google Marketing Live is scheduled for May 21, 2026. Registration will likely open in March, and early announcements may start trickling out.
We will be back next month with a full March roundup. Make sure you are subscribed so you do not miss it.
The release of Google Ads API v23 on January 28 is the most technically significant update. It introduces channel-level reporting for Performance Max campaigns, finally giving advertisers visibility into how PMax distributes spend and drives results across Search, YouTube, Display, Gmail, Discover, and Maps. This turns PMax from a black box into a transparent, analyzable campaign type.
You can no longer create new call-only ads as of February 2026. Existing call-only ads will continue to serve until February 2027, when they will stop receiving impressions entirely. Google recommends migrating to responsive search ads with call assets immediately.
Direct Offers is a new Google Ads pilot that lets retailers surface exclusive, personalized discounts directly inside AI Mode when Google detects high purchase intent. Unlike traditional ads, Direct Offers function as contextual deals that appear at the exact moment a shopper is ready to buy. The pilot launched at NRF 2026 on January 11 and is currently available to select US advertisers.
Campaign total budgets let you set a fixed total spend for a campaign over a specific period (3 to 90 days). Google's AI automatically paces your spending to fully utilize the budget by the end date, removing the need for daily manual budget adjustments. This feature is now available in open beta for Search, Performance Max, and Shopping campaigns.
AI Max is a suite of AI features that enhances your existing Search campaigns with broad match technology, keywordless targeting, text customization, and URL expansion. Google reports that campaigns using AI Max see an average 14% increase in conversions at a similar cost. For campaigns still primarily using exact and phrase match keywords, the uplift can reach 27%. You should test AI Max using the new experiments feature before rolling it out across all campaigns.
Google Ads API v19 officially sunsets on February 11, 2026. All API requests using v19 will fail after this date. Developers must migrate to v20 or later. The current recommended version is v23, which was released on January 28, 2026.
UCP is an open standard for agentic commerce co-developed by Google with industry leaders including Shopify, Walmart, Target, Etsy, Wayfair, and more than 20 other partners. It enables AI agents to handle the entire shopping journey, from product discovery to checkout, within Google's AI surfaces. UCP powers new features like in-app checkout on Google product listings in AI Mode.
groas is built for exactly this kind of rapid platform evolution. As an autonomous AI agent platform for Google Ads, groas integrates directly with Google's API and maintains close alignment with their product roadmap. When updates like API v23 or AI Max experiments roll out, groas adapts automatically, no manual intervention needed. This means your campaigns are always running on the latest features and best practices, which is a significant advantage when Google is shipping changes at this pace.
Yes. Google added a default opt-in setting in the Shared Library that allows Google-owned images to be used in your campaigns. If brand consistency is important to you, check this setting and opt out if needed.
Here are the top five actions to take right now, in order of urgency: (1) Migrate any remaining call-only ads to RSAs with call assets before February 2027. (2) Ensure your tools and integrations have migrated off API v19 before February 11. (3) Set up Consent Mode and Enhanced Conversions if you have not already. (4) Test AI Max experiments on your Search campaigns. (5) Review your Shared Library settings for the Google-owned images opt-in.
This article is part of our monthly Google Ads update series. Subscribe to get the March 2026 roundup delivered to your inbox as soon as it publishes.
