Posted by & filed under PPCChat.

During this week’s PPCChat session, host Julie F Bacchini had Navah Hopkins from Optmyzr as a special guest who talked about their new study on pausing low-activity keywords. Her study can be found here.

Q1: Have you heard and/or read about Google automatically pausing “low activity keywords”?

For reference – Low-activity keywords are considered those that were created over 13 months ago and have zero impressions over the past 13 months. Documentation here: @NeptuneMoon

I got the email from Google about 10 times, so I definitely saw it. @robert_brady

Yes, I was aware of this and also read the Optmyzr study yesterday. @NeptuneMoon

I may have heard something about it…What’s really interesting is that when this first came out, I was thrilled and it felt like a true shock that folks would be upset about it. Collecting that feedback and understanding the friction was really helpful (especially for smaller accounts). @navahf

I hadn’t heard about it until Navah posted, but it seemed very in line with some of G’s initiatives over the past, oh, 7 or so years? @ferkungamaboobo

Interesting @ferkungamaboobo – to me it felt new. Are you thinking about the phasing out of modified broad? @navahf

Well Google does like to “make our lives easier” @NeptuneMoon

One of the findings that really surprised me is how many “small” accounts had low activity keywords. @navahf

I’d put this in the “Google knows best” category of moves they have made. @NeptuneMoon

I read Navah’s post and also got a few emails from Google, so feel a part of the conversations. @revaminkoff

So like, one thing I constantly think about in a similar context is the automatic flagging of overlapping keywords. @ferkungamaboobo

Yes! and that keywords just stop serving. @navahf

I’ll admit I was surprised how many “small” accounts had low activity keywords @navahf

Given that in theory most accounts would migrate to a small structure (less than 400 words) @navahf

I would argue that “low activity” and/or “low volume” search terms have been on the rise of late generally. @NeptuneMoon

Definitely seeing Google favor some keywords over others, despite what’s in the campaign. @revaminkoff

It is important to give Google the benefit of the doubt that low volume and low activity are different. @navahf

Yes, although low volume will become low activity cause they don’t serve…@NeptuneMoon

One is a keyword that could have done things but didn’t, while the other has calculated low search volume (and isn’t held against you) @navahf

If you’re bidding competitively and there are 0 impressions over a 13 month period, what is the justification for keeping that keyword active? @robert_brady

you hit the nail on the head @robert_brady – a lot of times the bidding (or bidding strategy) is the problem. @navahf

Yeah, I’m a bit baffled at the pushback – this just seems like a nice little automatic cleanup thing. @ferkungamaboobo

@robert_brady Sometimes the client just needs to see/know it’s active, regardless of whether Google is serving it/it has any volume, etc. @revaminkoff

I guess my question is really why should Google care to the point of pausing for me if there are low activity keywords? @NeptuneMoon

The language from Google has been – for at least 5 years now – that cruft does not help your account. @ferkungamaboobo

@revaminkoff I cited that in the “if this then that” on whether to keep it live. @navahf

@navahf  – I saw that! Really appreciated it. Have been meaning to write more about it too. @revaminkoff

I can’t stay in the chat long, but here is my tin-foil hat theory on this: Google is trying to cut down on computational lift. The fewer keywords it has to calculate Ad Rank for in a given auction, the “cheaper” it is for them to run that auction. Plain and simple, it always comes down to the $$$ with Google these days. @robert_brady

I have similar theories about GA4 @robert_brady – Big G does not necessarily want to be the repository for the web’s performance data. Both for privacy regulation concerns & sheer cost. @NeptuneMoon

Is ad rank really that complex though? @ferkungamaboobo

You can’t tell me that a company running its own LLMs is worried about CPU usage @ferkungamaboobo

I wish folks would have been given more clarity on why changes are made. @navahf

When we understand we shed fear @navahf

@ferkungamaboobo If you multiply the volume of searches they run each day and can cut 10% of the keywords eligible from everyone, then it might add up. @robert_brady

That might be a good question for the Ads Liaison — might be a fun factoid press release if nothing else, or an opportunity to put into context of net 0 efforts @ferkungamaboobo

Google doesn’t feel it needs to really explain this stuff – as evidenced by their comms on these types of things. Which is their choice. But it leaves a lot of room for misinterpretation, misunderstanding, etc. @NeptuneMoon

@ferkungamaboobo Well if nothing else we’ll be able to report back with round two on what happened to these accounts. @navahf

The thing which always boggles me when it comes to things like this is that Google always tells me that I should pause my exact Brand term… Even when it’s driving a high level of exact traffic (Redundant Keywords in recommendations) So like is Google going to somehow ruin my brand campaign..@Meriem

But if it’s 0 Impressions then really ultimately why are they even in your account still after that long if they aren’t driving impressions!? @Meriem

@Meriem that’s super interesting, regarding branded.@navahf

I think a lot of us have a “well if it’s not hurting anything just leave it just in case” mentality — and i think based on some of the discussions I’ve had with Google reps of different flavours “what’s the harm” isn’t the question google wants us to be asking, they want us to focus on “what’s the benefit?” @ferkungamaboobo

@ferkungamaboobo 100% agree.@navahf

So like, w/r/t redundant terms — if you’re using automatic bids with each, the harm is that your true intent of the bid in the [modified but let’s ignore that rn] 2nd price auction isn’t being reflected — and the game theory answer to a 2nd price auction is to bid what the thing is worth to you instead of gaming it. @ferkungamaboobo

a huge issue with Google’s auction is that no one actually bids what something is worth to them @ferkungamaboobo

Like, idk, back in the day, I used to train “bid for position 1 [except in niches where that meant $250 a click and idk even then]” @ferkungamaboobo

I feel like I was taught that if a keyword hasn’t had impressions in over 3 months why is it there, review, amend, clean up. Back in the day with the cross-match negatives stuff it used to get very messy so easy to by accident add a negative, so we always monitored things which don’t do anything… So for HUGE accounts, it would be beneficial.. as long as it’s 0 impressions but I’m guessing this is just the start of an ‘idea’ from Google. @Meriem

The other huge issue is that most people don’t actually understand the auction anymore because most practitioners today only know smart bidding. @navahf

So the empathy for mechanics gets lost and folks make innocent accidents @navahf

Q2: Do you have any or many keywords in your accounts that meet this criteria? If so, do you have a lot or not that many?

Watching with much interest. @navahf

I’m not deep in paid search right now, but I can think of a TON of examples throughout my career. the “impr > 0” filter got a lot of use in the days of liberally using keyword stem combiners. @ferkungamaboobo

Here’s what we saw : @navahf

Google accounts showing low volume keywords

Not for my accounts. @robert_brady

We have a lot. @revaminkoff

I’ll shamelessly plug Optmyzr that our quick audits, regular audits, and Rule Engine help you identify low activity keywords and move them if need be (or pause them) @navahf

Also, some keywords are flagged as low search volume despite the fact that they get good activity. @revaminkoff

That’s why low activity isn’t the same as low volume. @navahf

I do not have 200 low volume keywords. I inherit those constantly however. @runnerkik

I like that it’s 200 consistently. @navahf

So, Google’s threshold for this is 13 months. I’d be very curious to see data as to whether the terms in this category now ever had traffic. That would be super interesting to me…@NeptuneMoon

I think for me anyway, this is really helping me understand how to built campaigns in a googly manner, which I genuinely believe is the right way to create campaigns. @ferkungamaboobo

It’s like people treat adding doomed keywords as busy work (not you @revaminkoff ) @navahf

@navahf similar to 4x ROAS @runnerkik

So our study applied the criteria that Google says it’s applying: the keyword cannot have had any impressions in the last 13 months. @navahf

Google: we are going to do a thing Me: Optmyzr already takes care of that for me but good luck anyways.@beyondcontent

I guess my question is how much does it really matter? I don’t feel like it does, although streamlined accounts are always easier to manage. But from a performance perspective, seems like it won’t affect much. @revaminkoff

Right! that’s why we wanted to dig in. @navahf

So I was super surprised at how many accounts would be impacted, but comforted that most should be fine @navahf

@navahf Can’t wait for part 2 of this study. @NeptuneMoon

I think a valid question would be – what’s the impact on the accounts 13, 26, 39 months back if those keywords were paused originally? what’s the desired past state that Google is trying to create now? @ferkungamaboobo

There was an outlier class of high-volume keywords with 50-75% low activity that outperformed every other cohort. @navahf

That’s the group I’m keeping my eye on @navahf

Because that’s the group that would have a legitimate reason to cry foul @navahf

It reminds me of when an old company changed up our incentive structure and everyone (except me) was going to make more money and be happier @navahf

But I smiled and nodded along because I knew the upside was going to be better sticking around. That’s how I feel a lot of marketers feel about Google @navahf

Or like idk, what is the expected future state of those campaigns if those terms are paused, given standard YOY forecasting? @ferkungamaboobo

It definitely raises questions about trend forecasting and seasonality. @beyondcontent

So like, that’s a question I have as I’m noodling on this – what’s the experiment design? @ferkungamaboobo

So 13 months should in theory allow for seasonality. @navahf

But I agree life happens @navahf

Maybe forecasting is the wrong word. More like, topic-coverage-when-things-start-to-happen-without-relying-on-random-broad-testing-campaign. @beyondcontent

Q3: What are your thoughts on Google automatically pausing these types of keywords?

@PPCKirk’s take in the study struck a chord with me: @NeptuneMoon

Google's changes and policies blur the line between ad partner and ad platform

Initially I was totally fine with it – but as  @PPCKirk and @revaminkoff and others have brought up, there are use cases that require more advertiser control. @navahf

Depends how paranoid we want to get. How close are we to total management by a PPC AI bot? Should we start studying creative design?  @beyondcontent

Yes, we should study design XD. @navahf

creative and video are the future @navahf

BUT….that’s not strictly tied to tin-foil hat mindset @navahf

I  agree, I was totally fine with it until @revaminkoff and @PPCKirk both of them made me think differently about intent. @runnerkik

That’s just a market shift. @navahf

It feels like Google has been softening us up for the inevitable “make some entries of critical data and leave the rest to us” for the past few years. Moves like this further that aim, but in a way that feels less threatening because it seems harmless on its face. @NeptuneMoon

I also don’t know if Google is organized enough for strategic decision to hurt advertisers. @runnerkik

I…. don’t read this as an antitrust thing tbh. @ferkungamaboobo

I know that’s nitpicky on a larger pull quote @ferkungamaboobo

I know that’s nitpicky on a larger pull quote @ferkungamaboobo

The antitrust is that G owns the vast majority of the display ad inventory. @ferkungamaboobo

I feel like it’s Google’s own doing though – if there were clear explanations the “reach” conclusions wouldn’t come into play. @navahf

I know my accounts I manage very well… it’s my job to. But I know when asked to audit accounts how little some accounts get touched… Will some even notice this taking place… I mean, how many accounts are opted into recommendations still…@Meriem

Sentiment is at an all-time low for Google and even well-meaning and objectively helpful tools come across as shady to those who have been burned in the past or are so jaded they only can see the “shake the cushions” Google. @navahf

For all of its vast, vast resources, Google sucks at comms and messaging. I actually find it quite funny. Because it’s absurd. But then again, PPC professionals are not an audience they care about communicating effectively with… so their comms often are incomplete and/or tone deaf to our ears. @NeptuneMoon

But the reason why I focus on that is because I think it’s indicative of the weird vibes in a lot of pushback here. we’re supposed to be data-based and yet we’re talking about tangents of tangents to argue that they shouldn’t automatically pause a targeting method that’s been ineffective for over a year. @ferkungamaboobo

I’m just scratching my head out loud, not calling anyone in particular out @ferkungamaboobo

Totally fair! @navahf

@ferkungamaboobo But why not just leave it in recommendations? Make it bold. Put a star on it, etc. I think I get twitchy when Google is going to automatically do anything “for me.” I opt out of all auto applied anything. And that is what this is at its essence. @NeptuneMoon

I would love to know how many accounts are Opted out of auto recommendations. @Meriem

As when you think about the amount of Google Ads accounts out there I feel like there are a LOT of accounts which are fully opted in. @Meriem

I look at it the same way as them cancelling ad accounts that don’t run after a few weeks. Cleaning up server space. @navahf

For me, I don’t see an issue with 0 impression KWs for 13 months… Yes it could be an opt-in feature for Recommendations @NeptuneMoon It would have be the one of the 3 things I actually opt into LOL @Meriem

I think there’s a difference too between “doing something that can increase spend” and “doing something that can only decrease spend” @ferkungamaboobo

e.g. I actually love the redundant keywords auto-apply — it helped me understand keyword matching at a level I didn’t before and is – when used in an ad-group-level bid structure like I typically use – a good call. However, it’s one that you might turn off if, say, you bid lower on exact match vs. phrase because the phrase bid will apply to the exact-match search term and your bids won’t work @ferkungamaboobo

But I would think that’s a correct structure in well less than 0.01% of accounts @ferkungamaboobo

I’m also of the opinion that if Google wrote a book that literally says “this is how this works and the right way to set up the accounts, notarized by our attorneys to literally be the truth” most PPCers would still not follow it @ferkungamaboobo

(with knowledge that there would be plenty of cases in the periphery that would not fit inside the normal distribution that advice falls into) @ferkungamaboobo

It’s already happening re: certification exams XD. @navahf

I’m trying to get caught up on everything. @ferkungamaboobo I agree it’s a little bit of a stretch, but the reason I referenced the anti-trust lawsuit is that part of the DOJs case against Google (and other monopolies) is the ability for consumers to easily move to a competitor if they are unhappy with (or even being harmed by) services provided. My point was that in a broader way, Google leaving its role as a Platform and making decisions for an advertiser isn’t normally behaviour an advertiser would accept… but with Google, there is no other recourse but to now let them mass-pause everything in our accounts. It’s pointing to the fact that Google’s ability to just wholesale make changes in the account without the advertiser really having other recourse (we couldn’t just quit because we hate this behaviour and go to Bing without all of us losing like 80-90% of our PPC revenue overnight) is the monopoly. Does that help make sense why I referenced the antitrust thing? @PPCKirk

Q4. Do you plan to re-enable paused keywords? Why or why not?

It depends. If I think they have a true chance, I’ll create them in a new campaign. Not quite re-enabling, but not quite leaving them paused either. @navahf

But for the most part, if they get paused, I’m going to leave them paused unless I miss out on queries I was getting before that I valued @navahf

For me, it would involve looking back beyond the 13-month window to see if they ever had traffic and going from there. @NeptuneMoon

If Google is pausing my keywords, I would review why has Google spotted something I haven’t. What is the reason behind that, and what are the learnings here. Also, how do I do my job better LOL @Meriem

@NeptuneMoon if they have traffic they wouldn’t get paused. @navahf

An impression (never mind a click) restarts the 13-month clock @navahf

I meant if they had traffic before the 13-month window for no traffic. @NeptuneMoon

Oh! that’s fair! @navahf

And if any terms used to get traffic, but then did not for 13 months, I’d want to try to figure out why. @navahf do you want to talk about how bidding can impact this from the study?

If something in the market shifted and new terms became a norm. @runnerkik

@NeptuneMoon – But don’t you feel that if something like an ad group dropped off for 13 months we should have noticed it by now right?! @Meriem

@Meriem I would hope that would have been noticed, yes! @NeptuneMoon

Sure! So one big pitfall for keywords is if you set a budget that can’t fit enough clicks in your day (so Google holds your keywords back), or that you set too aggressive of a smart bidding goal (CPA or ROAS – especially with bid caps without floors). These can lead to you inadvertently underbidding and taking yourself out of the auction. @navahf

There’s also the conversion factor and whether you trust your conversions at all @navahf

This is why bid floors are just as important as bid ceilings @navahf

PPCChat Participants

Related Links

Stop wasted ad spend with Karooya

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.

Leave a Reply

Your email address will not be published.