Posted by & filed under PPCChat.

In this week’s PPCChat session, experts shed light on the differences they are noticing in the Google Ads keyword matching, What are experts doing to mitigate the impact of the looser matching, What kind of audiences they are currently using in Facebook and more. This week’s session was hosted by Julie F Bacchini.

Q1: Have you noticed that keyword matching has gotten worse (matching to more irrelevant or questionable things) in Google Ads lately? If so, what are you seeing?

Yes! mostly it’s brand terms matching with their generic product. which is good, question mark, but messes up brand/generic budget setups. @JuliaVyse

Oh boy. Is it time to break out my waste rant again? The match types are getting worse, and they’re getting worse because Google wants to reduce ROAS on the platform so they can make more money. They’re doing that by stuffing waste into campaigns. @CJSlattery

Google is running with scissors essentially. Match types have almost become irrelevant and keeping up with negatives has become much more of my process than it used to be. @lchasse

Very much so. Noted earlier that I’m seeing matches with omitted words: “Blue widget” matching +blue +widget +system. It’s a massive change in intent. @armondhammer

Yes. Many accounts started getting noticeably worse in March. Conversion Rates and Cost-per-Conversions getting worse along the way. Case in point is this extremely reliable and stable client. Until recently… @Realicity

It’s not been good for a while. Every relaxation around what exact, phrase and broad really means has made keyword matching harder and harder. Intent or implied intent is far harder now than it has ever been. @SarahPixelHappy

I am seeing a very broad interpretation of what is a “close variant”. I have one client that sells chemicals and they way that different chemicals are matching to other chemicals is nuts. @NeptuneMoon

Every advertiser getting an 8x ROAS is something Google views as a failure. Because some of that profitability is money they want in their pocket. There are two was they can lower ROAS for advertisers. 1. Monetize waste and force it on advertisers. 2. Raise CPCs. @CJSlattery

Close variants are being a bit more annoying that usual. Making budget split & planning more difficult. Working on a classic @Brainlabs script to help with this! @mindswanppc

Yep, and in some accounts is overwhelming deciding where to begin. And those darn Recommendations are really beyond annoying (but informative on what Google AI “thinks” is relevant) @ynotweb

I’m beginning to wonder if they’re getting ready to drop support for BMM. It was never an official feature. @armondhammer

Google is doing both of these things. Google’s long term goal is to drive down ROAS to the lowest level possible while preventing a mass exodus from the platform. Because the lower the ROAS the more profit margin for Google. @CJSlattery

It feels like it’s gotten worse. The cynic in me wonders if this is google’s way of creating more competition for ad inventory – to make up for lower demand/revenue during lockdowns. But, surely not… doing that would be evil… @stevegibsonppc

However all targeting in Google is problematic – geotargeting, placement targeting. I have both set up in a number of accounts and I’m seeing adverts serving where they shouldn’t be. It’s not just keyword targeting making life difficult. @SarahPixelHappy

So this article that @gregfinn wrote is actually from one of my clients as I have seen in the wild cypressnorth.com/paid-search-ma…. I have been seeing Google drop important modifiers that separate BoFu from ToFu KWs like system, software, tools etc across clients. @Mark_from_MKTG

Q2: Does match type seem to have any bearing on how queries are being matched on Google Ads?

It’s getting harder to tell the difference, particularly factoring voice search. I tell Google to ‘call Alex’ and… @JuliaVyse

All match types are matching to horrible queries, at least for me. I still get twitchy whenever I see the Exact Match (close variant) in query tables..@NeptuneMoon

I agree with @JuliaVyse in that it is getting more difficult to tell the difference. I am also see just horrible match types on some brands. It really depends on the brand and products. @lchasse

And also, we had a hard enough time already! competitors trademarking generic terms, weird no-vowel brand names, make this easier google! @JuliaVyse

Match type is a farce nowadays. Aside from close variant matching being problematic, I’ll have a query that should map to a mod broad KW map to the exact version of the KW as a close variant which really messes up reporting. Mapping just goes to the highest bid. @Mark_from_MKTG

I feel like Google WANTS to make it harder on agencies and people who really fine tune. Like, what does it even matter! @ynotweb

The “Search Term” to Keyword Logic has changed so much recently. Simplest explanation: Exact = What BMM used to be without adding Extra Words. BMM = Exact but adding Extra words to your term. Phrase = BMM with “some” word order logic. Broad = WTF @Realicity

The EM close variants are SO bad for most of our clients. We got EM that are running at like 4x ROAS and then the CVs are sub 1. And using almost the same amount of budget. It’s terrible. @CJSlattery

We still embedded negative keywords to try and force all exact/phrase match traffic to exact kw. It’s not perfect and sometime broad slips in but nothing out of the ordinary. Then let BMM act as a sweeper for everything else. A bit more work to build but still works. @duanebrown

Q3: What are you doing to try to mitigate the impact of the looser matching that seems to be happening in Google Ads?

I use negative exact matching against phrase and broad keyword ad groups, but this can be a pain to set up and also the Brainlabs script but that’s a lot of steps to ‘fix’ something that used to be easy. @SarahPixelHappy

Personally – I’m moving more accounts up the matching ladder. What was BMM, I’m moving to phrase, etc. I’m a huge user of BMM, so it’s big. And a big PITA. @armondhammer

My new go to course of action for this close variant wasteland is: A. Create a master negative list of exact match negatives of my keywords without important modifiers like “software” B. Pull close variant search term reports separately C. Watch PHM KWs like a hawk. @Mark_from_MKTG

Yelling at my screen in case it helps! So much negative work though now vs. historically though. If you ever get a client with the work American in the name, just know you will be extremely busy. @lchasse

More frequent monitoring of queries and addition of negative keywords. I am not bothering with multiple match types nearly as much, because, why bother? @NeptuneMoon

I’m just doing sqrs way more now, and adding negatives along match types in my batch keyword lists. hyper vigilance. @JuliaVyse

Getting more aggressive on negative keywords and more frequent about it. Also, screaming into the void, crying in my coffee, and debating a new profession. @CJSlattery

Scripts to automatically add negatives and make sure regular SQRs don’t get missed. @mindswanppc

Much more frequent monitoring of accounts and getting pretty much surgical with the amount of negs and exclusions – some things I find scripts won’t pick up that need human intervention. @AzeemDigital

Time to go old school…. 1. Embedded negative keywords at ad group/campaign 2. Negative Keyword lists… majority of accounts we audit and win don’t use this or stopped using it. THIS IS HOW YOU WIN 3. Being ruthless on synonyms and other similar matched SQR @Realicity

I’m also making sure that my analytics are air tight. I can automate some things if I see conversions well. But if there’s low quality conversions (bad form fills for example) I’m hurt badly. @armondhammer

Definitely going more aggressive on negatives; more aggressive use of audiences as a primary control lever. More use of scripts to parse through STRs. Google is actively making it more difficult to eliminate crap from campaigns, so gotta take what you can get. @DigitalSamIAm

Q4: Is the continued fuzzying up of keyword matching in search changing your strategies for search? If so, in what ways? If not, why not?

Usually within a couple of weeks you can clean up the offensive mismatches in the form of negative exact keywords. It’s manageable IMO. @scottclark

Not fundamentally. As a long-time fan of broad match, I accept there’s a cost of doing business. @stevegibsonppc

Rethinking SEO vs PPC campaign recommendations. PPC used to be about Controlling your Reach and quicker ramp up to reach and audience (sales). The current Keyword Logic is affecting Conversions and wasting Ad Spend TOO MUCH. @Realicity

I’m all about simplifying the terms we directly target – I don’t miss the days of zip-code keywords and misspellings and whatnot. I really need to do a better job of looking at the “redundant keywords” recommendation to ID where Google is merging on close variants. @ferkungamaboobo

You mean other than shopping for a new line of work? (mostly kidding). I’m really stressing good ad copy for one. More audience usage. More automation if my conversions are very good. @armondhammer

Where I used to work stopped creating Ad Groups for all keyword match types – Phrase was cut so that when looking through SQR on the Ad Group there were fewer to check. @SarahPixelHappy

I’ve gotten a lot tighter with my keyword lists, we are spending significantly more time on negative keyword mining, and I’m staggering new campaign launches/tests within accounts to better control overall client performance. @Mark_from_MKTG

More of a workflow change, understanding it will take more work to create more negatives around your keywords to keep things where the brand needs them to be. @lchasse

I said it before and I meant it: Keywords are becoming a hint. So we need to market. Fundamentals. @armondhammer

I think the keyword targeting fuzziness is manageable (guess it depends on the amount of keywords) – You have SQR, Negative KW Lists and Scripts. It’s the other fuzziness that is a bigger issue to my clients – showing outside of area or alongside poor content. @SarahPixelHappy

Realizing that even something that seems like it has a fairly narrow focus will end up having a broader focus when it starts running and making clients aware of this. Expectation setting continues to be very important. @NeptuneMoon

Been more or less consistent the last few years. Build out for exact & BMM. Grab shopping data and see how you can grow the business. However, understand how people search for the product and be there on that search. Convert them when they are ready. @duanebrown

I wish @GoogleAds had the same process like with @MSFTAdvertising When we see a close variant that we don’t like, we contact our reps and they submit something to break to connection for that specific account. @360vardi

Q5: What kinds of audiences are you using currently in Facebook – broad, interest, lookalike, all of the above?

So I’m not the biggest FB expert so take this with a grain of salt, but for DTC clients I’ve been seeing lots of great results with the ATC/IC/PUR LAL audiences of late. @CJSlattery

I’m really only running web remarketing, CRM remarketing, and lookalikes. I’ve rarely had success with interest based audiences (except for layering on top of LAL) and job titles have been OK for ToFU content promo. @Mark_from_MKTG

I am using all of the above in my Facebook accounts. @NeptuneMoon

Broad interest for TOF demandgen stuff. Tested detailed interest expansion but I’ve had mixed results, even tho FB’s pushing it hard. Mixed LLAs for PUR intent (ATC, PUR, ENG), and then the main dependable RT and DPAs. Definitely a RT-only more and more lately. @timmhalloran

All of the above plus Custom. @JuliaVyse

For lookalike bases I’m trying to only use contacts that are sales qualified lead or higher for quality purposes. @Mark_from_MKTG

I do a lot of job targeted FB. Less interest for most campaigns. But I’m also b2b a lot, so it’s a different model, very narrow. @armondhammer

All of these, but custom is my absolute favourite. @SarahPixelHappy

All of the above — just depends on the client. FB is a royal PITA when it comes to certain audience types (jobs comes to mind), but actually pretty good on others. And some of them are just brutally broken right now, which isn’t (really) FB’s fault. @DigitalSamIAm

Also, I’ll give FB some credit for their “Power 5” BS that they’ve injected into the minds of young media buyers as some kind of panacea to all the ills of, well, everything. It doesn’t work, but it’s good enough that most people don’t know the difference. @DigitalSamIAm

Broad in a lot of cases. A few accounts do really well with well thought out interest targeting like our cannabis client. Look at GA demo or Audience Insights in Google Ads Audience Manager for ideas to target on FB. @duanebrown

Q6: Have you found that Facebook is “picking winners” more lately when it comes to the demographics it is serving ads to? Is it showing a lot more to women, for example, or a certain age group despite your targeting?

I feel like Facebook audiences are CERY self-reinforcing. One client had a couple of ads that were overwhelmingly appealing to women and now EVERYTHING is shown to 80%+ women, despite targeting both genders and regardless of what the ad is featuring. @NeptuneMoon

I feel like Facebook audiences are VERY self-reinforcing. One client had a couple of ads that were overwhelmingly appealing to women and now EVERYTHING is shown to 80%+ women, despite targeting both genders and regardless of what the ad is featuring. @NeptuneMoon

Not really. It just kinda picks on its own. I can do a broad audience with no demo filters and it’ll still choose the age+gender most likely to convert. I do put on those filters usually but it generally does a good job of finding that on its own. @timmhalloran

FB loves to pick “winner” sooner but usually, they are wright we find. They have 4 billion people on their platforms. @duanebrown

PPCChat Participants

Related Links

Leave a Reply

Your email address will not be published. Required fields are marked *

Website Protected by Spam Master