Olivia
Online
Alex | Google + Meta Ads
@AlexMartkovich
Google + Meta Ads for e-commerce and businesses that need calls, leads, and booked jobs. Founder, New Digital. Free account teardown ↓
Toronto, Ontario
Joined July 2023
257 Following
56 Followers
196 Posts
Pinned Tweet
Real estate agency. Google Ads were on, leads were not.
We rebuilt search terms, negatives, and the landing path.
Conversions +250% week over week. CPA stayed flat.
If your account gets clicks and the phone does not ring, I will look at it for 15 minutes and tell you the first thing I would cut. Book here: https://t.co/N1aRWsg2St
Claim your business in ChatGPT is finally here.
Now in beta: Businesses in ChatGPT.
Claim your business, manage your information, and get insights on how people engage with you.
Eligible businesses can also build a Business Agent to answer customers’ questions and help them explore what you offer.
https://t.co/N2GEZX0mFT
Google hides 50%+ of your search terms & [Exact] match is dead.
Just last week, one of my clients generated 9 high-ticket conversions in a single day, yet the Search Terms report showed only 1 search term associated with 1 conversion.
As someone who has personally managed $2M-$3M+ per month in Google Ads spend for years, I can tell you the problem goes much deeper than that.
If you've been wondering why your Google Ads campaigns are performing worse and worse, read on.
- Less accurate search term reporting
Until 2020, advertisers could see essentially all search terms that triggered impressions and clicks.
Starting with the September 2020 update, Google began withholding low-volume or privacy-sensitive queries. Queries that fall below Google's internal thresholds are not disclosed individually and are instead included in aggregate reporting.
Today, visibility varies by account, match type, and search volume. Many advertisers report seeing search terms for only around 50% of their Search campaign traffic, sometimes less in broader-match or PMax campaigns, particularly in accounts with substantial long-tail or broad-match traffic.
Early analyses saw ~28% of budget disappear overnight. A lot of accounts now have 50–90% of spend stuck in “Other search terms.
The bigger problem is all the traffic that DID NOT convert.
You have no idea what many of those searches were or what action to take.
Large-scale analyses (including analyses of tens of millions in spend) have found that these hidden terms often carry higher CPCs and lower conversion rates than the terms you can still see.
A November 2026 analysis of more than 400 accounts found that AI Mode activity (which Google says produces roughly 3-times-longer queries) is nearly invisible in the search terms report.
Google's stated reason is user privacy. Rare or unique queries could theoretically identify someone. Conveniently, this also prevents advertisers from identifying and eliminating irrelevant or poorly performing search terms that never convert, allowing Google to continue charging for traffic advertisers don't need.
- Exact match keywords aren't that exact anymore
To make matters worse, Google has made exact match increasingly less exact.
In addition to reducing search term visibility, Google has relaxed keyword match type restrictions and introduced AI-powered campaign features that further reduce advertisers' control over keyword targeting.
Google used to offer precise keyword targeting that allowed advertisers to select individual search phrases and bid on them, with exact and phrase match providing much tighter control over which searches could trigger their ads.
This was one of the core advantages of Google Search Ads. Advertisers could run highly targeted campaigns based on exactly what potential customers typed into Google.
I had a client who wanted their ads to appear only when someone searched for one specific phrase related to their service, and the campaign did exactly that.
Today, even when you bid on a specific keyword using the strictest exact match setting, Google can show your ads for numerous other searches it considers to have the same meaning or intent, labeling them as "exact match (close variant)".
No switch to turn close variants off. They hit every match type by default.
Google has expanded close variants over the years, particularly since 2021, with AI and intent-based matching further broadening how keywords are interpreted.
As a result, exact match keywords can attract irrelevant or loosely related traffic, such as "vegan leather wallet" for [leather wallet], research-oriented searches, or other queries that don't align with the advertiser's commercial intent, and advertisers pay for that traffic.
In one detailed account audit, 71% of phrase-match spend went to close variants and those queries cost 45% more per lead than the literal matches.
In practice, exact match increasingly behaves like the old broad match, while broad match can become an extremely expensive source of irrelevant traffic.
Advertisers can identify and partially fix these issues through the Search Terms report, which makes the declining transparency of that report even more problematic.
Negative keywords do not automatically block close variants the same way positive keywords match them, so singulars, plurals, and stems often still need to be added manually.
- Is the 2nd-price auction model gone?
Google still describes the Search auction as a generalized 2nd-price model (you normally pay just enough to beat the next Ad Rank), while the well-known switch to 1st-price applied mainly to display.
In a simplified example, if you bid $10 per click and the next-highest bidder bid $1, you might pay approximately $1.01 rather than your full $10 bid. Actual Google Ads pricing also accounts for factors such as Ad Rank and ad quality.
This was a clever strategy by Google because it encouraged advertisers to bid their true maximum willingness to pay instead of constantly adjusting lower bids to narrowly outbid competitors without overpaying.
Google now appears to be moving further away from traditional 2nd-price auction dynamics, with advertisers sometimes paying closer to their maximum bids, seemingly without a corresponding increase in competitive pressure.
The lack of auction transparency makes it difficult to determine exactly how much of this change is attributable to auction mechanics versus competition, quality signals, and automated bidding.
- Does this mean Google Ads no longer works?
Absolutely not.
Google Ads remains one of the most powerful sources of targeted, high-intent traffic available. For many businesses, there is simply no comparable alternative at the same scale.
But it has become increasingly important to understand how to build, structure, and manage campaigns properly.
The right campaign structure, campaign types, bidding strategies, targeting settings, and optimization approach depend heavily on your business, products, budget, and goals. That includes knowing which automated features to enable and which to disable.
Practical controls that still work include regular use of Search Terms Insights (which surfaces themes for hidden queries), cross-checking landing-page queries in Google Search Console, and turning off Search Partners and Display expansion when tighter control is required.
What works for a business spending $200,000-$500,000 per month on Google Ads could bankrupt a business spending $5,000-$15,000 per month.
All of this is bad for advertisers, but ultimately, it’s bad for Google too.
Advertisers who cannot regain control over their campaigns, and doing so is becoming increasingly difficult, are more likely to reduce their Google Ads budgets or consider leaving the platform altogether.
What share of your Google Ads search terms can you actually see in your account today?
If you are a business owner and want me to check your account for this, send me a DM.
![AlexMartkovich's tweet photo. Google hides 50%+ of your search terms & [Exact] match is dead.
Just last week, one of my clients generated 9 high-ticket conversions in a single day, yet the Search Terms report showed only 1 search term associated with 1 conversion.
As someone who has personally managed $2M-$3M+ per month in Google Ads spend for years, I can tell you the problem goes much deeper than that.
If you've been wondering why your Google Ads campaigns are performing worse and worse, read on.
- Less accurate search term reporting
Until 2020, advertisers could see essentially all search terms that triggered impressions and clicks.
Starting with the September 2020 update, Google began withholding low-volume or privacy-sensitive queries. Queries that fall below Google's internal thresholds are not disclosed individually and are instead included in aggregate reporting.
Today, visibility varies by account, match type, and search volume. Many advertisers report seeing search terms for only around 50% of their Search campaign traffic, sometimes less in broader-match or PMax campaigns, particularly in accounts with substantial long-tail or broad-match traffic.
Early analyses saw ~28% of budget disappear overnight. A lot of accounts now have 50–90% of spend stuck in “Other search terms.
The bigger problem is all the traffic that DID NOT convert.
You have no idea what many of those searches were or what action to take.
Large-scale analyses (including analyses of tens of millions in spend) have found that these hidden terms often carry higher CPCs and lower conversion rates than the terms you can still see.
A November 2026 analysis of more than 400 accounts found that AI Mode activity (which Google says produces roughly 3-times-longer queries) is nearly invisible in the search terms report.
Google's stated reason is user privacy. Rare or unique queries could theoretically identify someone. Conveniently, this also prevents advertisers from identifying and eliminating irrelevant or poorly performing search terms that never convert, allowing Google to continue charging for traffic advertisers don't need.
- Exact match keywords aren't that exact anymore
To make matters worse, Google has made exact match increasingly less exact.
In addition to reducing search term visibility, Google has relaxed keyword match type restrictions and introduced AI-powered campaign features that further reduce advertisers' control over keyword targeting.
Google used to offer precise keyword targeting that allowed advertisers to select individual search phrases and bid on them, with exact and phrase match providing much tighter control over which searches could trigger their ads.
This was one of the core advantages of Google Search Ads. Advertisers could run highly targeted campaigns based on exactly what potential customers typed into Google.
I had a client who wanted their ads to appear only when someone searched for one specific phrase related to their service, and the campaign did exactly that.
Today, even when you bid on a specific keyword using the strictest exact match setting, Google can show your ads for numerous other searches it considers to have the same meaning or intent, labeling them as "exact match (close variant)".
No switch to turn close variants off. They hit every match type by default.
Google has expanded close variants over the years, particularly since 2021, with AI and intent-based matching further broadening how keywords are interpreted.
As a result, exact match keywords can attract irrelevant or loosely related traffic, such as "vegan leather wallet" for [leather wallet], research-oriented searches, or other queries that don't align with the advertiser's commercial intent, and advertisers pay for that traffic.
In one detailed account audit, 71% of phrase-match spend went to close variants and those queries cost 45% more per lead than the literal matches.
In practice, exact match increasingly behaves like the old broad match, while broad match can become an extremely expensive source of irrelevant traffic.
Advertisers can identify and partially fix these issues through the Search Terms report, which makes the declining transparency of that report even more problematic.
Negative keywords do not automatically block close variants the same way positive keywords match them, so singulars, plurals, and stems often still need to be added manually.
- Is the 2nd-price auction model gone?
Google still describes the Search auction as a generalized 2nd-price model (you normally pay just enough to beat the next Ad Rank), while the well-known switch to 1st-price applied mainly to display.
In a simplified example, if you bid $10 per click and the next-highest bidder bid $1, you might pay approximately $1.01 rather than your full $10 bid. Actual Google Ads pricing also accounts for factors such as Ad Rank and ad quality.
This was a clever strategy by Google because it encouraged advertisers to bid their true maximum willingness to pay instead of constantly adjusting lower bids to narrowly outbid competitors without overpaying.
Google now appears to be moving further away from traditional 2nd-price auction dynamics, with advertisers sometimes paying closer to their maximum bids, seemingly without a corresponding increase in competitive pressure.
The lack of auction transparency makes it difficult to determine exactly how much of this change is attributable to auction mechanics versus competition, quality signals, and automated bidding.
- Does this mean Google Ads no longer works?
Absolutely not.
Google Ads remains one of the most powerful sources of targeted, high-intent traffic available. For many businesses, there is simply no comparable alternative at the same scale.
But it has become increasingly important to understand how to build, structure, and manage campaigns properly.
The right campaign structure, campaign types, bidding strategies, targeting settings, and optimization approach depend heavily on your business, products, budget, and goals. That includes knowing which automated features to enable and which to disable.
Practical controls that still work include regular use of Search Terms Insights (which surfaces themes for hidden queries), cross-checking landing-page queries in Google Search Console, and turning off Search Partners and Display expansion when tighter control is required.
What works for a business spending $200,000-$500,000 per month on Google Ads could bankrupt a business spending $5,000-$15,000 per month.
All of this is bad for advertisers, but ultimately, it’s bad for Google too.
Advertisers who cannot regain control over their campaigns, and doing so is becoming increasingly difficult, are more likely to reduce their Google Ads budgets or consider leaving the platform altogether.
What share of your Google Ads search terms can you actually see in your account today?
If you are a business owner and want me to check your account for this, send me a DM.](https://pbs.twimg.com/media/HWxNRJiXYAAXoQb.jpg)
AI content disclosures in California
Google now shows AI generated content disclosures on ads in California, South Korea and Vietnam, on top of New York, the EU and India.
Designated AI created or AI edited assets can carry a visible label right on the ad. The rule covers Search, YouTube and Discover campaigns. Google presents it as active and gives no separate effective date.
Google generated assets can be labeled automatically, and those labels cannot always be overridden. That is the part I would pay attention to for US clients targeting California or New York.
In client accounts I would do 4 things.
- Add the AI label column to asset reporting and audit AI created assets in US campaigns
- Record the disclosure decision in creative intake and approval
- Keep any custom disclosure away from image edges, because automated cropping can remove it
- Review Performance Max and Demand Gen enhancements that may create or change assets on their own
Do you know which of your current ad assets were made or edited by AI?
#GoogleAds #PPC #AIAds #PMax #PerformanceMax
@glenngabe We saw a significant decline in clicks and click-through rate while impressions grew.
@herrmanndigital @MetaforBusiness We see the same reporting trap, so we decide spend on click based columns before peak season.
@blvckledge Problem searches combined with a negative list for research only terms that never convert worked well for many of our clients!
@shivamA2k You def need to run a search campaign. I'd ignore Google's broad and phrase match nudges. Low-volume exact match like "driving lessons [suburb]" often brings the most targeted bookings. How many suburbs are you covering?
Pretty common now. With so many bots and hacks, Meta re-checks even verified accounts. Worth checking Account Quality in Business Suite. If the restriction sits on your personal profile rather than the ad account, the review option shows up there, not on the ad account. What do you advertise?
@sohaibgou Over 1900 projects delivered with Google and Facebook Ads. Happy to join the team. What products are you selling across Google and Meta?
If you want to know how your business shows up in AI search today, you can reach me at https://t.co/m1T94HyvA6
https://t.co/Gg3lhMkaZr
@VengeonsP Until it is live you can simply continue with AIO/SEO efforts.
@reallygabriella Delivered over 200 Meta ads projects across the US. Happy to join the team. Are you optimizing for demos booked, free trials, or paid signups?
Google Ads can now show up in follow up questions inside AI Mode in the US, not only after the first search.
That means one person asking several questions in a row can see ads more than once.
Search, Shopping and Performance Max campaigns are eligible automatically. There is no separate targeting, no separate exclusion and no separate budget for AI Mode. Its traffic sits inside your normal reports, search terms included.
Google can match ads to the AI answer as well as the question. Broad match, AI Max, Shopping and PMax qualify. Phrase and exact match only appear when the question itself shows clear buying intent.
Next to your ad, Google may add an AI explainer drawn from the conversation, your website and other public sources. You have no way to edit or switch it off.
What I would check in a US account
1. Search terms with longer, question style wording
2. Negative keywords, the searches you choose to block
3. Website claims, since Google may quote them beside the ad
4. Conversion tracking, before testing broad match or AI Max
AI Mode results cannot be split out, so I would not credit or blame it for a change in volume unless the search terms and timing show it.
Have your search terms started to look more like full questions?
@Charles_SEO We have seen AI engines cite a brand within a few hours after a focused press release + a matching entity page.
@ericosiu On some paid media audits we often find a ~60% drop from clicks to real link clicks before the page even loads.
@bloggersarvesh Agree, generic claims are invisible.
@EcomShri Vertical scaling on winners is the way to go. I'd also add a custom label for products with no spend after 14 days.
Last Seen Users on Sotwe
Most Popular Users
Elon Musk
@elonmusk
241.4M followers
Barack Obama
@barackobama
118.8M followers
Cristiano Ronaldo
@cristiano
114.5M followers
Donald J. Trump
@realdonaldtrump
111.9M followers
Narendra Modi
@narendramodi
107M followers
Rihanna
@rihanna
98.7M followers
NASA
@nasa
92.3M followers
Justin Bieber
@justinbieber
91.8M followers
KATY PERRY
@katyperry
90M followers
Taylor Swift
@taylorswift13
84M followers
Lady Gaga
@ladygaga
75.5M followers
Virat Kohli
@imvkohli
73.4M followers
Kim Kardashian
@kimkardashian
70.8M followers
YouTube
@youtube
68.7M followers
Neymar Jr
@neymarjr
66.4M followers
Bill Gates
@billgates
65.2M followers
Selena Gomez
@selenagomez
63.1M followers
The Ellen Show
@theellenshow
62.2M followers
CNN
@cnn
61.7M followers
X
@x
60.7M followers