Google’s bidding change demands smarter budgets

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, Cardinals experts explore innovative ways to build your digital presence and attract more patients. Buckle up for another episode of Ignite.

Lauren: All right, everybody, welcome back to Ignite: Healthcare Marketing Podcast. I’m here with Evan Ilgenfritz. Ignore the last names. Too hard. Evan is our VP of paid media, and he is here today to speak specifically to one of Google’s latest updates, the target-based bidding change. I’m super excited for this episode, Evan, because I just returned from maternity leave, and I don’t even know everything about this change yet. You’re going to teach me live on this podcast episode what I need to know going into the next couple of weeks. Good to have you.

Evan Ilgenfritz: Thanks for having me. Yes, it’s exciting, really important topic, I think, for a lot of people in the industry.

Lauren: Yes, I’ve been hearing amongst our clients. The teams, the change management, the adaptation, the plans A, B, and C, depending on how it affected each business. Excited to talk through some specific examples. For those of you listening, if you’re not super clued into this, August 17th, Google changed how their Target CPA and Target ROAS objectives behave for campaigns that are limited by budget.

Now, I know for most of you listening, and most of our clients, because we are running hundreds, sometimes thousands of campaigns delineated at the location level, oftentimes limited by budget is one of our biggest challenges. When you cannot consolidate because you need to have control over budget, or where the traffic is going, or how much a clinic or group of clinics is spending, that limited by budget is probably the flag that is on all of your campaigns.

This was something we had to navigate with quite a few of our clients in that situation. If you are in it, this is the episode for you. The challenge was in the past– actually, you know what, I’m not even going to try to take the words out of Evan’s mouth. Evan, tell us about the challenge. Frame it up for us. What’s really going on here? Why did Google do this, and what are we up against?

Evan: The main change, as you mentioned, is on August 17th. That change specifically applies to campaigns that are limited by budget, although I would say keep an eye out for all your campaigns. The main concept is that when you’re using Target CPA or Target ROAS bidding in Google Ads, oftentimes you may have noticed that your campaign performance outperformed your target.

Let’s say you’re a campaign and you have your CPA target that you’re operating against set to $100, but you’ve been seeing that the past 30 days, even several months maybe, it’s been delivering at an $80 CPA, so below what you were asking it to deliver. That is going to change. I think basically it applies to campaigns that are limited by budget because Google is trying to make sure that it is distributing the spend as best it can

Ultimately, what’s going to happen is in that scenario, if you have a campaign Target CPA set to, say, $100, your performance has been overperforming your target and delivering a lower CPA. After this change, now the campaign is going to start leaning towards your actual target that you’re requesting. Where before, if you were overperforming your target, now it will not. Now it will start delivering right at where your target is, around $100 for a CPA.

Lauren: There’s a challenge because who doesn’t want the platform to overdeliver? Who doesn’t want it to come in lower? What are some of the controls that we have as the people managing, or that our clients have in advocating to whomever is managing their campaigns, to ensure that, okay, I had a $80 CPA and my target was set to $120. Now it’s just going to jump up to $120. What do I need to be doing about that, and how do I walk it back down?

Evan: What we did here at Cardinal in advance of this change, we methodically tried to walk our CPA targets closer to where it had been delivering. Again, in your example saying that you had it set to $120 Target CPA and you were delivering at $80. In the weeks preceding this change, we were slowly lowering our target closer to that $80 we were getting so that it didn’t have an arbitrary jump up and therefore a loss of efficiency when the change took effect. That’s a major one.

I think it’s always good to align your CPA targets with what you’re getting. We like to talk about them, or at least I do like to talk about targets for Target ROAS or Target CPA like magnets. If your performance is at $100, your CPA is at $100, and you set a target, you want them really close together when you start. If you want to improve that CPA, you slowly drag your target down and try to drag the actual performance down with you.

If you try to make a big jump, say you’re at $100, and your actual performance, and you say, “I’m going to try to hit $50,” it’s too far apart. You’ve lost that connection. That’s a really important thing when it comes to making adjusting for this is, first of all, you’re making incremental change. Google recommends to make somewhere in the neighborhood of around 15% adjustment per time when you’re adjusting your targets, either in Target CPA or in Target ROAS. You don’t want to do that big jump and therefore confuse the algorithm.

That’s a major one is making sure you’re being very incremental with it. The other thing that goes with it is you’re looking at how long your campaigns are going to be in the learning period. Google also says that the learning period is typically around three conversion cycles. That’s dependent upon how long your conversion cycle is. If you have an action targeting someone who comes in for an assessment, that may take a week or two after the initial lead. You have a delay. For actions that you’re optimizing like that, you probably have a longer window when you make a change to let it sit and stabilize before you make another change.

Lauren: Evan, in the old world, if you were getting a CPA of $80 and your target was $120, you most likely had it set as such because you were willing to tolerate $120, and you were trying to up your budget and see how far against that ceiling it would go. Now what happens when you’ve set your Target CPA to the $80? Let’s say you walked it down, you’ve got it there, but now you need to scale again. What is the reverse situation here? How do I then go back out and scale?

Evan: There’s multiple factors in here. I’d like to address the limited-by-budget component first. That’s a problem we see across the industry. I think, as you mentioned at the top of the podcast here, that people have multi-locations or tons of different things that they’re covering, and therefore their budgets are low. Limited-by-budget basically tends to mean that your budget is so low, it often is too low in relation to the CPA that you’re driving. Again, Google want to make sure you’re fishing in a big enough pond for your conversion. Google recommends that your daily budget is three to five to 10 times that of your expected CPA.

Lauren: A lot of people listening are like, must be nice, Evan, not going to happen.

Evan: There’s a lot of workarounds. To finish the analogy, if your CPA that you’re driving is $100, your campaign is better able to perform for you if your daily budget is maybe say $500. It is often when we get into your budget is $100 and your CPA is $100. Google really only has maybe one chance a day to convert. That’s when you run into struggle with the algorithm. Now, it’s not always an option. It’s rarely an option. In fact, just to increase your budget. There’s a lot of ways that we try to work around that.

One of them is, if you’re able, running shared budgets and attaching them with portfolio bid strategies is a major unlock. You get to retain the individual campaign design. Let’s say you’ve got a campaign for an Atlanta campaign and a campaign for Orlando. If there isn’t a reason that you need to isolate the budgets between them, you could combine their budgets. Now, the budget collectively is larger, and it can collectively drive better performance for these two regions. That’s one of the things we see as it relates to daily budgets.

I think, additionally, when it comes to making some of these adjustments is, like we said, you should often regularly be monitoring your targets for Target ROAS or Target CPA. I think in the previous world, when CPA may have been outperforming the target, you said. It may have been an indication that you could have, or should have, been lowering it anyways, not leaving at $100 when your actual delivery was $80. You may want to be lowering it.

I think we just have to be, especially for campaigns that are limited by budget, just very deliberate and aware of what the setting is, how do we march it down methodically without jumping too far away from the algorithm, and then repeating the cycle.

Lauren: Are there situations in this change, Evan, where you are suggesting changing bid strategies? Are there organizations that are better suited for something like a max conversions because this new dynamic is really threatening their performance or their ability to get to performance?

Evan: I think it’s certainly worth testing. There’s a lot of minor differences between, say, a max conversions and Target CPA. Testing is always the name of the game if you’re able to. In a lot of ways, target–

Lauren: Not just taking your whole account and flipping it over by picking a [unintelligible 00:09:13], too. Yes.

Evan: Limit your exposure to the test and make sure that you can isolate it, prove it works, and then consider rolling it out more widely. If you’re on Target CPA or Target ROAS, you can consider, if it works for you, increasing your targets to give yourself more leeway. If you’re running into trouble there and budget continues to be a problem, we recommend either shared budgets and/or portfolio bid strategies.

What I’ll tell you, that’s a really important consideration, especially related to the limited-by-budget crowd. We run into a lot of situations where a client, a business, can’t afford to consolidate budgets. They have to have very deliberate budgets for each clinic, each location, each market, and they’re not able to combine and share them. You can still gain major benefits algorithmically from just using a portfolio bid strategy.

Your campaigns retain their individualized budgets, so you’re not worried about muddying the waters or investing too heavily in one. However, you’re able to share more deliberately the performance results collectively for those groups, and the system can tend to perform better with that increase in data volume.

Lauren: Yes. I guess another question I would have, I’m never an advocate, usually never an advocate, for choosing a conversion action that is farther away from your end target, but is there a world in which organizations might be driving a little bit too hard? Let’s say I want someone to fill out a form, complete an assessment, come in for a tour, and actually start a service.

I’ve got my whole campaign oriented around starting the service, and that’s really what’s causing my limit in my budget because there’s just too few of them. Is there a world in which we’re recommending people really assess where in the conversion funnel and that maybe actually taking a half step back to something one step removed is actually the right move for their business? Should they be thinking about and testing that as well?

Evan: 100%. If you say what is the best-case scenario for Google to perform is that you have sufficient budget, you’ve got sufficient conversion volume, and that typically is per campaign, 30 of the ones you want in the past 30 days, more is always better. It’s volume of impressions. Do you have a lot of impressions that you can go after? Again, the bigger the pond you’re fishing in with the right conversion signals, the better the performance the system can drive for you. When you have limitations on any one of those areas, that’s when you need to get creative. We run into that situation a lot, Lauren, where we may be optimizing towards the best-

Lauren: Best business KPI.

Evan: Best business KPI, the conversion action that has the highest correlation with the outcome, typically new patient acquisition that they want. However, you’re also struggling by doing that with you’re starving the campaigns of that volume of conversion signals. In that scenario, you may actually perform better by re-including some less qualified signals. Now, what I’ll tell you is you have to do it the right way. That’s typically going to require value-based bidding.

Either maximize conversion value or Target ROAS, assigning the right relative values to these conversions. Then what you’ve achieved there is you’ve given the campaign greater signal volume of conversion actions while still telling it that these over here are more valuable. You can get out of the muck a bit because you’ve solved the problem of the conversion volume being too low for the campaign to figure itself out, but you’re still keeping a finger on the pulse of quality that it’s delivering.

Lauren: Just because something is a solution for where you’re at today doesn’t mean set it and forget it either. I think that’s something, if you were to take that approach, for example, and maybe reintroduce a conversion signal one step removed, keep an eye on it because algorithms will feed and learn from themselves. If that starts to head in the wrong direction, it becomes a death spiral. Again, that may be a solution for one of your markets, but not the other. It may work here, but not there. There’s all of these adjustments and tweaks that have to be made in response to this. There’s not one right or wrong answer.

I think that brings me to Evan to wrap this up in terms that we’re talking hypothetically, but this has been rolled out now for about 30 days. Have you seen major wins? Have you seen major losses? Have you caught any gotchas that you were like, “I hadn’t even thought about that,” or, “Man, I should tell other people that happened, and so we don’t make that mistake again?”

Evan: Yes, it’s a good question. In the wide scope of things, I think, certainly it’s a learning period. It’s a change. That requires, if we were properly prepared for it by adjusting those CPAs, any of those adjustments have some form of learning period attached to it. In the shortest of terms, I think there’s probably a performance decrease in the near term while you’re acclimating, while your campaign, while your plan to address it as the different options we’ve gone over today.

While you roll those out, you may likely see a ideally brief window of performance reduction because the system, the campaign, is learning with your new settings. I think one of the wins is that it’s a good reminder to be more deliberate and vigilant about your process and what our opportunities are for combining resources. Again, I can’t reiterate enough. You can unlock performance by increasing sharing budgets, or increasing the data volume of conversions going to a campaign, or increasing the number of relevant impressions you’re capturing.

Again, the bigger pond you fish in, the better performance you can go get. I think in searching for those options when dealing with a change like this, it clarifies some of the things that we maybe had opportunity before we made this change, and now we have a new pursuit of it.

Lauren: There’s going to be more to come, more learnings on something like this. It’s been live for about 30 days. Evan, if you’re talking to marketers who maybe saw a dip and didn’t quite know why, I think this is something for them to go check what their Target CPA is relative to what they think it had been. What is it set at? Can they work to walk it down? That’s a very quick here-and-now situation.

For people planning for next year in this adapted landscape, any advice you would make on how they think about setting their budgets or their ability to be more flexible next year that would help them be more future-proofed against stuff like this?

Evan: Yes. I think I would just say that it’s interesting that this change, I know I’ve already reiterated this a number of times, but this change is, for right now, specifically related to campaigns that are marked as limited by budget. That, while adjusting to this change in that the behavior of the campaigns will now be different when using Target CPA or Target ROAS, we have to make an adjustment. If we were overperforming our target and we want to stay there, we’ve had to make an adjustment.

I think what it also really speaks to is the challenge and the creative opportunities we have when dealing with limited-by-budget campaigns. They will always be a struggle for performance. If you have a campaign that’s got $50 a day and its CPA, its driving, is $100, you will always struggle to deliver performance reliably for that campaign. That’s how Google operates. This is an opportunity to align with the new setting, but also to re-explore those options.

Can we be flexible with sharing budgets? Can we put them in a portfolio bid strategy to give the campaigns more data to use? Can we expand some of the targeting we’re doing so we have impression-share headroom, so we have a greater ability to find and deliver good people to drive performance? I think that’s the major takeaway, is that if you’ve got a lot of campaigns limited by budget, start asking yourself those questions. Can we make some of these adjustments? Can we operate with how Google is going to work? Can we adjust how we’re running our strategy to squeeze out the best total performance?

Lauren: Evan, I would even take it one level up, maybe out of the media sphere, and question whether that clinic is getting enough volume from this channel to warrant continuing to invest in it. If I don’t have enough to spend to get a reasonable CPA and enough volume, should I be powering that clinic with more organic next year and taking that clinic’s budget and giving it to another clinic? All of a sudden, that clinic is no longer in the limited by budget. We can actually play aggressively in the algorithm.

I would say even more broadly, looking at your mix next year and assessing whether every clinic needs this low level of support, or we’re better off going a little bit deeper in some areas where this is the battle that multi-location marketers have to deal with every single day. The answer one day can change from the next day, but it’s this constant configuration between how much volume do I need and how efficient do I need to be. Think bigger than just your media mix as well, would be my recommendation.

Evan: I think that’s a great one. The last one I’ll add to is in terms of how do you get creative with giving the system more elbow room on data and spend. To your point of, do we remove clinic support off the agenda so that we can seed that budget elsewhere to hopefully get them out of the budget? Another thing to consider is ad scheduling. Can you turn off the weekends or turn off three days of the week so when you’re running that budget on those four, three, or four remaining days, it is higher? The budget is higher for the days it’s on, that’s another way to get creative and see if you can solve this limited-by-budget problem.

Lauren: Yes. Maybe in digging into operational data, you find that the ability to answer the phone on Fridays is not strong enough anyway, it’s just more in spend. There’s so many other things we could dig into. It’s really about how can I concentrate my effort in a period of time or with the right dollar amount to make it work really hard for me and not just hum along and chug along and never really get anywhere. Lots of creative ways to think about it. Again, you could have a multi-location environment where you’re deploying every single one of the things, Evan, that I talked about in different use cases depending on what’s going on.

Evan, thanks for teaching me something new today. Appreciate it. For all of you listening to this episode, more from Evan and our media team to come in our Scaling Up event coming up October 13th and 14th, so make sure to tune into that. See you, Evan.

Evan: Thanks, Lauren.

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 blog at cardinaldigitalmarketing.com.