Podcast #64

Adapting to Google’s Push Towards Automation: The Importance of Broad Match

Google is actively promoting the use of automation and AI in its advertising solutions. This shift is causing traditional "exact match" to become obsolete and the limitations of other keyword matching options are becoming increasingly apparent. To keep up with this change, advertisers are being pushed to use "broad match" type in their campaigns.

Episode Highlights:

Evan Ilgenfritz: “Essentially what’s happening is broad match now has increased functionality that the other match types do not. This is a big one because right now broad match can leverage are a variety of things such as previous user search history… Ultimately what this means is we’re evolving, our own understanding of the importance of broad match and Cardinal, in particular, is now testing and kind of looking forward to the future here for what this improved broad match could offer.

Related Resources

Announcer: Welcome to The Ignite Podcast, the only healthcare marketing podcast that digs into the digital strategies and tactics that help you accelerate growth. Each week, Cardinal’s experts explore innovative ways to build your digital presence and attract more patients. Buckle up for another episode of Ignite.

Lauren Leone: Hey everyone. Welcome back to Ignite Digital Marketing Podcast. My name’s Lauren Leone. I’m the SVP of healthcare marketing. Today I’m joined by our director of paid media, Evan Ilgenfritz and our topic of the day is a discussion about Google’s move towards what I’m going to loosely call the keyword list future. What we want to talk about today is how Cardinal is approaching testing within Google’s platforms to adapt, to learn now before Google forces our hands and to kind of lean into machine learning and AI. What we’re noticing in platform is that the future is going to look like gone are the days of traditional exact match and phrase match types. Ultimately, Google is going to say, tell me the core keywords that describe what you do. Give me a link to your domain and I will read that and understand what it is that you do and what keywords to bid on. What we want to talk about today is maybe how some of our testing of that is going and how we see Google using other signals besides just keywords to ultimately find us the conversions that we’re looking for. Evan, I guess fundamentally expand on that really broad description that I gave and just tell our listeners today what it is that we’re seeing in the change or what we expect to come.

Evan Ilgenfritz: Sure. Google has been making shifts over many years towards more smart automated systems. Historically, we had four major match types, broad match, broad match modified, phrase match, and exact match. Of course, exact being the most specific. Phrase, making sure the string of words go to the right in the right order, and broad match, and broad match modified, were variations of each other doing more broader reaching searches. I’ll be the first to admit that historically I have not been a huge fan of broad match because it has been a bit unruly. You can put a keyword that you think is relevant and it will capture a search out for stuff that does not appear even remotely relevant.

This relationship is changing, Over the years, I believe last year they eliminated broad match modified. Broad match modified is a bit more nuanced control, broad match requiring certain words appear in a search query but they eliminated that in favor of moving towards adding that functionality to phrase and exact. What we historically knew about exact is that it was literally what keyword you put in it had to match that exactly. That’s been evolving. We’ve added plurals and close related terms to that list. Exact match is not quite as exact as it used to be. Phrase match has taken on more of broad match characteristics. The order of words can still be important but the system will reach out and try to find things that it seems are very relevant. That puts broad match in an interesting realm. As I said, historically we’ve shied away from it because it was very hard to control.

Now as Google is moving towards this kind of automated machine learning future that we’re all aware of pushing more recommending more accounts on the smart bidding and tactics like that broad match is evolving as well. Essentially what’s happening is broad match now has increased functionality that the other match types do not. This is a big one because right now broad match what it can leverage are a variety of things such as previous user search history. Have they searched things relevant to what you’re offering? Proximity and other characteristics related to this person, the real of audiences, the quality of audiences. Phrase match do not have this at all. You can’t really in effect and overlay audiences the same as you can on search and display but we’re seeing this now on broad match. Ultimately what this means is we’re evolving. Our own understanding of the importance of broad match and cardinal, in particular, is now testing and kind of looking forward to the future here for what this improved broad match could offer.

Lauren: So you hear broad match like you said, we’ve not historically been a fan. All the audits I have done on accounts in the past five years I see broad match as a major red flag but that has totally shifted. The fears of broad match of it being unruly, there are ways that now given this kind of shift that those can be controlled and I think the biggest callouts for the group are going to be how to think about negative keywords and what is the right time to deploy broad match when in kind of the smart bidding cycle or does it make sense and what controls have to be in place before you just go out and do this.

Evan: Sure. One last thing I’ll say about there’s a couple of big benefits in broad match. For one, a lot of accounts and clients will see that they’ve reached the ceiling on relevant keywords and they’re having trouble spending the prescribed budget. We’ve got a very nuanced specific list of keywords and that list is hitting its ceiling of what it can offer, what we can capture. Broad match the benefit already there is that broad match offers a wider scope to capture and it opens up your available impression share.

Lauren: It essentially loosens the perimeter on semantics and semantics start changing and it allows us to adapt to those changes in the way people are typing their queries in rather than assuming we just happen to know all the different places.

Evan: That’s exactly right. It’s an ongoing challenge to know exactly. You’ll always be surprised and paid search about what people search and how relevant they are. You might think you know the exact phrases people would say but what broad match illuminates is that we’re trying to find people closer to the conversion point of a journey even if they don’t say the exact right phrases. People that could be looking for information in their searches, let’s say using a dental client as an example. We would always assume that exact match keyword like dentist near me is going to be much more qualified than a search like do I need a dentist. Now what broad match is doing though if you use signals from your landing page, from your website, previous user behavior, all this stuff, broad match is able to identify people searching more broadly but they may, in fact, be closer to the decision point than you anticipated. That’s where it’s getting us, it’s opening up this ability to reach broader while still prioritizing the results that are important to your account.

Lauren: The biggest thing you said was closer to the conversion point and until or unless you have identified that conversion point and implemented the proper tracking mechanisms, this is not somewhere that you want to venture. You’re not going to use broad match on maximize clicks or you better believe everyone’s searching, do I need a dentist or how should I put ice on my tooth? It’s going to end up here campaigning. Talk to us about the necessity of conversion trackers to be successful.

Evan: Yes, there’s basically a couple of things. Look, let’s talk about why you might consider doing this. Trying broad match. For one, as we talked about, if again using a dental group as an example if they have very specific locations, clinics. You ultimately are limited by geography. You can’t expand your geography to people that won’t be willing to drive in to go to the clinic so you’re limited there in terms of increasing your available impression. If your campaigns are struggling to spend and scale up with you that is a sign that it might be time to be considering broad match. Now, I’ll say also the requirement for this new broad match to use it correctly is that you have sufficient conversion history in the account. The old-school way is still available, just doing broad match, constantly curating the search terms, and adding negatives.

You can still do that. It still has the same risks that it historically has in reaching out to super-relevant searches. This new model will require smart bidding. That is really what makes this work. Your account will need to have sufficient, we tend to say a campaign will want to have about Google varies what it recommends that around 30 conversions a month tends to be enough volume to support switching to smart bidding that will really be necessary for this version of broad match that we’re talking about because it doesn’t get the benefits I’ve described without smart bidding applied.

Lauren: The audience benefits, the ability to tap into those other signals.

Evan: That’s right. If you’re not using smart bidding, you can still get a search for do I need to go to a dentist and it’s still just as unqualified. The smart bidding allows the system to be like this person has already been searching, they’re ready to go, we’re going to bid on them anyways.

Lauren: Yes, I think you had a good point. You brought up who is this right for and scale and kind of reaching outside of what you think your typical bottom funnel keywords is definitely one of the ways to know if this is applicable. I think the other one that we should talk about is the niche audience. When you’re trying to use search to find the needle in the needle stack, the keyword alone isn’t enough to help you find them. We’ve got an example in our current management that is, let’s say a home care client. In home care itself, there are certain limitations or certain types of customers that we’re looking for customers that are more affluent, that can afford this service. Customers who aren’t using things like Medicare, for example, there are plenty of home care agencies that do support Medicare, but this isn’t one of them. The nuance of someone searching home care near me and having those additional signals, information on their past behavior, their affluence, their proximity to the neighborhoods where people may be able to afford this, whether or not they have certain insurance types, those signals are only applicable or available if you are on something like this broad match.

Evan: Yes, that’s right. What we know, especially in those situations where in the medical field where they don’t access certain insurances or Medicare or Medicaid, people are not searching. That’s the one thing with paid search, they’re not really going to self-identify that say, “I need home care, but I also have Medicare.” It’s not going to happen. They’re making a few searches per month, but nowhere near sufficient. People don’t search that way.

Lauren: Evan, we’re talking about the necessities to deploy this properly, one of which is obviously conversions, but that not all conversions are created equal. What do we really need to have in place? Not just 30 phone calls, 30 button clicks, 30 what? What matters to ensure that we’re getting the right traffic and not just more calls into your call center that are bogging down your operations team, but are actually becoming patients.

Evan: Yes, that’s a good point. When we look at it, this is applicable to any, even if you’re not considering broad match, it’s always better, especially in the medical field to be trying to get your send signals to the system that are deeper conversions. Typically, we see in the medical industry that there’s qualifying information. Can they pay out of pocket? What insurance do they have? There’s several criteria. Even just generally only optimizing towards conversions of did they call can miss some of that qualifying information. It’s one of the hardest things in the industry to work with your client or in your agency to make sure that you’re passing back as deep of conversion information as you can. Ideally in this situation, someone has called in and has been qualified by the sales team.

That’s super valuable information that really tells us that those people that did this are the ones we’re looking for. That is especially helpful when we’re talking about broad match and we’re saying, okay, system, we’re going to give you some looser terms to go after. We’re going to trust you that you’re going to find people that are relevant to us. That’s the best thing to feed through the system is that qualifying information so that the system has more things to go off of to get you the right people.

Lauren: For things like, in other healthcare situations, like an online booking scenario, there is the appointment booked. If you can get the new patient qualifier, for example, that just elevates the performance, it trends the AI towards the right things.

Evan: That’s exactly right. We’ve had clients come in, we’ve worked with them to approve, with online booking. We’ve initially been limited at, did they click the button to book. It’s helpful, but it’s nowhere near as good as did they submit the booking request. Did they self-select, I’m a new patient on the way to doing that. That’s really valuable information.

Lauren: Thinking about the fact that Cardinal started testing this, do you have any really quick examples you can share? Maybe one that stands out in your mind of a client that we ran this test and that broad match test outperformed and really proved that this future is one that we can succeed in?

Evan: Yes, there’s a couple of things I’d want to call off. One interesting thing about testing this broad match is that unfortunately, the value of broad match in your accounts is most obvious when it is not side by side with phrase exact keywords. What happens is typically phrase exact are more nuanced selections of keywords. The system, if you’ve got a dentist near me, broad match, and dentist near me exact match, the system will default to giving it to exact match. That’s the trick with testing this, is that you have to set up a situation where it’s completely segmented so they’re not existing in the same realm.

Lauren: We need clients to buy in to say, we’re going to test this, it’s going to be for 90 days, we’re going to shut off. What has generally been working, this isn’t an effort to ensure that we’re staying where Google’s puck is going. When that time comes and they say, gone are exact and phrase, we’re already teed up in the right position.

Evan: That’s right. What we’ve done to mitigate the risk of testing this, typically, if you have a new total net new account, you can start this from the get-go to some degree, but a lot of these accounts will already be on phrase and exact primarily. The trick there is to create that testing environment, pretty simple, select, you want to mitigate risk, but you also want to make sure you have enough data. What we’ve done is we’ve done Google ads experiment split tests. We take an existing campaign, leave that how it is as the control, and then the duplicate experiment campaign that runs side by side with it is only populated with the broad match search that we’re talking about here. That way that is an environment even in an account that currently you don’t want to eliminate phrasing exact and just cross your fingers. You can test this in an environment that does cleanly segment the two. What have we seen? We’ve run this across a number of accounts. There’s varying degrees assessed, so mostly we’re in a broad match is doing better area right now. That’s our learning so far.

Lauren: Is that both in the efficiency and volume categories? It’s not a drive towards efficiency, sacrifice, volume, it’s able to accomplish both.

Evan: That’s right. Okay. Let’s talk about the results we’ve seen, we’ve been running these tests for many months now across many different accounts and verticals, and we’ve seen really strong results. Universally broad match there’s some accounts that are still learning and have not quite reached confidence but across the board, we have seen better efficiency in the broad match test than we have seen against the original controls using phrase and exact. One example would be we have a plastic surgery group and we’ve seen every test we’ve run, broad match has proven maintaining the volume, increasing overall efficiency. That’s what we want to see. We want to make sure that we’re able to still maintain the focus and the efficiency while also opening up the potential impression share volume by using broad match terms.

It’s a great success. We’re going to flip the switch and fully convert these accounts to broad match only things to consider here. We definitely want to make sure that we are monitoring and changing the search query reports. Any terms over time you’re going to make sure that the keywords are in fact relevant. Anything that is not efficient, we’re going to eliminate you’ll keep doing that, but broad match will be able to open up that volume for you. What we’ve seen also, as a side note, just in terms of metrics that’s important to note, is that we’ve seen increased clicks, increase impressions increase conversions. Now interestingly, the conversion rate tends to be slightly lower on the broad match terms by its broad nature.

However, the efficiency is gained there because the CPCs tend to be significantly lower with broad match, there’s more searches, there’s more volume. It doesn’t have to be quite as competitive. That’s where we’re seeing the efficiency gains. Well, of course, with that in mind, as we discussed before with conversion tracking, you want to make sure to keep an eye on quality. If we’re giving up a little bit of room on conversion rates in favor of overall efficiency and scalability, you still need to keep an eye on the quality of leads. You’re going to have to make sure that the leads you count in platform calls, forms, books, submissions for appointments, that those are effectively translating into actual appointments and revenue. Overall, this has been very promising. For Cardinal, at least this is part of the future. It’s going to be a slow evolution. A lot of stuff in Google’s changing and moving towards smart bidding, but this is 100% one of the tools we’re going to be using significantly as we look to scale accounts, existing accounts as well as onboard new accounts.

Lauren: You heard it here first people, this is something you should be asking your agencies, your in-house teams to consider a test. It doesn’t mean it’s the perfect solution for you today. If it’s not, think about the steps that you can take to get to a point where you can test it, and just don’t be waiting around on exact match-only campaigns when Google officially terminates those match types. Evan, thank you for joining us. This has been an awesome episode of Ignite. Please like, share, comment wherever you’re listening, and we’ll see you guys next time.


Announcer: Thanks for listening to this episode of Ignite. Interested in keeping up with the latest trends in healthcare marketing? Subscribe to our podcast and leave a rating and review. For more healthcare marketing tips, visit our [email protected].


[00:20:28] [END OF AUDIO]

Healthcare Marketing Insights At Your Fingertips

Listen and subscribe to Ignite wherever you get your podcasts.

Get Started

Ready to Grow?

Great partnerships start with great discoveries. We start with your business goals and budget, and then help you find the right digital marketing strategy to fuel growth.

Fill out the form to get started!

"*" indicates required fields

This field is for validation purposes and should be left unchanged.