Episode 111: Google Broad Match Is Back

Drive and Convert (Ep. 111): Google Broad Match Is Back

In this episode, Ryan and Jon discuss Google's broad match keyword feature and the best ways that brands can utilize it.

Listen to this episode:

About This Episode:

Just like Google is changing, Ryan’s stance on broad match is changing too. In this week’s episode, Ryan gets candid about his opinion on broad match and why he now believes it to be valuable… in certain situations.

Google has used AI for its broad match keyword feature for years, but haven’t always been forthcoming about it. Now that there is transparency about how the feature works, Some interesting use cases have become clear. Google isn’t just throwing ads around, there is calculation and intent behind search results when this feature is in use.

Check out the full episode to learn:

  • How your brand should control keyword match types
  • What user data Google uses to assume intent when showing ads
  • In what ways the different match types have changed

If you have questions, ideas, or feedback to share, connect with us on LinkedIn. We’re Jon MacDonald and Ryan Garrow.

Subscribe To The Show:

Episode Transcript:

Announcer: You’re listening to Drive and Convert, a podcast about helping online brands to build a better e-commerce growth engine, with Jon MacDonald and Ryan Garrow. Jon: Hey, Ryan. So you’ve been on a broad match killing soapbox for years. I don’t have a better way to say that. Ryan: Yes. I have. Jon: But you’ve been pretty vocal about it. Although I will say recently, it seems like you’ve possibly been changing your tune on broad match a little bit, so you thought it’d be a great topic. I’m really intrigued when we change opinions, because usually, that means something juicy is happening. We have our reasons for doing it, so love to chat about that today, too. Ryan: Yeah. Broad match has been the dumpster fire of Google Ads for years, and there’s just no other way to put it. It was just … wastes disgusting amount of money, and we’ve, as a sales org, at Logical Position, have just sold against that for years. Because it’s the easy button, and Google makes it very easy to start with broad match. So I’ve had to not like it, because I’ve seen the waste that it’s done. But after spending some time down at Google a couple of weeks ago, we read at their Google Cambridge office with some clients, they actually had a big session just on broad match and what it’s doing now. Again, I’ll put an asterisk by all of this, because it’s still early on in this. But usually, when Google does a session, they’re trying to ram something down your throat, because there’s incentives around it for the team that’s pushing it at you. Jon: Okay. Ryan: In this case, they talked to us before the presentations, and our team got together and actually has been doing some tests around this, and adding it into accounts, and watching to be able to back up some of the things that Google’s saying. So there’s some very smart people at Google, and I do think they’re doing some valuable things when it comes to broad match now, not for everybody, but it’s for sure there. Jon: Okay. I know that you don’t change your mind without good reasons, so that’s good to hear. How have broad match types changed, then? Ryan: Well, the names of match types in Google haven’t changed. We still have broad, phrase, exact. There’s only three of them. We have eliminated, over the last few years, modified broad. That’s essentially phrase match now. But I guess I should probably start with the foundation, make sure everybody listening is on the same page with us. If your search is furniture store, that’s the actual search term for the exact match, furniture store. It’s been that way for years. You’re going to show for furniture store. No surprise. That’s the exact match keyword. Somebody searches furniture store. You show. But it does now have the same meanings. So that exact match keyword will also show for somebody searching home furnishing shop. Jon: Interesting. Ryan: That is a pretty big change that’s now in there for exact match. We still like exact match. It’s still functioning very good, and we’ve seen Google keep that thesaurus pretty much in check. Jon: Right. But they might want to check a dictionary, because that’s not the definition of exact. Ryan: It is not. So Google’s thesaurus on exact and what they think exact means, obviously not there. It is no longer exact exact, but it’s the closest Google’s going to give us. Again, Google makes money with clicks, and if they think somebody’s going to click an ad because of that, they have some incentive to show it. If you don’t want to show for home furnishing shop, you would need to be exact match negativing that out, but it’s there. It’s the closest we’ve got to exact. Then, the next layer up is phrase, which, in reality, is more like the old-school broad match keyword of yesteryear with some pretty loose thesaurus is on there. But if you have a phrase match keyword for furniture store, that’s going to show for searches that include the meaning of the keyword, not just … It has searches that include. So it’s not necessarily it has to be in the right order. So if you have a phrase match, furniture store, in your account, it would show for cheap furniture stores. It would show for living room furniture deals. Jon: Interesting. Ryan: It would show for, what store is the best place to find a bedroom dresser? Jon: Wow. Okay. Ryan: That’s a lot of different searches rolling up into a phrase match that used to be, by the way, you had to have furniture store in the search term together. Nothing could come in between. Things could go on the end, but you had to have furniture store in the phrase of the search query. That’s the name phrase match. Now, it’s like it has the same meaning, we think, or the same intent, but it’s going to be all over the place still. So that’s still a very broad version of a term. Jon: Right, but there’s a third one called broad. Ryan: There is. Jon: So how do we get more broad than that? Ryan: Well, now it’s just instead of having the meaning of the terms in there, it just relates. And so the term furniture store would show for home decor or cream-colored leather sectional. It relates, but it has not even nearly the same intent in my opinion. If I have a physical storefront selling furniture, and somebody is looking for a cream-colored leather sectional, I may not even have that in inventory. I may have wood furniture, and I put a keyword in there for furniture store, because I sell furniture. I can now show ads for cream-colored leather sectional, which is not ideal, because that’s also maybe an e-comm intent. Somebody wants to buy that online. So I can have shopping ads up there, as well, and I’m, with this search, probably showing text ads for that. Jon: Interesting. Okay. So besides Google not knowing how to exactly define broad phrase and exact according to what a dictionary would, it sounds interesting. You just have to be able to speak their language, and as we’ve said a hundred times on this show in 111 episodes, probably more than a hundred, it’s Google’s world, and we’re just living in it. Ryan: Exactly. Jon: So fair enough. Okay. We’ll speak their language. So Google still goes way beyond a keyword with that broad match settings. Right? So what’s changed to make you think positively about broad match now over what it was before? Ryan: Well, the simple answer is just AI, and that seems to be everybody’s simple answer right now. All you have to do to raise your stock price is- Jon: Right. So that everyone’s favorite answer these days, if nothing else. Ryan: AI, my stock price goes up 40%. This is great. Jon: There you go. Boom. Done. Ryan: Yeah, and all the holdings I have in my stock portfolio, if they would just mention AI on their next earning call, I would be a very happy person. Jon: Cash out and retire. Ryan: The larger answer is Google’s had AI behind the scenes for years, but they’ve been very late to the game in talking about what AI is doing and where and how it’s working. That’s kind of been on Google’s end that they’ve missed that boat in letting us know. But the recent transparency around it, about how broad match keywords are leveraging AI, is actually pretty cool. It’s not been something Google’s spent a lot of time and energy of actually talking about. In fact, I think this was my only Google presentation in 15 years that actually dove into what is this keyword doing and then how to understand what it can show for. Before, it was like, “Broad match keywords.” I’m like, “Okay. Well, what does that mean?” “It means broad match.” I was like, “Great. Pretty excited about that definition, Google. I’m not going to use it.” So we didn’t. Some key things now, based on our recent understanding of what Google’s doing in AI, about what broad match keywords are. And so not only is Google going to use that keyword, but they’re going to look at other keywords in the ad group to decide the context of that keyword. So, for example, if you have a keyword, a broad match keyword in your account for pink socks, and then the search term is salmon socks, Google’s going to look at the other keywords in your account and in that ad group, like red socks, blue socks, and purple socks, and they’re all broad match. Google says, “Oh. You’re showing colored socks,” and Google knows that when somebody searches salmon socks, they’re generally understanding that that’s referring to the color, not pictures of salmon on socks, the fish. Jon: Got it. Okay. Ryan: And so Google has data around this, and they’ve watched clicks and say, “Most of the time, salmon is clicking on socks for color, not fish.” They’re also going to look now at previous searches, and this is the fascinating one to me, probably most fascinating, is that they’re going to say, “All right. This user has done these searches in the past. Therefore, we know what they intend on this search,” or they think they do. And so if you’ve got a keyword in your account … The example they used at this event was Chicago versus New York baseball. So if that was a term you were on, and that was a term you were trying to show ads for, they would say, “Okay. Somebody searches Chicago versus New York.” So they forgot the baseball piece, which was pretty important in your keyword. You’re trying to show baseball-related things, but this user had previously searched baseball scores. And so Google says, “All right. Based on their previous searches, this user is interested in baseball scores, and then we know that when they search Google versus New York, they’re looking for the baseball score. They’re not trying to figure out the food scene between Chicago and New York. They’re looking for baseball.” So they would show an ad for that. Jon: Okay. No. That’s really helpful, because in the past, you probably would even get Chicago versus New York pizza coming up. Right? Ryan: No. You just, you wouldn’t get that search. And so I think that is a really good one. Now, there’s a lot of thinking that has to go into all this by the way, and a lot of asterisks, because you’re not actually getting search queries for most of these terms now. Google is keeping a lot of that held tightly, and you don’t get to see it. Some of that, again, it’s probably on purpose so that you can’t sculpt the traffic as much as you want and eliminate searches. Because Google says they’re good, you will believe them, and trust them, and it is great. IRS doing your taxes, you’re going to be paying more generally, but there is value here. So another one is going to be user location. This is pretty easy. If you’re trying to show ads for restaurants in New York, and somebody looks for restaurants near me, and Google knows they’re in New York, they’re going to show an ad. Pretty basic. They’ve probably been doing that for a while. But again, you’re going to start showing based on user location if that seems to be part of the intent of your ad. Then, the other real, I think, big piece of AI in this is Google going to the landing page and trying to understand the context of your business. So this becomes more important as you’re designing ad groups now, to make sure they’re sending it to landing pages specific to the keywords in your ad group, not just sending everybody to the homepage. Because if you don’t have a bunch of content on your home page related to these keywords, it’s going to be difficult for this piece of AI to really do what it’s supposed to do. Jon: Yeah. Definitely not going to convert. Ryan: Yeah. For sure won’t convert. If you’re sending all your traffic to your home page, call Jon and I immediately. Pause the podcast, and we’ll help you. But again, this, I think, is a dumb example, because if you’re putting this in your ad, you’re way too broad. But if you had the keyword shoes, is the one they brought up, you put that broad match keyword, shoes. I’m like, I don’t know why that makes sense for anybody unless you’re spending gobs of money. But the search term was Run Far Footwear 1000, which is a brand of shoes and their specific model. Your landing page that you had sent the shoe ad group to was for ultra marathon shoes, which are a different subset of running shoes, that included the Run Far Footwear 1000. So it actually included the search query, but you just decided to make an ad and a keyword shoes, not even running shoes. You would show an ad for that, which would be appropriate. But again, if you are selling shoes of a specific type, and that keyword is in your account, again, call me. Because that’s terrible. But the idea is that Google has that ability to look beyond the ad group and say, “Oh. The intent is to sell these running shoes. Let’s show ads for that.” Jon: I think that’s going to be great for folks who are just starting out and don’t know much about Google. So I feel like that’s a win for beginners, probably a loss for folks who have an agency running their account or are a little more advanced. Ryan: Well, it can be, but also, it’s not that Google is always going to do that and do that difficult work for you. So Google is making it very easy to get an ad set up and run, which is part of their role of saying, “Hey. Every business should be able to get an ad up without an agency to start.” I mean, if you’re only going to spend 100 or 200 bucks a month, don’t pay an agency to do that. You have to get going on your own. But if you’re selling running shoes for men, women, and trail running, you’ve got all these segments on there, you’ve done a good job merchandising, if you put the keyword shoes in your account, you’re going to show for a lot of crap. And so you don’t necessarily want to put that keyword in there when you’re only spending one or 200 bucks a month. Jon: Not to mention that’s going to be more expensive per click to do such a broad term. Ryan: Yeah. Well, you might get some cheap stuff, but it’s going to be cheap stuff that nobody else wants, that nobody else that can spend good and have experts working on their behalf is going to compete on. So there’s a lot of moving pieces here still. I think that it’s still fairly early on for most companies moving into broad match at scale. And so I withhold my opinion on the value for small advertisers still, because I haven’t seen it work yet. Announcer: You’re listening to Drive and Convert, a podcast focused on e-commerce growth. Your hosts are Jon MacDonald, founder of The Good, a conversion rate optimization agency that works with e-commerce brands to help convert more of their visitors into buyers, and Ryan Garrow of Logical Position, a digital marketing agency offering pay-per-click management, search engine optimization, and website design services to brands of all sizes. If you find this podcast helpful, please help us out by leaving a review on Apple Podcasts and sharing it with a friend or colleague. Thank you. Jon: Oh. Okay. So these other keywords, the previous searches, using user’s location, and having context of that landing page you’re sending folks to. Should everyone just be using broad match now? I mean, it seems like it’s better. Ryan: It’s better. I’ll say that. It’s for sure better. So I don’t want to kill broad match. Yeah. I’m off that kill broad match soapbox. Jon: Okay. Ryan: That’s not my goal anymore. Jon: Keep it alive, but maybe at an arm’s length. Ryan: Yeah. It’s, for sure, not for everybody. I would say a brand still has to be very careful and controlled how and when they deploy broad match keywords. And so I would probably boil it down to building from the bottom of the funnel up still. So if you haven’t maxed out exact match keywords for your brand … We go back to the running shoe one. Run Far Forward 1000, if there’s still a bunch of those searches left, and you sell those, and you want to sell more, capture those first before you start moving into a broad match of shoe and try to sell some Brooks, or Saucony, or ASICS, and all the other shoes you have. Get very specific where you have the margin to play and you know the lifetime value of those customers. If you know that people that buy Run Far Forward 1000 shoes from you are more likely to buy another pair of shoes in three to six months, sell more of those. Adjust your ROAS goals before you start moving into some broad match keywords. If I had to … You’ve got a gun to my head and tell me, “You have to come up with an idea of when broad match makes sense,” I’ll tell you if you’re spending less than 10 grand a month on Google, broad match probably doesn’t make sense yet. There’ll be exceptions across the board, I’m sure, but that’s a general rule of thumb. It’s probably even higher than that, but for sure, if you’re spending less than 10 grand a month, and you don’t have an agency working for you, it would be difficult for me to advocate for broad match keywords. I’ve had an analogy for a while that I’ve used to kind of explain Google Ads management, and I think it applies in this case. AI is powerful, and there’s a lot of great things you can do with it. But it’s kind of like a Formula 1 car. It’s very complex. It can be fragile, but man, if it is working for you, you are going extremely fast and doing some really cool stuff. But just because you have a driver’s license, doesn’t … You conceptually know how a car works, accelerator, brake, steering wheel, road, gears. Great. It goes forward. Maybe you can change the oil. If you’re my wife, you can’t drive manual, so you have to stay on automatic or a one-pedal car. But knowing how to drive a regular car does not mean you could get behind the wheel of a Formula 1 car and drive, let alone compete in a race. I mean there are 20 humans on the planet that are qualified to drive that type of car in a race. And so I don’t think controlling broad match keywords is that extreme. There’s definitely more than 20 people on the planet that can do it for you, but there’s so many moving pieces and so much analysis that has to be done to protect your brand from that. I cannot, in good conscience, advocate for somebody starting up on their own deciding that broad match is a great one. Google defaults to broad match. You have to actually go into Google Ads, find the settings to turn off keywords to say, “I don’t want them all to be broad match in my account. I want you to use these settings that I set.” And so if you’re going into a Google Ads account, you’re running it, go into settings. It’s under additional settings, even below that on your campaign, that you have to go unclick broad match and make sure that, “If I put phrase match and exact match in, I want you to use those, Google.” It’s tough. It’s not an easy button for people for sure. Jon: So how should brands control this key match, keyword match types, then? Ryan: The best is going to be controlling your bids and making sure that you’ve got them in the right order. So I still advocate for tightly-controlled ad groups, and so most brands should have exact and phrase in their ad group and know exactly what people are searching for and where they’re going on the page. So staying on that Run Far 1000, you would have an ad group with phrase and exact on that, landing on the product page for that Run Far 1000. If you had Run Far Forward as a brand, you would land on a category page for all the Run Far Forward shoes you have on there, phrase and exact. You would bid the exact match higher than phrase. So you would know maybe it’s $2 for exact match as your bid in the same ad group, and then your phrase match would be bid at, say, $1.50. That would ensure that the exact match gets all the exact queries there, and then the phrase match would pick up the things that are extra that have the same meaning. Maybe somebody may mistype and say, Run Far Forward 2000, and maybe there isn’t a 2000 shoe. I don’t know. I don’t wear those shoes, but it would be the same meaning, or ultra marathon run far forward. Those types of things would show for the phrase match, go to the right landing page because of that. Most companies don’t put broad match in there. But if you’ve got enough spend, and you’re hitting thresholds that say, “Hey. This ad group is not spending anymore. I can see impression share. I’m at the top most of the time. I’m getting above 80% impression share on this exact match keyword,” great. Let’s start putting some broad match campaigns together, possibly, or broad match keywords where you keep the bids down lower. And so I would say if I was going to go broad match in that account, I would say broad match would have a bit of a dollar. Then, I would say I’m going to use negative keywords to sculpt out those exact match keywords that I know should be hitting the exact match ad group. So there’s a lot of work that goes in to say, “Google, I want you to start using this a little bit and to try to take advantage of some of your AI there based on past searches you’ve done and all this. But if they search exactly what I want to be showing for, make sure that goes to the ad group, and I can see the search, the impression, and I can see the conversion so I don’t have messy data.” So what I’ll see a lot of times is you’ll have mismatched bids all over the place, because somebody sees a conversion go to a phrase match. I’m like, “Oh. The phrase match is working really well. This is great.” They’re going to bid phrase match up. It gets up to $3, $3.50, and then you’ve got exact and phrase competing against each other. Because the phrase can take the exact, and all of a sudden, exact match search queries go down to zero. It’s all going to phrase. Well, you’ve just given Google permission to take more money from you. And so that phrase, they’re going to do it, because you gave them permission. Then, if you have broad in there, it can suck broad match out, and then you’ve got messy conversion data. It becomes very difficult for a company like you, Jon, to understand, how is this keyword converting? This traffic is going to the Run Far Forward page and not converting. Well, you’ve got a bunch of messy data. They’re actually searching for ASICS. They’re not converting, because they’re not on the page. Jon: There’s a lot to think about here. Okay. So I’m with you in understanding why you’re no longer on the soapbox of killing off broad match. I’m still confused why Google refuses to read the dictionary, but I’ll speak their language. It does seem like it could be helpful for folks who are starting out in Google to get more success, to kind of help accelerate things. I can see why Google would do that, and it does seem to give you more flexibility overall. So that’s positive, but all in all, okay. I hear you. I think that as a user, this could be helpful for me, so I do appreciate that. Ryan: Yeah. The past history stuff is good, and the problem I would have with a new user starting off on this would be you don’t get the search queries. And so when you don’t know what they’re actually searching for that’s converting, it becomes difficult to say, “Oh. I know that we’re competing really well on purple running shoes, because I have some.” They’re actually searching red, and getting to your site, and not converting. So you just don’t get the transparency that I love and have grown to love on Google. Jon: That makes sense. Ryan: So you have to accept that, and it is what it is. Also, when you mix all this in with Performance Max, it even gets more complicated, which is a whole nother conversation. But when you have broad match in your account, Performance Max can take precedent, and take those queries out, and put them in broad match. So if they’re converting really well, broad match can suck those in and create text ads for you. Because it’s converting well, because it’s helping Performance Max accomplish its goals. Jon: Right. Right. So last question on my side. When does this roll out for everybody? Is this something … Ryan: It’s already out. Jon: Because I know you just heard about this. It’s already there? Okay. Ryan: It’s been out for a while, and Google’s just now being a little more transparent, I guess, and at least telling some of their larger agencies like us what it’s doing and giving us access and some transparency into that, which I love. Because it helps us understand what to expect from a broad match. Whereas before, it was across the board, pause broad match, because the thesaurus is so loose, almost guaranteed it’s wasting money. Now, it’s like, “Okay. We’ve maxed out at the bottom of the funnel. We’re moving up into that intent, creating demand.” I think it’s going to be some of the areas that are trying to push into social media, creating demand when you can’t, when you’ve maxed everything out at the bottom, like, “Hey. Let’s go run some broad match,” because it’s going to be better, generally speaking, than running a display ad. So it’s going to be the layer before that. So that’s why you’ve got to have enough spend to accept the fact that it’ll be some lower return, and it fits into most of the larger brands. When we’ve ran broad match in larger brands … So a quick example. We were able to spend only 10% more by adding broad match in year over year. This is ’24 versus ’23, but conversions went up 25%. We actually dropped the CPA, the cost of acquisition, I think 15%. It was somewhere around those numbers. Jon: Wow. Ryan: But it was pretty powerful in being able to add that match type in when the spend was high enough, and they could actually leverage all of our analysis and leverage we added to control it. Jon: That’s great. Well, a lot of interesting stuff here, Ryan. I appreciate you sharing that with me today, and as always, I’ve walked away learning something new and poking fun at Google a few times. So all in all … Ryan: It is Google’s world, and we do like it. Because it grows businesses. Jon: Yeah. There you go. Ryan: It’s just, it’s challenging. Jon: Yes. It can be. That’s why we have you here. Right? Can tell us some of this. Ryan: Hopefully, I stick here, stick around and keep helping. Thanks, Jon. Appreciate it. Jon: Thank you, Ryan. Announcer: Thanks for listening to Drive and Convert with Jon MacDonald and Ryan Garrow. To keep up to date with new episodes, you can subscribe at driveandconvert.com.

Jon MacDonald smiling at the camera for The Good

About the Author

Jon MacDonald

Jon MacDonald is founder and President of The Good, a digital experience optimization firm that has achieved results for some of the largest companies including Adobe, Nike, Xerox, Verizon, Intel and more. Jon regularly contributes to publications like Entrepreneur and Inc.