Published on
Zach Jackson

February was another busy month in search, bringing a mixture of welcome news and fresh challenges for website owners, publishers, marketers, advertisers, and content creators. Here we spotlight the most important developments, the ones that matter most to your digital strategy.

Updates are colour coded by importance:

🔴 Major developments likely to impact strategy

🟡 Worth watching or understanding

🟢 Informative - but lower impact for most

SEO

🔴 Microsoft launches AI performance report in Bing Webmaster Tools

Bing now offers AI performance insights within Bing Webmaster Tools, the Bing equivalent of Google Search Console. This new report shows site owners and marketers how content appears across Microsoft Copilot, AI summaries within Bing, and select partners (including ChatGPT).

It’s not as comprehensive as it could be, but it’s definitely a step in the right direction, offering greater clarity into performance across Bing’s AI systems. It shows:

  • Total citations - how often your content is cited on the whole
  • Average cited pages - the average number of unique pages cited in AI-generated outputs per day over the selected time range
  • Grounding queries - the key phrases that Bing’s AI has used that led it to your content. Note: not the user’s actual prompt
  • Page-level citation activity - citation counts for specific page URLs from your website
  • Visibility trends over time - a timeline showing how citation activity has changed over a given period

For site owners and marketers…

Click data is still absent, and that’s the metric everyone wants to see, but this is a huge step in the right direction from Bing. We hope to see Google make similar moves soon.

We see this report being helpful for both developing digital strategy and tracking campaign progress as search becomes increasingly AI-based. 

We’ve been using advanced, bespoke reporting solutions to give our clients a window into AI performance for some time, but we're integrating Bing’s new insights to provide an even more robust account of campaign impact. Reach out today to learn more about our clear, comprehensive, and transparent reporting methodology and how it benefits your digital strategy.

🔴 Microsoft introduces content marketplace to pay publishers when content is used for AI

At the start of February, Microsoft announced Publisher Content Marketplace (PCM). Through PCM, publishers will be able to licence premium content to AI builders seeking quality content for grounding AI outputs.

According to Microsoft, usage reporting will be transparent, and publishers will determine the licensing terms. Microsoft’s stated goal is to protect the web’s content ecosystem as we shift towards AI-led search, while ensuring their models have access to quality information that improves outputs.

For publishers…

Although the PCM is being developed in partnership with some of the US’ largest publishers (think Associated Press and USA Today), Microsoft aims to open it up to publishers of all sizes, including ‘specialised and independent voices’.

It’s early yet, but those interested in opting in could start considering how PCM might impact their content strategy. For instance, how will you decide which premium content is put on the marketplace, and which should be reserved behind paywalls for human users. Crucial to note is that AI-licenced content effectively grants free access to human users, so deciding which to offer may require careful thought.

How strong a revenue stream PCM could provide is yet to be determined, but it’s a development worth watching closely.

🔴 Google downranks self-serving listicles

SEO data appears to show that Google has taken a stand against self-serving “best of” listicles that promote the content publishers as the top choice (or one of) to readers. 

For example, imagine we published a post called ‘Best UK SEO and PPC agencies in 2026’, and put TDMP in the top spot.

This content format emerged as a “hack” to game Google’s (and others') AI search platforms into recommending a company’s own offerings to users in outputs. 

While Google hasn’t mentioned anything about this, many SEOs believe it to be related to one or more recent unannounced updates.

For businesses and content strategists…

If you were considering drafting some self-promoting listicles, don’t. Not only does it leave a bad taste in the discerning reader’s mouth, it will now likely be crushed by Google’s algorithm, leading to wasted resource.

While this tactic was effective, it’s clearly manipulative and black-hat leaning.

With AI search being so new, several tactics have and will continue to emerge that genuinely help to boost visibility on these platforms, but before implementing, it’s important to consider sustainability.

If a tactic feels cheap or in poor taste, it likely is. It may work today, but tomorrow it could cause more harm than it ever did good.This doesn’t mean all AI visibility tactics are bad. The key is evaluating which are fair and mutually beneficial to users, AI, and Google.

At TDMP, we assess not just the efficacy of content visibility techniques, but the longevity. Our clients benefit from strategies that work now whilst protecting their digital interests as algorithms shift. Learn more about our SEO services.

🟡 First Google core update of 2026 goes live (US only)

Google launched what they’re calling the February 2026 Discover core update in early February. As the name suggests, it focused specifically on “systems that surface articles in Discover”.

In terms of scale, it targets English language users in the US, but will eventually roll out globally. Google has been unusually forthright about the objectives behind this update, writing that it will make the following improvements:

  • Showing user more locally relevant content from websites based in their country
  • Reducing sensational content and clickbait in Discover
  • Showing more in-depth, original, and timely content from websites with expertise in a given area, based on our systems’ understanding of a site’s content

We’ll be covering the February 2026 Discover core update in detail in our continuously refreshed guide to Google’s algorithm updates.

🟡 Google settles on prominent link system for AI Overviews and AI Mode

For while now, Google has been testing different mechanisms of linking from AI Overviews and AI Mode to source websites. It seems they’ve settled (for the time being at least) on pop-up link cards that appear when you hover over a citation.

A screenshot of hover-over link cards in Google's AI Mode

Source: X

For site owners…

This is a far more prominent linking system, so it could realistically help to improve clicks from Google’s AI interfaces. And even if someone doesn’t ultimately click through to your content, they’re more likely to see and remember your site or brand name.

This can be a good awareness and trust builder. The user sees you as the source chosen by Google’s AI, validating you as an authority on the topic.

🟢 The markdown for AI bots debate intensifies

We recently published an article on the worth of serving markdown versions of webpages to benefit AI bots, concluding that it’s not currently an effective or responsible tactic. Read our full guide.

Shortly thereafter, Google’s John Mueller weighed in on the debate, reinforcing our own analysis by stating there’s no need or benefit to serve markdown to AI bots, questioning whether these crawlers can even parse markdown links effectively.

Then, in mid February, Cloudflare announced a feature that auto-converts HTML into markdown, giving AI bots access to both without any manual work.

This feature is freely available in beta for those on eligible paid Cloudflare plans.

Strictly speaking, Cloudflare’s HTML-to-markdown auto conversion tool is doing something slightly different to what we discussed in our guide and what Mueller has advised against.

Instead of serving parallel pages to AI bots, one in standard HTML and the other as markdown, Cloudflare’s feature enables your single URL to serve one or the other based on what the bots request. This is called “content negotiation”.

Within this feature, there is also the option to include content signals that explain whether content can be:

  • Used to train AI
  • Used for AI search
  • Used for AI input (i.e. agentic search and other operational tasks)

As remarked in our guide, Robots.txt, AI crawlers, and what (if anything) you should be doing about it, nothing is forcing bots to comply with these directives, and there’s no evidence that they work.

For site owners and marketers… 

If you have access to Cloudflare’s HTML-to-markdown converter, we’d recommend against opting in at this moment in time. It could be something to watch, but, as we mentioned in our guide, there currently isn’t any hard evidence of markdown for AI bots being beneficial, and it’s still unclear how search engines will perceive this practice in the future.

In response to the new Cloudflare feature, Mueller has stuck firmly to his guns, suggesting it would be better to focus efforts on improving the main webpage for both human and AI users.

Paid Media

🔴 Google introduces new security feature for Google Ads

Google has launched ‘Multi-party approval’ (MPA) for Ads accounts, a new security feature that requires a second account administrator to approve any “high risk actions”, such as changing user roles or adding/removing users.

For businesses and advertisers…

There has been a spike in Google Ads account hijackings as of late, even of accounts with two-factor authentication enabled, so we’d recommend using this additional layer of security where practicable.

You can learn more about MPA implementation and the approval process from Google’s help doc.

🟡 OpenAI launches broader test of ads in ChatGPT (US only)

After an initial testing period, OpenAI has officially launched a larger, more formalised test of paid ads in ChatGPT in the US - but with an unanticipated twist. 

Users have access to a set of controls that gives them more agency over the sort of ads they see while using ChatGPT. OpenAI stated:

‘You can control the ads you see in ChatGPT, including dismissing ads, sharing feedback, learning how and why you’re being shown a particular ad, deleting your ad data with one tap, and managing ad personalisation at any time.’

When a user opts out of personalisation, ads remain contextually relevant to the conversations they appear beneath, but are not fine-tuned based on user history.

🟡 Google removes Parked Domains from the Search Partner Network

Google has now fully retired Parked Domains (AFD) as an ad surface within the Search Partner Network. 

Parked domains are undeveloped web pages with little or no content, long associated with low‑quality traffic. Google had already stopped including them by default, and as of 10 February 2026, they’ve been removed entirely, along with the related content‑suitability setting.

For advertisers…

For advertisers, this is a positive quality‑of‑traffic update. It means Search Partner placements should now deliver more meaningful user intent and fewer low‑value clicks.

🟡 Data reveals which Google ad spaces generate the most invalid clicks

Data posted to LinkedIn by Mike Ryan shows which of Google’s ad networks leads to ads getting the most invalid clicks. 

“Invalid clicks” are non-genuine engagement, i.e. accidental clicks or fraudulent clicks by bad actors and bots. For each ad space, the data shows the split between these accidental and fraudulent invalid click types.

Ryan’s findings were based on a Q4 2025 observation of over 550 million ecommerce clicks across the Google Ads ecosystem:

  • Google Display Network: 17.5% invalid clicks. The majority were accidental clicks.
  • Search Partners: 10.3% invalid clicks. The majority were fraudulent clicks
  • Discover: 8.1% invalid clicks. Almost entirely benign - only 0.1 clicks were fraudulent
  • Cross-network: 6.8% invalid clicks. The majority were accidental clicks
  • YouTube: 5.6% invalid clicks. Almost entirely benign - only 0.2% were fraudulent
  • Google Search: 5.3% invalid clicks. A relatively even split between accidental and fraudulent
  • Maps: 2.4% invalid clicks - all benign
  • Gmail: 1.6% invalid clicks - all benign

A bar chart showing the percentage of invalid clicks per Google ad network

Source: LinkedIn

For businesses and advertisers…

This is interesting data that can help you decide how to structure your paid media strategy, but there are some crucial points to note:

  1. High invalid click rates don’t necessarily mean wasted budget - Google filters out any invalid clicks it detects. You are not billed for these clicks.
  2. It’s normal for different networks to behave differently - Display and Discover naturally see more accidental taps due to mobile usage patterns. Search Partners can vary more widely because they include third‑party sites. YouTube, Maps, and Gmail tend to be more controlled environments, which is reflected in their lower rates.
  3. This data shouldn’t scare advertisers away from any channel -  Every network still drives valuable, incremental reach. The key is understanding the context so you can set realistic expectations and evaluate performance with the right lens.
  4. The real takeaway is about performance analysis - Although invalid clicks are filtered by Google and removed from your billing, they can still influence:
  • CTR
  • CPC volatility
  • Conversion rate stability
  • Traffic quality signals

For example, a spike in accidental clicks on Display might inflate CTR or distort early‑stage engagement metrics. Understanding how involuntary clicks impact different ad networks can therefore help you interpret performance fluctuations and optimise using reliable data.

If you want support navigating performance fluctuations across Google’s networks, TDMP is an award‑winning Google Premier Partner PPC agency with a proven track record of delivering exceptional results for clients, even in highly regulated sectors. Explore our PPC services here.

Stay current with TDMP

As ever, the search industry is adapting to better utilise technological advancements and better serve searchers. When so many key aspects shift on a monthly basis, keeping up and maintaining a robust digital strategy can be difficult, but we can help.

TDMP is an award-winning digital marketing agency specialising in paid media for complex markets and AI-aware SEO, empowering brands to stay ahead of change and turn digital disruption into opportunity. If you’re ready to work with an experienced and impactful digital partner, let’s talk.

Keep your finger on the TDMPulse

Sign up to our newsletter for monthly insights, news & guides

Â