CMA Google Search Choice Screens: What SMEs Should Do
CMA Google Search Choice Screens: What SMEs Should Do
CMA Google Search Choice Screens: What SMEs Should Do
CMA Google search choice screens could shift where UK search traffic starts. See what to track, what to fix on your site and how to prepare.
CMA Google search choice screens could shift where UK search traffic starts. See what to track, what to fix on your site and how to prepare.
9 min read

CMA Google search choice screens change which box a person types into. They do not change what Google rewards once they have typed. Keep those two things apart and most of the panic in your inbox resolves itself.
Here is the problem with the coverage so far: nearly all of it is written for policy people. Designation notices, consultation documents, paragraph numbers. Useful if you are a competition lawyer. Not much help if you run a business that gets most of its enquiries from organic search and you want to know whether to change anything this quarter. So this piece covers the regime quickly, accurately, then spends the rest of its length on what you should actually do.
What the CMA has decided, and what is still a proposal
The legal machinery is the Digital Markets, Competition and Consumers Act 2024. The digital markets competition regime under it took effect on 1 January 2025, and the Competition and Markets Authority opened a strategic market status investigation into Google's general search and search advertising services that same month. A proposed decision followed in June 2025, alongside a roadmap of the measures the CMA might impose. Designation was confirmed later in 2025.
Designation is not a remedy. It is permission to make remedies. Once a firm holds strategic market status, the CMA can impose conduct requirements on it, and each requirement is consulted on and introduced separately. Choice screens are one item on a list that also includes fair ranking rules, publisher control and attribution over how content is used, and data portability.
That is why the dates in news coverage contradict each other. Four different moments get reported as if they were one: the CMA proposing a measure, the consultation on it, the final decision, and the day users see something on a phone. A headline about choice screens in spring tells you nothing about when a screen appears on a handset in Doncaster. Before you plan around any date, check the CMA's case page for the current status, because this moves.
Who would actually see a search choice screen, and where
The screens under discussion are tied to Android device setup and to the Chrome browser. Not to every search a person makes. Reported strengthening of the proposals goes beyond a single one-off question at setup: prompts on first use, and periodic reminders, so the choice is not buried in the five minutes nobody remembers after unboxing a phone.
AI assistants have been pulled into the same conversation, which matters more than the browser detail. If the prompt a user is offered includes an assistant rather than only a traditional search engine, the question shifts from "which engine indexes you"to "which system is summarising you". Those are different problems with different fixes.
Plenty stays untouched. Desktop Safari. Existing devices already configured, unless a reminder reaches them. And every user who sees the screen and taps Google anyway, which, on the EU evidence, is most of them.
What the EU's choice screens already tell us about real movement
We have five years of precedent here, and it is not dramatic.
Google ran a paid auction choice screen on Android in the EEA from March 2020: engines bid for a slot. That model was replaced in September 2021 with free inclusion for a wider set of providers, which removed the complaint that only the richest challengers could appear. Since March 2024, under the Digital Markets Act, Google has shown choice screens on Android and in Chrome across the EU.
What happened? Alternative engines publicly reported sharp spikes in installs and downloads around the launch. What did not follow was a durable change in overall share of any size that would reshape a marketing plan. Spikes are easy. Habit is hard.
Weigh any "the default is up for grabs"claim against StatCounter's UK search engine market share data, which has shown Google holding the overwhelming majority of UK search referrals for years. That is the ceiling a prompt is working against. A few percentage points of query volume redistributing over a couple of years is plausible. A collapse in Google referrals is not, and anyone selling you an urgent pivot on that basis is selling you something.
What changes for your organic traffic, and what genuinely does not
Nothing in the choice screen proposals makes helpful, specific, first-hand content less valuable. The ranking criteria are not the subject of the measure. The default destination is.
What does change, quietly, is dependency risk. A business that gets nearly all of its non-paid traffic from one engine has a single point of failure, and the fix for that was always the same regardless of regulators: be properly indexed and properly findable in more than one place. If choice screens give you a second meaningful referrer, your traffic gets less fragile. That is the honest upside.
Now the part most coverage underplays. For anyone who publishes, the fair ranking requirements and the publisher content attribution requirements are the bigger story. Choice screens govern which box someone types into. Fair ranking and attribution govern how your pages are surfaced, how they are reused, and what happens when your content is summarised inside an answer rather than clicked. If you write for a living or your website is your main sales channel, those are the consultations to read. We cover the publishing side of that in our notes on Google's AI content guidelines for publishers.
One more wrinkle. A small share shift hurts narrow commercial pages more than broad informational ones. If a page earns only a handful of clicks a month on a high-intent query like "emergency boiler repair Reading", losing even a small percentage of those clicks is an enquiry you can feel. A guide pulling 3,000 impressions a month absorbs the same percentage without anyone noticing.
Set your baseline before anything rolls out
This is the work that actually pays, and almost nobody does it until it is too late to be useful.
Export the before picture
Pull a Search Console performance export split by device and by country, covering at least the last 12 months. Then build a GA4 table of sessions by source and medium for the same period, and write down, in plain numbers, what share already arrives from Bing, DuckDuckGo, Ecosia and direct. Most SMEs have never looked. The combined share is usually small, often only a few percent, and knowing your actual figure is the whole point.
Save both exports somewhere a human will find them in a year. A screenshot in a Slack thread does not count.
Split your rank tracking by device
Most SME rank tracking quietly assumes one engine and one device: a weekly desktop scan on google.co.uk, scheduled in 2021, never revisited. If a slice of mobile queries moves to another provider, that scan still reports healthy green numbers while your enquiry form goes quiet. Treat it as two jobs. Track mobile rankings separately from desktop, because the screens under discussion are mobile-first, and the gap between the two is where the early signal will show. Our piece on what rank tracking numbers do and do not tell you goes further into why averaged positions hide this sort of thing.
Learn to tell the two causes apart
A choice screen effect and a core update look nothing alike once you know where to look. A share shift shows up as fewer clicks from one engine while your impressions and average position for the same queries hold steady: the pages still rank, fewer people are asking that engine. A core update shows up as position loss concentrated in specific query clusters or page types, with impressions falling alongside clicks.
Check device split too. A regulatory shift should bite on Android mobile first and leave desktop broadly alone. An algorithm change rarely respects that line. If you are working through a suspected ranking drop, our walkthrough on diagnosing an unconfirmed Google update covers the sequence. Get this wrong and you rewrite perfectly good pages chasing the wrong cause, which we see constantly and which costs weeks.
The practical checklist for a lean team
Here is the good news. Several alternative engines lean on Bing's index rather than crawling the web independently, so "preparing for choice screens"is mostly an indexation job. It takes an afternoon and costs nothing.
Verify in Bing Webmaster Tools and submit your sitemap there. Not only to Google. You can import your Search Console properties directly, which saves the DNS faff.
Turn on IndexNow. It pings participating engines when you publish or update a page, rather than waiting on a crawl that might take a fortnight. Most CMS plugins support it with a tick box.
Read your robots.txt properly. We find blocked bots on client sites regularly, usually inherited from an old developer or a security plugin. Check bingbot, DuckDuckBot and Applebot are not quietly excluded, and check any firewall or bot-management rules at the CDN layer too, because that is where the real blocking usually lives.
Keep your Article, Breadcrumb and FAQ schema clean and un-duplicated. If you run an SEO plugin such as Yoast, confirm it is not already emitting the same markup your theme is adding. Two competing Article blocks on one page is worse than none.
Keep publishing to the same cadence. A consultation is not a content strategy.
That last point is the one we would defend hardest. The content that earns a click on a non-Google engine is the same content that earns one on Google: specific, first-hand, and genuinely answering the query someone typed. Nothing in the CMA programme rewards thinner or faster publishing. If anything, attribution requirements push the other way, towards work with a name and a point of view attached to it, which is how we write everything anyway.
Frequently Asked Questions
What are the CMA's Google search choice screens?
They are a proposed conduct requirement under the Digital Markets, Competition and Consumers Act 2024, which would make Google present UK users with a prompt to pick their search provider rather than defaulting silently to Google. The screens under discussion apply to Android device setup and the Chrome browser. They sit alongside other measures the CMA has signalled, including fair ranking and publisher content attribution.
When will UK users start seeing a search choice screen?
There is no single confirmed date, and the dates quoted in news coverage usually refer to different stages: a proposal, a consultation, a final decision, or a go-live. Each conduct requirement is consulted on separately after designation, so rollout follows the final decision on that specific measure. Check the CMA's case page on gov.uk for the current position before you plan around any date.
Will choice screens reduce my Google traffic?
Probably a little, over time, rather than suddenly. The EU has run choice screens on Android since 2020 and under the Digital Markets Act since March 2024, and alternative engines reported install spikes rather than lasting share change. With StatCounter showing Google well above 90% of UK search referrals for years, a few percentage points redistributing is the realistic scenario, and narrow commercial pages will feel it before broad guides do.
Should I start optimising for Bing and DuckDuckGo now?
Not optimising, exactly. Indexing. Verify in Bing Webmaster Tools, submit your sitemap, enable IndexNow and make sure no bot rules are blocking bingbot, DuckDuckBot or Applebot. Several alternative engines draw on Bing's index, so that afternoon of plumbing covers most of them, and you do not need a separate content strategy on top.
How do I tell a choice screen effect from a Google core update?
Look at impressions and device split. A share shift shows clicks falling from one engine while impressions and average position for the same queries hold, concentrated on mobile. A core update shows position loss on specific query clusters across devices, with impressions dropping alongside clicks. You can only run that comparison if you exported a clean baseline beforehand, which is why we keep telling people to do it now.
CMA Google search choice screens change which box a person types into. They do not change what Google rewards once they have typed. Keep those two things apart and most of the panic in your inbox resolves itself.
Here is the problem with the coverage so far: nearly all of it is written for policy people. Designation notices, consultation documents, paragraph numbers. Useful if you are a competition lawyer. Not much help if you run a business that gets most of its enquiries from organic search and you want to know whether to change anything this quarter. So this piece covers the regime quickly, accurately, then spends the rest of its length on what you should actually do.
What the CMA has decided, and what is still a proposal
The legal machinery is the Digital Markets, Competition and Consumers Act 2024. The digital markets competition regime under it took effect on 1 January 2025, and the Competition and Markets Authority opened a strategic market status investigation into Google's general search and search advertising services that same month. A proposed decision followed in June 2025, alongside a roadmap of the measures the CMA might impose. Designation was confirmed later in 2025.
Designation is not a remedy. It is permission to make remedies. Once a firm holds strategic market status, the CMA can impose conduct requirements on it, and each requirement is consulted on and introduced separately. Choice screens are one item on a list that also includes fair ranking rules, publisher control and attribution over how content is used, and data portability.
That is why the dates in news coverage contradict each other. Four different moments get reported as if they were one: the CMA proposing a measure, the consultation on it, the final decision, and the day users see something on a phone. A headline about choice screens in spring tells you nothing about when a screen appears on a handset in Doncaster. Before you plan around any date, check the CMA's case page for the current status, because this moves.
Who would actually see a search choice screen, and where
The screens under discussion are tied to Android device setup and to the Chrome browser. Not to every search a person makes. Reported strengthening of the proposals goes beyond a single one-off question at setup: prompts on first use, and periodic reminders, so the choice is not buried in the five minutes nobody remembers after unboxing a phone.
AI assistants have been pulled into the same conversation, which matters more than the browser detail. If the prompt a user is offered includes an assistant rather than only a traditional search engine, the question shifts from "which engine indexes you"to "which system is summarising you". Those are different problems with different fixes.
Plenty stays untouched. Desktop Safari. Existing devices already configured, unless a reminder reaches them. And every user who sees the screen and taps Google anyway, which, on the EU evidence, is most of them.
What the EU's choice screens already tell us about real movement
We have five years of precedent here, and it is not dramatic.
Google ran a paid auction choice screen on Android in the EEA from March 2020: engines bid for a slot. That model was replaced in September 2021 with free inclusion for a wider set of providers, which removed the complaint that only the richest challengers could appear. Since March 2024, under the Digital Markets Act, Google has shown choice screens on Android and in Chrome across the EU.
What happened? Alternative engines publicly reported sharp spikes in installs and downloads around the launch. What did not follow was a durable change in overall share of any size that would reshape a marketing plan. Spikes are easy. Habit is hard.
Weigh any "the default is up for grabs"claim against StatCounter's UK search engine market share data, which has shown Google holding the overwhelming majority of UK search referrals for years. That is the ceiling a prompt is working against. A few percentage points of query volume redistributing over a couple of years is plausible. A collapse in Google referrals is not, and anyone selling you an urgent pivot on that basis is selling you something.
What changes for your organic traffic, and what genuinely does not
Nothing in the choice screen proposals makes helpful, specific, first-hand content less valuable. The ranking criteria are not the subject of the measure. The default destination is.
What does change, quietly, is dependency risk. A business that gets nearly all of its non-paid traffic from one engine has a single point of failure, and the fix for that was always the same regardless of regulators: be properly indexed and properly findable in more than one place. If choice screens give you a second meaningful referrer, your traffic gets less fragile. That is the honest upside.
Now the part most coverage underplays. For anyone who publishes, the fair ranking requirements and the publisher content attribution requirements are the bigger story. Choice screens govern which box someone types into. Fair ranking and attribution govern how your pages are surfaced, how they are reused, and what happens when your content is summarised inside an answer rather than clicked. If you write for a living or your website is your main sales channel, those are the consultations to read. We cover the publishing side of that in our notes on Google's AI content guidelines for publishers.
One more wrinkle. A small share shift hurts narrow commercial pages more than broad informational ones. If a page earns only a handful of clicks a month on a high-intent query like "emergency boiler repair Reading", losing even a small percentage of those clicks is an enquiry you can feel. A guide pulling 3,000 impressions a month absorbs the same percentage without anyone noticing.
Set your baseline before anything rolls out
This is the work that actually pays, and almost nobody does it until it is too late to be useful.
Export the before picture
Pull a Search Console performance export split by device and by country, covering at least the last 12 months. Then build a GA4 table of sessions by source and medium for the same period, and write down, in plain numbers, what share already arrives from Bing, DuckDuckGo, Ecosia and direct. Most SMEs have never looked. The combined share is usually small, often only a few percent, and knowing your actual figure is the whole point.
Save both exports somewhere a human will find them in a year. A screenshot in a Slack thread does not count.
Split your rank tracking by device
Most SME rank tracking quietly assumes one engine and one device: a weekly desktop scan on google.co.uk, scheduled in 2021, never revisited. If a slice of mobile queries moves to another provider, that scan still reports healthy green numbers while your enquiry form goes quiet. Treat it as two jobs. Track mobile rankings separately from desktop, because the screens under discussion are mobile-first, and the gap between the two is where the early signal will show. Our piece on what rank tracking numbers do and do not tell you goes further into why averaged positions hide this sort of thing.
Learn to tell the two causes apart
A choice screen effect and a core update look nothing alike once you know where to look. A share shift shows up as fewer clicks from one engine while your impressions and average position for the same queries hold steady: the pages still rank, fewer people are asking that engine. A core update shows up as position loss concentrated in specific query clusters or page types, with impressions falling alongside clicks.
Check device split too. A regulatory shift should bite on Android mobile first and leave desktop broadly alone. An algorithm change rarely respects that line. If you are working through a suspected ranking drop, our walkthrough on diagnosing an unconfirmed Google update covers the sequence. Get this wrong and you rewrite perfectly good pages chasing the wrong cause, which we see constantly and which costs weeks.
The practical checklist for a lean team
Here is the good news. Several alternative engines lean on Bing's index rather than crawling the web independently, so "preparing for choice screens"is mostly an indexation job. It takes an afternoon and costs nothing.
Verify in Bing Webmaster Tools and submit your sitemap there. Not only to Google. You can import your Search Console properties directly, which saves the DNS faff.
Turn on IndexNow. It pings participating engines when you publish or update a page, rather than waiting on a crawl that might take a fortnight. Most CMS plugins support it with a tick box.
Read your robots.txt properly. We find blocked bots on client sites regularly, usually inherited from an old developer or a security plugin. Check bingbot, DuckDuckBot and Applebot are not quietly excluded, and check any firewall or bot-management rules at the CDN layer too, because that is where the real blocking usually lives.
Keep your Article, Breadcrumb and FAQ schema clean and un-duplicated. If you run an SEO plugin such as Yoast, confirm it is not already emitting the same markup your theme is adding. Two competing Article blocks on one page is worse than none.
Keep publishing to the same cadence. A consultation is not a content strategy.
That last point is the one we would defend hardest. The content that earns a click on a non-Google engine is the same content that earns one on Google: specific, first-hand, and genuinely answering the query someone typed. Nothing in the CMA programme rewards thinner or faster publishing. If anything, attribution requirements push the other way, towards work with a name and a point of view attached to it, which is how we write everything anyway.
Frequently Asked Questions
What are the CMA's Google search choice screens?
They are a proposed conduct requirement under the Digital Markets, Competition and Consumers Act 2024, which would make Google present UK users with a prompt to pick their search provider rather than defaulting silently to Google. The screens under discussion apply to Android device setup and the Chrome browser. They sit alongside other measures the CMA has signalled, including fair ranking and publisher content attribution.
When will UK users start seeing a search choice screen?
There is no single confirmed date, and the dates quoted in news coverage usually refer to different stages: a proposal, a consultation, a final decision, or a go-live. Each conduct requirement is consulted on separately after designation, so rollout follows the final decision on that specific measure. Check the CMA's case page on gov.uk for the current position before you plan around any date.
Will choice screens reduce my Google traffic?
Probably a little, over time, rather than suddenly. The EU has run choice screens on Android since 2020 and under the Digital Markets Act since March 2024, and alternative engines reported install spikes rather than lasting share change. With StatCounter showing Google well above 90% of UK search referrals for years, a few percentage points redistributing is the realistic scenario, and narrow commercial pages will feel it before broad guides do.
Should I start optimising for Bing and DuckDuckGo now?
Not optimising, exactly. Indexing. Verify in Bing Webmaster Tools, submit your sitemap, enable IndexNow and make sure no bot rules are blocking bingbot, DuckDuckBot or Applebot. Several alternative engines draw on Bing's index, so that afternoon of plumbing covers most of them, and you do not need a separate content strategy on top.
How do I tell a choice screen effect from a Google core update?
Look at impressions and device split. A share shift shows clicks falling from one engine while impressions and average position for the same queries hold, concentrated on mobile. A core update shows position loss on specific query clusters across devices, with impressions dropping alongside clicks. You can only run that comparison if you exported a clean baseline beforehand, which is why we keep telling people to do it now.
BEHIND THIS ARTICLE
Get your free content marketing snapshot
Enter your website and see the blogs and social posts Awareness Marketing would create for your business. Free, no account needed.
Enter your website and see the blogs and social posts Awareness Marketing would create for your business. Free, no account needed.
See What Weโd Create

