During this week’s PPCChat session, host Boris Beceric sought experts’ views on Colin Slattery‘s article on hidden search terms, performance gap on visible and hidden search terms, tactics used to limit wasted spend from hidden terms and more
Q1: Did you read Colin Slattery’s article on hidden search terms in Google Ads? What were your key takeaways or reactions?
Yes – an excellent and thorough analysis! Key takeaways are that, as we all suspected, hidden search terms are generally poor-performing terms, and we need a way to opt out of these. @beyondthepaid
Yes, seems like another way for Google to waste our money. @AnthonyVillanueva
I thought Collin did some heyvy lifting by pulling all that data (you can’t pull just from the MCC).
I wasn’t surprised to see the analysis, it confirmed what I’d suspected all along. It also made me do similar analysis on my accounts (ongoing) – for reference, monthly spend in the 7 figures. @BorisBeceric
Yes, completely agree. For some small accounts with high CPC’s – hidden search terms end up being an additional tax for having small budgets with no transparency. @alimehdimukadam
Yes. Really well thought out and super interesting conclusions. @JuliaVyse
I still see super long tail “there is no way more than one person searched this” queries in my SQR though, so I’m not sure what to make of the findings/theory. @revaminkoff
Exactly @revaminkoff– there is no rhyme or reason to what’s hidden and what’s not. @beyondthepaid
@revaminkoff yes, the whole “privacy” thing doesn’t sit well with me. Like I can see searches for “mortgage low credit score and here is my social security number” with 1 impression, but somehow I can’t see the costly searches that didn’t convert? @BorisBeceric
Right – I don’t think it’s “privacy” accordingly? Or I don’t think privacy explains all of it, I guess. @revaminkoff
It’s so easy to just assume nefarious reasons at this point. And even if it’s indeed privacy, Google has never given a better argument than just “trust me bro”. @BorisBeceric
It was an incredible way of putting numbers behind (everyone’s) suspicions. It was entertaining, well-written, and highly informative. Google could react, but my gut is that this “privacy curtain” will hold and provide the shield they need to continue extracting revenue. At least we all know now, with data in our collective back pockets. @teabeeshell
well and the old adage – never assume nefarious intent when simple incompetence explains it. Google has a galactic dataset to work with daily, so running a dumb algo that sweeps some things but not others WHILE gemini is moving into AI overviews and parsing search terms means arbitrary weirdness is very likely. @JuliaVyse
@JuliaVyse dunno, I always assumed that Google, with the highest density of PhDs in the world, would be able to make more sense of the giant pile of data they’re sitting on. @BorisBeceric
Q2: Performance Gaps: Colin’s data showed a big performance gap on visible vs hidden search terms. Have you observed similar gaps in your own accounts?
No, but my assumption is that on straight lead gen or ecom, it’s more apparent than in my foot-traffic attributed accounts. @JuliaVyse
Yes – it’s something I’ve noticed for 15+ years at this point though. Other has always had lower CTR and higher CPAs than whatever is shown. @revaminkoff
I see varied performance between accounts when it comes to visible vs other search terms.
Most notably, it’s much worse on high CPC, high competition lead gen accounts. @revaminkoff and it’s almost always on lead gen accounts? @BorisBeceric
Absolutely. But not always in the same direction. Sometimes the hidden terms are the only ones with conversions; other times they’re complete junk. @beyondthepaid
Yes, for some accounts and that’s where it gets infuriating if your hidden search terms have 30% of cost and contribute only 10% of conversions as an example. @alimehdimukadam
In broad strokes, within “Other search terms” in mostly DTC/ecomm I’ve seen:
- Higher CPCs & Lower CTR (somewhat vanity metrics)
- Lower CRs
- Lower CPAs
- Lower ROAS
The argument (sometimes made by Google) is that these “other” terms represent incremental revenue/conversions, albeit at a higher cost. My questions remains: Have you exhausted ALL available search queries/auctions within my bid and budget parameters BEFORE going this direction? If yes, I’m grateful. If not, I’m immensely frustrated. @teabeeshell
@BorisBeceric Not necessarily – two of the big examples I’m thinking of were eCom. @revaminkoff
@beyondthepaid ah that’s so frustrating! Just show me what’s working so I can double down on that! @BorisBeceric
@alimehdimukadam we’ll get to the % of hidden in another question, but yes that’s broadly what I’m seeing as well. @BorisBeceric
@teabeeshell do you think this has to do with the push towards broad match? @BorisBeceric
When 80-90% of search terms are hidden (this is B2B lead gen) it’s super frustrating. @beyondthepaid
@beyondthepaid sometimes I think the privacy argument could hold value when it’s in a sensitive industry with strict policy etc. @BorisBeceric
@BorisBeceric – If broad match works (as designed) within true BM keywords or via AI Max, the system should get smarter over time. I’m not seeing this, either in the search terms themselves or in the hard, calculated data. It seems more opportunistic, in the sense that Google realizes that there are queries they can mask for profit. If Google only made money on the smartest advertisers, they would be looking at much smaller revenue figures each quarter. @teabeeshell
Q3: Cost of the Hidden: The article says 30% of spend often goes to hidden terms. How much of your spend do “Other search terms” now represent?
I don’t have a percentage to share, but the team is working on it! @JuliaVyse
In ecomm, I’m seeing closer to 20-23%, but in a way, it’s semantics. IMO, all search terms should be visible. My retort to Google’s “we mask for privacy” is always: What could an advertiser reverse-engineer, within a given search query, to determine personally identifying information about a given consumer? What use cases are there to support this behavior? Obviously, Google is under no obligation to divulge this, but it adds to the lunacy of their rationale. @teabeeshell
I’m just now looking at a couple accounts again, and the range is between 30% – 50%. BUT I have seen an account recently where it was 10 / 90 visible/hidden. @BorisBeceric
80-90% @beyondthepaid
Seems like there’s a big divide between lead gen and ecomm. @teabeeshell
Always! Google hates B2B @beyondthepaid
For reference, all the accounts I looked at are lead gen. @BorisBeceric
Referencing a could of lead gen accounts, I’m seeing 35-40% @teabeeshell
@teabeeshell I don’t know about ALL search terms. I can imagine scenarios in local lead gen for specific industries where I could deduce from a search term who that person might be. But that would almost always involve intimate knowledge of a neighbourhood & people living there. So yeah…fringe cases basically. @BorisBeceric
Speaking as someone who runs addiction support and mental health support campaigns, I could reverse engineer quite a bit with the right location data. I do like a world where Google does more in terms of privacy. But as usual, their methods are for the birds. @JuliaVyse
Slight detour, but isn’t it funny how it’s “privacy for me, but not for thee”? We get reduced visibility, privacy laws, etc, while Google does fingerprinting, etc. We always get the short stick. @BorisBeceric
YES! I work with govs and non-profits that are SO CAREFUL with everything, and the most vulnerable populations who we support are just mined. Just mined for data because what choice is there? @JuliaVyse
And don’t get me wrong, privacy is important! @BorisBeceric
20-35% @revaminkoff
It’s different for different accounts. I have seen hidden search terms account for 10-20% of the terms and one some days account for up to 60% of the cost. Average: 10-30. @alimehdimukadam
@JuliaVyse & @BorisBeceric That makes a lot of sense, hadn’t considered those verticals. There are other ways around the negativity Google has inspired: We make judgment calls when people reference their own names and/or ages. Here’s a (neutral) sample of some of the queries we’re masking. @teabeeshell
I have seen zip codes pop up in searches before. @revaminkoff
@JuliaVyse I think it’s also important for us as consumers to be mindful of the data we share willingly or unwillingly. And yes, I do realize that not everyone has either the ability or even possibility to do so, or to divest from specific tech platforms/companies/ecosystems. At this point, we’re pretty locked in. @BorisBeceric
A previous response I recall getting from Google is, “When there are not enough queries, statistically speaking, we often withhold these queries from our reporting.” This is easily debunked with the single-impression queries we see, which is why I belief Google changed their narrative. @teabeeshell
@revaminkoff OH I’ve seen some revealing searches! People are…interesing. @BorisBeceric
@teabeeshell I feel like what I’ve seen still reflects that narrative about the statistical significance and therefore it’s removed for privacy reasons @teabeeshell But I agree that all the 1 impressions searches that DO show up make me think that’s not actually true. @revaminkoff
@teabeeshell I believe the search term insights are Google’s way of pushing us towards acceptance of less visibility into data. @BorisBeceric
Just to keep 100, my old client was Viagra from Pfizer. A key point was that people using this medication did NOT want people knowing about it. Even discussing the ailment Viagra treats was treated with such care and sensitivity. There are better ways to approach privacy than hiding 1:1 search terms from people actually asking for help. @JuliaVyse
@BorisBeceric – Agreed. Getting advertisers used to “Insights” vs. “Search terms” is a way to directionally provide what’s being requested without divulging the precision requested. @teabeeshell
In all fairness, I just checked one of my Bing campaigns and the total SQR report spend > than the actual search ads spend total, so… ? @revaminkoff
Q4: Optimization Impact: How does the lack of visibility affect your ability to optimize?
On a simple level, it’s harder to mine negative keywords and optimize spend. @JuliaVyse
With so much data hidden in B2B accounts, we’re not able to efficiently learn what themes are performing and what aren’t. Is our messaging hitting the right topics? Are people confused about what we offer? Search terms offer insight into these things, and without them, we’re flying blind. Then again, in a keywordless world, this won’t matter haha @beyondthepaid
I find that I leverage Negative Keywords more than many advertisers. (I know we all use them, but based on “inherited accounts” and conversations with many, it seems I’m in the minority. I’m talking 10+ lists, each with 1,000s of KWs.) Search terms help unearth, with precision, intent-based phrasing to avoid. Most brands would agree that they shouldn’t pay Google (too much) to “learn” what’s accurate in terms of converting queries. Adding negative terms from actual search phrasing is the fastest way to avoid this. Through the lens of Google, negative KWs should be treated as a valuable “signal” from advertisers, but it’s not as welcome as we’d hope. @teabeeshell
Sometimes it almost feels like flying blind. For example, I still do n-gram analysis, but the amount of data you get back nowadays often doesn’t allow you drawing conclusions. It just makes all aspects of account management harder. I guess all parts of an ad account are now viewed as “inputs” (keywords, ads, landing pages, conversion tracking etc). Negative keywords should be an important signal to Google as well. @BorisBeceric
I’ll have clients who want to invest in what’s working, but the conversion came from a search query that Google isn’t showing, so we get stuck… @revaminkoff
@teabeeshell yep, I used to regularly run into negative keyword list limits and had to split them up into multiple lists. One thing (theory) I recently read is that smart bidding “should” take care of unwanted search queries as otherwise it can’t reach it’s target. While I think that might be partially true for D2C accounts, the opposite holds true for lead gen – you just gotta be ruthless with negative and make the budget work extra hard. @teabeeshell
Q5: Mitigation Tactics: What tactics have you used to limit wasted spend from hidden terms?
At this point, it’s still mostly a ruthless attention to the SQR report and adding negatives, hoping this translates to both visible and hidden. Occasionally, I’ll revise copy in accordance to findings, but that’s rarer when good copy is employed at the onset. @teabeeshell
Exact match focus usually helps a bit. Even though that also comes with the downside of (anecdotally speaking) lower search volumes. The dreaded “low search volume label shows up more and more. I’ve also started to try and have an LLM come up with an extensive as possible keyword list BEFORE I launch an account/campaign. Just can’t afford to pay first so we can learn what’s working and what’s not. Solid tracking also helps. @BorisBeceric
Where possible, I try to pull as many search terms from Console as possible. It’s not perfect, but at least with brand and top-ten generic searches, you can triangulate. @JuliaVyse
I try to think about negatives in a more proactive way than just adding that exact query in the SQR as an exact match negative to try to pick up what I’m not seeing. @revaminkoff
Exact match and negatives still work @beyondthepaid
@beyondthepaid don’t tell the PMAX & broad match kids . @BorisBeceric
@BorisBeceric i recently did this for a client we were onboarding, while my colleague was manually working up a list of negatives manually. colleague’s manual list blew the LLM list out of the water in terms of quantity of negatives, but the LLM did come up with some that we missed! so nice to supplement our usual method of excluding the junk before launching. @janinemonico
PPCChat Participants
- Melissa L Mackey @beyondthepaid
- Anthony Vilanueva @AnthonyVillanueva
- Boris Beceric @BorisBeceric
- Ali Mehdi Mukadam @alimehdimukadam
- Julia Vyse @JuliaVyse
- Reva Minkoff @revaminkoff
- Travis @teabeeshell
- Janine Monico @janinemonico
Related Links
Stop the wasted ad spend. Get more conversions from the same ad budget.
Our customers save over $16 Million per year on Google and Amazon Ads.