This week’s PPCChat session was inspired by the post from Kirk Williams where he focused on keyword match type segmentation & how we can organize our search campaign accounts. Host Julie F Bacchini sought PPCers views on the segmented match types in Google Ads & if Google’s machine learning does it better if they do not segment. Here is the screencap of the entire session.
Q1: Which keyword match types are you currently targeting in your Google Ads campaigns? Why are you using or not using the available match types (broad, phrase or exact)?
Exact (90% of KWs) and Phrase (9.99% of KW) Broad only under specific circumstances. @Realicity
Phrase and Exact. We are leaning heavily into voice search and while meanies won’t show us which device people actually use, we are seeing more and more conversation and questions in our education and public sector queries. @JuliaVyse
Mostly sticking to exact match and phrase match for control. There is a great case for using broad in Pmax, but we haven’t quite taking the step yet. @soanders
Mainly phrase & exact. Our clients are on the smaller side and can’t afford to waste huge chunks of budget on sloppy matches by Google so very little broad. @selley2134
Depends on the client and account goal. If it needs to be super strict I tend to go with exact w some phrase sprinkled in. Currently working on an enterprise level account where we are seeing stellar results also using broad & smart bidding. @BorisBeceric
I currently use phrase and exact. I am going to more exact and less phrase as I become increasingly frustrated with (a) phrase being what broad match modified was and (b) not being able to see so many damn queries. @NeptuneMoon
I start with phrase it offers the best of both words – with exact and the small radius hyper local campaigns I run we do run into “low search volume” so phrase is a happy medium. @runnerkik
Recently analyzed a Broad based keyword campaign for 2021 results where we analyzed Offline Sales. 0 Offline Sales for 2021 even though good CPL and volume. Keywords that affect the Funnel are important. @Realicity
I primarily use phrase and exact match. I will test out broad match occasionally in ad groups with well-established negatives. But I haven’t, and won’t any time soon, “upgrade my existing keywords to broad match” as Google says. @alexnicoll93
I should add that my client mix is currently all low volume (in any given month) clients. So, much of the advantages of ML just misses them… @NeptuneMoon
Broad is just that too broad but sometimes I have to use it to explore keywords or to just spend money when the money has to be spent (some have to spend regardless). @runnerkik
Only use phrase match and exact match. I will use broad match sometimes but only if I really need to – I prefer not to though because the results haven’t won me over yet. @dylanppc
I should say I use broad quite a bit in Microsoft Ads. It’s fully just a different thing for them. And there are times when broad testing in a TARGET audience framework can be helpful. @JuliaVyse
Our agency mainly uses phrase match (RIP to BMM, which used to be our go-to), as well as exact match. The main reason is to better control the traffic we are driving, as we’ve rarely seen good results from pure broad match. @adclarke10
That said, we’ve seen better performance from broad match keywords when we layer a remarketing (or other) audience into the campaign as well. @adclarke10
We primarily use Exact and Phrase, with controlled Broad and DSA usage. A lot of it depends on client need, goals, industry, how good Google is in that vertical for targeting intent (yes I find it ranges by account), etc. @PPCKirk
Have started Pausing PHRASE match keywords for tenured accounts where Exact is covered extensively and we are Limited By Budget. @Realicity
I use all match types however I’m strategic about which keyword concepts get put on each match type and use ad group negatives to direct budget. @navahf
Broad match typically gets its own ad group (sometimes campaign). Exact will be for the longer tail concepts and phrase match is for the 2-4 word keywords. @navahf
Last tweet on this one – I inherit fully broad match kw accounts all the time and I sometimes have to live with the waste because the client can’t cope with a drastic drop in conversions, even if it is explained. @runnerkik
Exact and phrase. We will use broad for query filtering strategies, especially on newer accounts where we’re still experimenting on which words work. But by and large, P and E. @timmhalloran
I stick to phrase and exact pretty much. Have some broad keywords being tested here and there. (big exception is Google Grants where it’s all broad). @robert_brady
Mainly sticking to exact and phrase match for the control but we have with broad match I’m some accounts. @adwordsgirl
Old school, always do phrase and exact. I am now forced to consider broad, and dip my toe in occasionally but I still don’t trust it. @JonKagan
We typically use Exact and Phrase Match. We use Dynamic Search and Broad Match KWs in certain cases, primarily with eCommerce clients because the system can optimize toward actual revenue – which is much more difficult for Lead-Gen clients without Offline Tracking. @MenachemAni
Q2: Are you currently running campaigns with segmented match types in Google Ads? If so, why and if not, why not?
I haven’t done this since 2016. Segmenting match types out (save for broad match as a catch all) creates too many opportunities for accidental duplicates and impression share lost to rank. Absolutely team power data ad group campaign. @navahf
Overall: no (see my article). Practically in certain instances, yes. Two examples: – limited budget accounts where we primarily use exact-brand campaigns where we want to control brand/non-brand segmentation (close variants sometimes suck at this). @PPCKirk
Yes. we have some very brand-sensitive clients. Think regional retailer competing with a national big dog. So we do separate brand from Generic, and then exact from phrase for easier management through the year. Less needed on conversational public campaigns. @JuliaVyse
I do not have my campaigns segmented by match type. I put phrase and exact in the same ad groups and separate them more by theme. If I were to run broad, I would do that in a separate campaign. @NeptuneMoon
Not really. Occasionally, we will set up a broad match campaign separate from the other keywords to test but just do it to control budget so as not to take away from phrase/exact. @selley2134
No. You might get away with it on smaller accounts, but the negatives outweigh the positives. You just create more work for yourself and you can’t win the negative keyword battle. @BorisBeceric
Yes because I built them that way but more and more I am building phrase only accounts and not even creating the exact and broad versions that I would segment into their own ad groups. @runnerkik
Yes. It Depends. One of the biggest reasons for this recently is when certain verticals map to too many competitor Brand terms for Broad and Phrase match keywords. Like Credit Unions, Chiropractors, Dentists… Anything that business names are XYZ Chiropractor. @Realicity
Generally, I’m still doing the same ol’ while being aware it works differently. I’ll make a phrase + exact ad group and sometimes a broad ad group and then hike up bid adjustments on exact and throttle broad to a level that’s tolerable in the interim. @timmhalloran
I don’t segment match types, but if I have any DSA ad groups then they’ll usually go in their own campaign. Otherwise it’s mostly phrase with a few extra exact matches in the same ad group. @dylanppc
Not segmenting match types but segmenting campaign types instead. One client will have brand, smart shopping, pmax, products. @soanders
I used to! I’ve moved away from that in the last year or so. I intermix match types (again, primarily phrase + exact) now. I think close variants and my increased “lean” into the algo, it doesn’t seem as necessary for me to segment. @alexnicoll93
We actively used/tested SKAGs a few years ago, but have mostly moved away from this approach. After updates to match types & close variants, we found that SKAGs were driving similar traffic, so it made more sense to consolidate or further segment by KW theme/intent. @adclarke10
We stopped segmenting by match types several years ago. Instead, we’ll segment by theme. @adwordsgirl
Yes, honestly I do it solely because I am old school and incredibly stubborn. I also segment by the device to make people’s life harder. @JonKagan
Never been a fan of Match Type Ad Groups. @MenachemAni
Q3: Part of the argument for not segmenting match types in today’s Google Ads accounts is that machine learning (ML) does better if you don’t segment. Has this been your experience?
Nope. @selley2134
LMFAO. No. @adwordsgirl
Mixed results. I feel like on bigger accounts it might be the case, but on some of my smaller accounts it doesn’t make a difference. Or WORSE. @BorisBeceric
On Facebook Ads, I have definitely seen better performance after consolidating over-segmented ad sets (ex: significantly lower CPMs). But I don’t think we’ve really seen the same on Google Ads. @adclarke10
To be honest, my decision to move away from segmentation was as much about ease of management as it was about the data. It’s easier to just drop well performing search terms as exact rather than put it in the exact match ad group and worry about negatives. @alexnicoll93
When I moved away from segmentation I didn’t see a decrease in performance and I can still manage my keywords effectively. So I just kind of went with it. @alexnicoll93
Sometimes a better job is how I keep my head straight – with mixed match type – it seems the broad or phrase just always serve – and I am only diluting the data with the 1 or 2 clicks from exact. @runnerkik
Yes. Also, because all match types are looser, it’s much more difficult to segment properly and the added complexity is not worth it. @MenachemAni
Q4: Have you changed the way you set expectations for newly unsegmented or existing unsegmented campaigns in Google Ads?
luckily no. Our clients are very dialed and we’ve built a lot of trust. So overall expectations are about campaign goals and budget, less about executional details. It takes a time investment to get here, but it’s so worth it. @JuliaVyse
Changing expectations for GG Ads. Over the years we have come to make less and less promises. We are moving back into the marketers real role: analyse, hypothesize, explore, test, optimize. We get paid for brain juice and operational excellence. Yay. @soanders
Literally running with different sets of KPI’s depending on the audience segment or account structure. @JonKagan
We typically handle account Re-orgs in 1 of 2 ways: – the account doesn’t perform well: we just reorganize because it’s not like it’s going to get worse lulz – the account already performs well: we transition campaigns into our ideal over time in small bits. @PPCKirk
“Mostly” Keyword Segment by Theme in Ad Groups. This hasn’t changed much except fewer AGs. New: More reliance on ML and Audience Analysis. @Realicity
Yes, I have changed the way I set expectations for G Ads. The whole concept of a learning period for search ads is NEW. Recommend higher budgets for this period cause there is just so much waste w/ bad matches. Need more patience for search, which is WEIRD. @NeptuneMoon
Not necessarily other than I create way fewer keywords than I used to with long tail becoming obsolete – in terms of segmenting when I take on accounts that aren’t segmented I treat them as opportunities to learn vs immediately changing with my assumptions. @runnerkik
Yup. Same here. Worded well so I’m just gonna steal this answer. @alexnicoll93
Our expectation-setting hasn’t changed dramatically, because I think the same principles apply to both segmented & unsegmented campaigns. Either way, Google’s algorithm has to go through a learning/data collection phase, which we communicate to the client. @adclarke10
Not really. Clients are concerned about their KPIs, not the minutia. @BorisBeceric
PPCChat Participants
- James Svoboda @Realicity
- Julia Vyse @JuliaVyse
- Anders Hjorth @soanders
- Shaun Elley @selley2134
- Boris Beceric @BorisBeceric
- Julie F Bacchini @NeptuneMoon
- Sarah Stemen @runnerkik
- Alex Nicoll @alexnicoll93
- Dylan Online @dylanppc
- Ashton Clarke @adclarke10
- Kirk Williams @PPCKirk
- Navah Hopkins @navahf
- Tim Halloran @timmhalloran
- Robert Brady @robert_brady
- Ameet Khabra @adwordsgirl
- Jon Kagan @JonKagan
- Menachem Ani @MenachemAni
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.