Will Microsoft really make a dent in Google’s advertising empire?

The launch (or technically relaunch from a long time ago) of Microsoft’s PubCenter, a potential rival to Google’s AdSense, marks a currently small but likely significant moment in the online advertising industry. This move signals Microsoft’s ambition to carve out an increasing share of the digital ad market, a domain long dominated by Google. Exactly how PubCenter stacks up against AdSense remains to be seen, but Microsoft is already claiming it will be offering publishers higher engagement and likely greater revenues than Adsense, a bold claim.

At its core, PubCenter aims to provide a platform for publishers to monetize their websites through advertising, much like Google’s AdSense. However, Microsoft might bring a fresh perspective and innovative features to the table. These could include advanced targeting options, leveraging its extensive data from platforms like Bing and LinkedIn, and potentially offering better revenue share terms or more diverse ad formats. If PubCenter’s user interface and algorithm for ad placement and revenue generation are on par or surpass those of AdSense, it could entice publishers to make the switch.

For publishers and advertisers, the introduction of PubCenter is likely to be beneficial. Competition fosters innovation, improved service, and more favourable terms. Publishers might find themselves with more lucrative or flexible monetization options, while advertisers could benefit from more competitive pricing and potentially less saturated advertising spaces.

Regarding the impact on Google’s revenue, it’s important to consider the scale and entrenchment of AdSense in the market. While PubCenter might chip away at Google’s market share, it’s unlikely to cause a significant dent in the short term. Google’s advertising ecosystem is deeply integrated with its other services like Search, YouTube, and Google Analytics, creating a robust and comprehensive offering that’s hard to replicate. Nonetheless, even a small percentage shift represents a substantial amount in an industry worth billions.

Beyond advertising, Microsoft and Google have been head-to-head in several other arenas. One notable area is cloud computing, where Microsoft’s Azure directly competes with Google Cloud. Though Amazon Web Services leads this market, both Microsoft and Google have been aggressively expanding their cloud offerings.
In the realm of productivity software, Microsoft’s Office suite still holds significant dominance over Google Workspace, though the latter has been gaining traction, especially among startups and educational institutions. Moreover, in the search engine market, Bing continues to trail behind Google Search, but Microsoft has been making consistent efforts to improve it.

Microsoft’s launch of PubCenter as a competitor to Google’s AdSense is a bold move that could reshape the online advertising landscape. While it may not significantly impact Google’s revenue in the immediate future, it introduces healthy competition that could lead to better offerings for both publishers and advertisers. This development is part of a larger trend of competition between these tech giants, extending across various domains from cloud computing to productivity tools. As the landscape evolves, it’s clear that Microsoft is not shying away from challenging Google in key areas, signalling a more competitive and dynamic tech industry in the years to come. If you would like to learn more, please reach out to the Display team here. You can also sign up for our newsletter here and follow us on LinkedIn and Twitter.

Interested in more insights from the display team? Find out how X leans on Google to sell their ad inventory programmatically.

Google Expands AI Overviews and Introduces AI Mode

The search landscape is evolving at breakneck speed, and Google is leading the charge with its latest advancements. AI Overviews – Google’s AI-powered search responses – are getting a upgrade with Gemini 2.0, and a brand-new experimental feature, AI Mode, is rolling out in the US for some users.

Here’s what’s new and why it matters.

Expanding AI Overviews with Gemini 2.0

Google’s AI Overviews have already transformed the search experience for over a billion users. Now, powered by the advanced capabilities of Gemini 2.0, AI Overviews are set to improve.

Key Upgrades:

Enhanced Multimodal Capabilities: AI Overviews can now process complex queries, integrating text, images, and even coding-related questions.
Faster & More Accurate Responses: Gemini 2.0 ensures that AI-generated answers are not only quicker but also more precise.
No Sign-In Required: Access AI Overviews without logging into a Google account—making them more accessible than ever.
Expanded User Base: AI Overviews are now available to teenagers, allowing younger audiences to benefit from AI-driven insights. Its not clear how Google would know a teen user vs an adult for those not logged in.

With these improvements, AI Overviews aren’t just summarising search results—they’re becoming a powerful assistant for tackling deeper, more intricate queries across various domains.

For brands AI Overviews are a branding opportunity to occupy the top real estate on the SERP, however there are also concerns that AI Overviews reduce clicks to websites – the user receives the information they need with no need to click.

Introducing AI Mode: A Game-Changer for Search

Google’s AI Overviews have proven popular, but many users want even more AI-powered assistance. Enter AI Mode, an experimental feature designed to take search to the next level.

Here users will click a tab, the traditional SERP disappears and a whole page is generated with the answer to your question.

What AI Mode Brings to the Table:

Advanced Reasoning & Problem-Solving: AI Mode is built to handle complex, multi-step questions, providing well-structured responses that incorporate deeper analysis.
Conversational Interactivity: Unlike traditional search, AI Mode allows for follow-up questions, creating a more intuitive and personalised search experience.
Real-Time Data Integration: AI Mode doesn’t just rely on static web content—it taps into fresh data from the Knowledge Graph, shopping databases, and live sources.
Dynamic Query Processing: Using a ‘query fan-out’ technique, AI Mode runs multiple searches in parallel to compile the most comprehensive and accurate results.

For instance, if you ask AI Mode, “What’s the best fitness tracker for runners with heart rate monitoring and GPS?”, it won’t just list a few options. Instead, it will break down the key differences, highlight expert reviews, and even suggest related follow-up queries, such as “Which tracker has the longest battery life?”.

Video source: Expanding AI Overviews and introducing AI Mode

The Future of AI Search: What’s Next?

Google is rolling out AI Mode as a limited opt-in experiment via Google Labs in the USA, starting with Google One AI Premium subscribers. This phased release allows Google to refine the experience based on real-world feedback before expanding access more broadly.

This is the same approach that Google had with SGE which became AI Overviews.

While AI Mode aims to provide reliable and objective responses, Google acknowledges that early-stage AI products can sometimes introduce unintended biases. To mitigate this, Google is investing in:

Stronger Factual Validation: AI Mode prioritises high-confidence responses but defaults to traditional search results when uncertainty is detected.
Visual & Interactive Enhancements: Future updates will integrate richer visuals, videos, and interactive elements for a more immersive search experience.

What This Means for Search Marketers and SEO teams

The rise of AI-driven search means businesses and digital marketers need to rethink their SEO and content strategies. Here are some key takeaways:

Prioritise High-Quality Content: AI Overviews and AI Mode will surface content that is well-structured, informative, and authoritative.
Leverage AI for Visibility: Expect SGE (AI Overviews) Ranking Reports to become a crucial tool for tracking how AI-generated results impact visibility. Learn more about SGE Ranking Reports.
Optimise for Conversational Queries: With AI Mode’s interactive approach, optimising for long-tail, question-based queries will be essential.

For more insights on the evolving search landscape, explore our latest content on Integrated Search Marketing Services.Final Thoughts

Google’s AI-powered search evolution is more than just an upgrade – it’s a fundamental shift in how we access and interact with information. With AI Overviews expanding and AI Mode introducing a new layer of AI-driven intelligence, the future of search is more dynamic, intuitive, and tailored than ever.

As always, staying ahead of these changes is key. Whether you’re a brand, marketer, or everyday search user, embracing AI-driven search tools will be crucial in navigating this new era of digital discovery.

If you’re interested in learning more about how this update might affect the search landscape, sign up for our Future of Search event in which we explore cutting-edge technologies that are shaping the way we find information online.

Event details

Date: Wednesday, 2 April 2025
Time: 08:30 AM
Location: TIPi Group Offices, London

The rising cost of Google ads: Why CPCs are increasing in 2025

The cost of running Google Ads continues to climb – and 2025 is no exception. And it’s not just ROAST feeling the pinch, advertisers across nearly every vertical are feeling the pressure as Cost-Per-Click (CPC) rises to record highs (Ref 1,2,3). This inflation isn’t just a fluke or the result of increased competition alone: it reflects a confluence of structural changes within the advertising ecosystem.

From Google’s continued automation push and expansion of keyword matching to platform-level bidding wars and increased investment in digital across industries, we’re seeing a more aggressive auction environment. Add in reduced visibility into user data post-cookie, and it’s clear that advertisers are now paying more per click, not necessarily because demand is booming, but because precision and efficiency are harder to come by.

As an industry-wide trend, at ROAST, we’re seeing this trend reflected in our own client data, and it’s prompted a deeper look at both the why and the what next. ROAST Account Manager Charlie Chester unpacks our internal findings, sharing what we’re doing to drive efficiency in this climate and our ability to offer actionable tactics that balance both performance and cost.

A changing search landscape

A recent review of keyword performance across ROAST clients highlights a significant shift in the Google Ads keyword landscape in 2025.

Throughout Q1, ROAST have observed a sharp rise in the cost of exact match clicks, with Google appearing to prioritise phrase and broad match keywords. This shift presents new challenges for advertisers who rely on exact matches for precision targeting and cost efficiency.

Performance Trends: What the Data Tells Us

Analysing performance in January and February, we’ve seen:

A 10 percentage point (pp) decrease in exact match click share, aligning with a steady increase in phrase match impressions.
A 91% increase in exact match CPCs from December into January across multiple clients, making it clear that cost pressures are mounting.

This shift suggests that Google is favouring broader keyword matching while simultaneously increasing the cost of maintaining control over search queries via exact match.

The brand keyword impact

Google’s guidance recommends leveraging phrase and broad matches alongside smart bidding tools to optimise conversions. However, for critical terms like brand names, ROAST continues to prioritise exact matches to maintain control.

To illustrate why this is important:

One client’s phrase match brand keyword saw CPCs 4x higher than its exact match equivalent – despite no competitors appearing in auction insights.
When isolating brand terms, we observed a clear spike in exact match CPCs, while click volume remained flat.
If investment, competition, and volume remain unchanged – why are costs increasing? The trend suggests a gradual but persistent shift toward phrase match keywords, likely influenced by Google’s automation and bidding algorithms.

How advertisers can respond

Despite these changes, there are still ways to manage CPC inflation effectively:

Demonstrate the Value of Exact Match – While Google is pushing phrase match, our data still shows clear cost benefits to maintaining control with exact match.
Refine Keyword Structures – Ensure phrase and broad match terms are tightly controlled with negative keywords. Regularly review and build out exact match terms to limit unnecessary spending.
Leverage First-Party Data – Efficient audience segmentation and automation tools should be used strategically to maintain performance and advertiser control.
Monitor and Adapt to Google’s Algorithmic Shifts – Google’s machine learning is constantly evolving, so maintaining an agile approach to bidding and match type strategy is critical.

We hope to have highlighted how ROAST is monitoring this industry-wide trend. Afterall, managing search performance in an environment of rising CPCs requires expertise, adaptability, and a data-led approach. At ROAST, we help clients navigate these challenges by ensuring search strategies remain efficient, cost-effective, and aligned with the latest platform developments.

Get in touch with our team to discuss how we can help optimise your paid search campaigns in 2025.

References:

Search Engine Land: CPCs keep increasing – here’s what you can do about it
BrightBid: How to fight back Google CPC inflation in 2025
Foundery Digital: Rising CPCs and evolving trends in digital advertising

Google Expands AI Overviews and Introduces AI Mode: What SEO Managers Need to Know

Image source: Expanding AI Overviews and introducing AI Mode, Google

Google continues to reshape the search landscape with the expansion of AI Overviews and the introduction of Google AI Mode. These advancements mark a significant shift in how search results are generated and displayed, with far-reaching implications for the search landscape. Here’s what Rainier, ROAST’s Senior Organic Performance Executive has to say: what you need to know and how ROAST’s expertise ensures you stay ahead of the curve.

As AI becomes more integrated into search results pages, content strategies are evolving. Where content was once primarily created to serve users and improve rankings, the focus has now shifted toward ensuring visibility within AI-generated responses. This shift is a response to expected drops in click-through rates, as more answers are surfaced directly in the SERPs.”

Expansion of AI Overviews

Initially launched as part of the Search Generative Experience (SGE), AI Overviews provide concise, AI-generated summaries directly within search results. This feature has now expanded to cover in depth questions, offering users faster and more in-depth insights without the need to click through. The goal here is to streamline the information-gathering process, making it more efficient for users to find relevant information.

Introduction of AI Mode

Building upon the success of AI Overviews, Google has introduced AI Mode – a more interactive and personalised search experience. This opt-in feature allows users to engage in conversational interactions with the search engine. AI Mode aims to provide a more intuitive and engaging user experience, aligning with the growing trend of conversational AI.

Key Developments:

More AI-Powered Answers:AI Overviews will expand to cover more complex and multi-step queries, providing direct and synthesised answers within the search engine results pages (SERPs).
AI Mode Integration: A new setting, Google AI Mode, allows users to opt into a more interactive, AI-driven search experience.
Impacts on Organic Search: Traditional search rankings may shift below the fold as AI-generated responses take more visibility within the results.

What Does This Mean for SEO?

Shift in Click-Through Rates (CTR): With AI Mode offering more direct answers than multiple searches, users may find what they need without clicking through on traditional organic listings. This could lead to reduced traffic for informational queries, making it essential to optimise to appear within AI-generated results.

The Role of Content Optimisation: Google AI Mode will prioritise authoritative, well-structured, and semantically rich content. This means:

Structured Data Matters More: Schema markup will play a bigger role in helping Google understand and feature content.
Long-Tail and Conversational Keywords Are Key: AI-generated answers thrive on natural language queries, so optimising for question-based and long-tail search terms is crucial.
Promoting Expertise and Trustworthiness: Google prioritises sources with strong E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) signals.

A New Approach to Search Strategy

AI-Optimised Content Strategies: Brands must create in-depth, high-quality content that AI Mode can extract and display.
Diversifying Traffic Sources: With potential shifts in organic visibility, leveraging paid search, social, and other channels will be critical.
SGE (AI Overviews) Ranking Reports: Tracking AI-driven search performance through services like our SGE Ranking Reports will be essential.

Why ROAST Is Your SEO Partner for AI Search

At ROAST, we don’t just follow search trends: we anticipate them. Our expertise in SEO Services ensures that your brand stays visible and competitive in an AI-driven search environment.

AI-Driven SEO Strategy: Our team analyses AI Mode’s impact on rankings and refines content to perform in AI Overviews.
Technical Excellence: From structured data to entity optimisation, we leverage cutting-edge tactics to maximise visibility.
Data-Led Insights: Our proprietary tools track AI-generated search performance, helping brands make informed decisions.

Stay Ahead of Google’s AI Evolution

The expansion of AI Overviews and the introduction of Google AI Mode signal a fundamental shift in search. Ensuring your brand remains at the forefront requires expertise, agility, and an AI-first approach to SEO.

FAQs on Google AI Overviews and AI Mode

What are AI Overviews in Google?

AI Overviews are AI-generated search summaries that provide quick, synthesised answers directly in search results. They leverage machine learning to extract relevant information from multiple sources.
What is AI Mode in Google?

Google AI Mode is a new feature that allows users to opt into an AI-enhanced search experience, delivering more interactive and comprehensive AI-driven answers.

What’s the difference between AI Overviews and AI Mode?

AI Overviews: AI-generated search results appearing within standard queries, similarly to the historically popular featured snippets results.

AI Mode: Using a customised version of Gemini 2.0, AI Mode is designed to help with further questions, allowing users to explore a topic in detail, with comparison & reasoning, allowing users to essentially search within the search results page.

What is the impact of AI Mode on Google search results?

AI Mode could lead to a decline in traditional organic search traffic as more answers are generated within search results. SEO strategies must adapt by focusing on AI-optimised content, structured data, and high-quality authoritative content.

Google Launch Audio Overviews in Labs Experiment

Google have launched a new search experience with the introduction of Audio Overviews, a new feature that transforms search results into engaging, AI-generated audio summaries.

Part of Google’s AI overviews this new experiment is currently available in Google Labs, offers users a hands-free way to absorb information quickly and efficiently. What could Audio Overviews mean for users and publishers? Where could this lead to in the future with video content.

Google Launch Audio Overviews in labs experiment

Source

What are Google Audio Overviews?

Audio Overviews utilise Google’s latest Gemini models to create concise, conversational audio summaries of search results.

When users opt into this experiment and perform a search, they may see an option to generate a short audio overview on their query topic. This feature aims to provide a quick “lay of the land” for unfamiliar subjects, offering a convenient way to absorb information while multitasking or for those who prefer auditory learning.

Interestingly the audio file generated is in a podcast / conversational style where two people are answering the question. In the example below, the video has a male and female voice with an answer on how to improve your credit score.

Each example ends with “to learn more check out the list of sources below”

You’ll also notice that it takes 20 seconds to generate, a combination of the AI searching, then generating and creating the audio clip which is 3:56 seconds long.

Almost as a “sister” to AI overviews the Audio Overviews gather their information from a selection of sources which are shown below the audio player. But as many have commented, they don’t mention these sources explicitly within the audio.

Example script from “latest trends in NHS digital transformation 2025”

https://weareroastdevs.wpenginepowered.com/wp-content/uploads/2018/02/example.mp3Female: So, the NHS is going through a massive digital makeover, right? What are the main things we should be looking out for in 2025?

Male: Absolutely. We’re seeing big changes in AI, how patients are cared for, and making sure everyone has fair access to health care. And that’s just the start.

Female: AI seems to be everywhere. How will that actually play out in the NHS?

Male: Well, AI is helping to translate languages, making sure everyone understands their treatment. Also, it’s being used to spot problems early on and even help doctors make better decisions. But it needs to be managed properly so all practices, big and small, can use it effectively.

Female: That makes sense. What about getting all the different NHS systems to talk to each other? It’s always seemed like a challenge.

Male: You’re spot on. Getting systems to work together is super important. The focus is on making sure all the different parts of the NHS can share patient records easily. This way, doctors have all the info they need, no matter where you’ve been treated before.

Female: So no more repeating your medical history every time you see a new doctor?

Male: Exactly. And it also means that things like booking appointments and seeing test results can all be done through one app, like the NHS app.

Female:That would be a huge improvement. What about keeping all that data safe? Cyber attacks are a real worry.

Male: Data security is a top priority. The NHS is working hard to protect patient info from cyber threats. This means using the latest encryption and keeping a close eye on their systems to spot any problems early.

Female: It’s good to know they’re on it. What about people who aren’t so comfortable with tech? How does this help them?

Male: That’s a really important point. The NHS is trying to make sure that everyone, no matter how good they are with tech, can still get the care they need. This means things like making sure websites and apps are easy to use, and still offering support in person or over the phone if that’s what people prefer.

Female: So it’s not just about the tech, but also about making sure it works for everyone.

Male: Exactly. And it’s about using tech to free up staff time so they can spend more time with patients. Things like remote monitoring, where patients can check their health from home, can really help take the pressure off hospitals.

Female: That sounds like a win-win. What about the costs? Is all this digital stuff expensive?

Male: It can be. But the idea is that by using tech smartly, the NHS can actually save money in the long run. Things like cutting down on paperwork, making processes more efficient, and preventing problems before they become serious can all add up to big savings.

Female: So it’s an investment in the future?

Male: Precisely. And it’s not just about money. It’s about making the NHS more sustainable too. Digital solutions can help reduce the NHS’s carbon footprint, like cutting down on travel for appointments.

Female: I see. So what are the biggest challenges in making all this happen?

Male: One of the biggest is getting everyone on board. It’s not just about having the right tech, but also about making sure staff have the skills and training they need to use it properly. There’s also the challenge of making sure all the different systems can work together and that data is shared safely and securely.

Female:So it’s a team effort?

Male: Absolutely. It needs everyone, from healthcare professionals to tech experts to policymakers, all working together to make it a success.

Female: Alright, so it sounds like the NHS is really trying to embrace digital tech to improve health care for everyone. What do you think is the one thing that would make the biggest difference in 2025?

Male: If I had to pick one thing, it would be making sure that everyone has equal access to these digital tools. If we can close that digital divide and make sure that everyone can benefit from these advancements, that would be a game changer.

Female: I agree. It’s about making sure no one gets left behind. Thanks for breaking it all down for us.

Male: My pleasure. It’s an exciting time for the NHS, and I’m looking forward to seeing how these changes play out.

Female: To learn more, check out the listed sources below.

Audio Overview Test Results

As with any new feature on Google best thing is to test, try different key phrases, see what variances there is in the new feature. We conducted a small test, here are some findings.

On 21 random searches, audio overviews appears 10 times.
An Audio Overview can appear on its own or with an existing AI Overview
Most often it shows 10 resource and links, but there are examples of 8 and 9
They appear sometime below P1, sometimes further down
On average Google take about 20 seconds to generate
Average lengh of the audio fle was 269.9 seconds (4 mins, 29 seconds)
You can’t download the MP3, via source you can only see the first 13 second chunk

Benefits for Users: Accessibility and Convenience

The introduction of Audio Overviews brings several advantages to users:

Enhanced accessibility for visually impaired users
Hands-free information consumption for multitaskers
Quick overview of complex topics
Personalised learning experience for auditory learners

ChatGPT has long had the ability to have answers read out to users, once generated.

This feature could be particularly beneficial for students, professionals on-the-go, or anyone looking to quickly grasp the essence of a topic without extensive reading.

Tracking Audio Overviews

If audio overviews launch for real in the future we would look to track with the following methodology.

Does the Audio Overview appear (True/False)
If so what rank position (P1 to P10)
URLS in the sources (URL list)
Transcription (written text)
Brand mention (extracted from the transcription)
Length of audio

Potential Impact on Click-Through Rates

While Audio Overviews offer benefits for some users, there’s a potential concern for website owners and content creators. By providing comprehensive summaries, directly in search results, users might feel less inclined to click through to individual websites. This could lead to a decrease in click-through rates (CTR) for organic search results.

The decrease in CTR has already been noted with AI overviews, however with Audio, there is much more likely the user will put their device down and listen, or on a desktop click to another tab / application and less likely to click to a source.

Why Audio Overviews Make Sense for Google

Google’s venture into Audio Overviews is a natural progression of their existing technologies. The company has already developed advanced text-to-speech capabilities, evident in products like Google Assistant and Google Translate. By leveraging their Gemini AI models, Google can now offer more nuanced and context-aware audio summaries.

Future Possibilities: Video Overviews?

The introduction of Audio Overviews opens up exciting possibilities for the future of search. One potential evolution could be the development of Video Overviews, where AI generates short video summaries of search results. This could incorporate relevant images, animations, clips from YouTube and even AI-generated visuals to provide a more comprehensive and engaging overview of topics. We know that Google has the technology with VEO 3 from Deepmind, a Google company being one of the leading models for video generation.

However, it would also raise new challenges in terms of content creation, copyright issues, and the potential for misinformation.

Who Can Access Audio Overviews?

Currently, Audio Overviews are available as an experiment in Google Labs. The feature is limited to:

Users in the United States
Searches conducted in English

As with previous experiments, Google may then release to the public after this testing period.

How to Sign Up for Audio Overviews

If you’re interested in trying out Audio Overviews, follow these steps:

Visit the Google Labs page for the Audio Overviews experiment – https://labs.google.com/search/experiment/30
Opt into the experiment
Perform a search on Google
If available for your query, you’ll see an option to generate an audio overview

Users can provide feedback on individual overviews and the experiment as a whole through thumbs up/down options in the interface.

As we at ROAST continue to monitor and analyse the impact of innovations, like Google Audio Overviews, we’re excited about the potential of new ways to optimise content and improve user experience, but at the same time understand the possible impacts on clients existing performance. For more information on how these changes might affect your SEO strategy, check out our AI Overviews Ranking Reports.

AI Max and the End of Match Types

What Rising CPCs Really Mean for Advertisers

Advertisers have long debated the merits of different match types across Google, and the introduction of AI Max has been no exception. While arguments about control and reach opportunities remain valid, no one can ignore the overarching pressure: cost.

Looking at Google’s trajectory, the direction is clear: from exact, phrase, and broad match, to Smart Bidding, then Performance Max, and now AI Max. Each stage has shifted power away from advertisers and toward machine-led optimisation. Where marketers once had levers to tightly define queries and bids, those controls have steadily eroded, replaced by systems that prioritise Google’s automation and signal-driven decision-making. AI Max is not an isolated development, it is the culmination of this long process, where the final vestiges of keyword control give way to a fully automated model.

This shift has played out against a backdrop of rising CPCs throughout 2025 (1,2). As we approach the end of the year, advertisers face a double challenge: not only higher costs, but also diminishing control. Google is steering advertisers deeper into its AI ecosystem, where automation increasingly replaces traditional match types.

Google’s narrative is consistent: automation delivers scale, and with AI Max, this is no different. The emphasis has moved from keyword matching to predicting user intent, framed as better performance, more signals, and less friction for advertisers. Yet it remains unclear whether this shift is truly performance-driven, or whether advertisers have been gradually primed for adoption. Rising phrase match CPCs (3), in particular, have pushed many to test alternatives, making automation appear the inevitable next step.

Data taken from the ROAST MCC

Is Phrase Match Finished in the Age of AI Bidding?

Across the ROAST MCC, AI Max testing shows cheaper CPCs with CTRs comparable to phrase, further highlighting that phrase is no longer viable at scale. Exact match retains strategic value, especially for high-intent objectives where CTR matters most. This trend is consistent across verticals including property, insurance, and healthcare logistics.

Search query matching remained comparable to broad match, with clear signals that Google’s AI models now lean on contextual inputs such as landing page relevance and creative assets to inform query matching. Campaign strategies can no longer depend on keyword selection and bid adjustments; instead, success will hinge on providing high-quality signals for Google to optimise against.

This shift carries two major implications:

Budgets will need to flex around machine-led decisions.
Strategies will rely more on landing page content and asset quality rather than traditional keyword structures.

While Google has introduced reporting tools to provide transparency on which search terms trigger ads and how those queries perform, this requires more frequent checks and greater reliance on independent measurement to fill in the gaps.

Looking Forward

The core driver behind this shift is changing consumer behaviour. Google has confirmed the rollout of ads in AI Mode in Q4 2025, where queries are typically two to three times longer than in classic search. This evolution is reshaping how advertisers operate, moving campaigns away from keyword dependency and toward intent-driven targeting.

Google frames the change as augmenting, not replacing, Search. With traffic growing in categories where AI Overviews perform strongly, the message for advertisers is clear: prepare for smarter, more contextual targeting on AI surfaces.

Recommendations

Short-term

Test AI Max in parallel with phrase: With lower CPCs and comparable CTRs, AI Max can provide valuable insights when layered into strong campaigns.
Prioritise creative testing: As match types fade, ad copy, assets, and messaging become the most controllable levers. Run systematic A/B tests to feed automation better input signals.
Optimise landing page content: Strong on-page content supports both performance and automatic asset creation.

Mid-term

Strengthen first-party data & conversion modelling: With reduced transparency and stricter privacy updates, capturing robust data is critical.
Adopt independent measurement (incrementality / MMM): Run incrementality tests to measure true lift and incorporate media mix modelling to validate long-term trends, avoiding over-crediting automated channels (4,5).
Focus on data quality, while this has been a rule for a while. It is now essential to optimise towards the most valuable actions to guide automation.

Long-term

Build a culture of continuous experimentation: Incorporate lift tests, geo holdouts, and validation experiments into your media plan to prevent overconfidence in automation.

Conclusion

Rising CPCs may have defined 2025, but the bigger story is the shift in control. As Google doubles down on Ai-driven campaigns and ads in Ai-powered search, advertisers must adapt or risk falling behind.

The winners in 2026 will be those who embrace automation strategically, balancing AI Max adoption with strong creative, robust first-party data, and a diversified media mix.