Google shows you how to optimize Smart Bidding and other automations

Episode Description

There’s nothing quite like learning about something from the people who created it.

We’re excited to have Google join us on PPC Town Hall for a discussion on how to optimize Smart Bidding campaigns and other automations they’ve developed. This is a direct, closer look at Smart Bidding, use cases, and actionable tips to carry back to your accounts.

This panel covers:

• The difference between automated and Smart Bidding
• How to get started with Smart Bidding and validate success
• When to switch between bid types
• 5 stages of bidding from Google
• Connecting business data with optimizations in Optmyzr
• How to optimize Smart campaigns (shopping, performance max)
• Overall results on the Optmyzr platform & why Smart Bidding works well
• Optmyzr tools to use in conjunction with Smart Bidding

Episode Takeaways

Introduction to Smart Bidding and Automation

• Smart bidding is increasingly driven by Google’s advancements in machine learning and AI.
• The episode introduces experts to discuss how to optimize smart bidding and its implications for PPC professionals.
• Feedback indicates a strong interest in tactical advice about automation and maintaining relevance in a career heavily impacted by automation technologies.

Understanding Smart Bidding

• Smart bidding refers to Google’s automated, real-time bidding strategies based on conversion optimization, distinct from semi-automated or manual bidding methods.
• It is influenced by a multitude of signals, considering billions of data points to optimize bid strategies dynamically.
• Experts recommend setting realistic targets aligned with historical performance to avoid overwhelming the system and to allow for gradual optimization.

Setting Up and Optimizing Smart Bidding

• Proper setup involves aligning smart bidding goals with business objectives, utilizing accurate conversion tracking, and selecting the right attribution models.
• Continuous monitoring and adjustment are crucial; using tools like Google’s experiment features can help test and refine bidding strategies without fully committing.
• Integrating broader marketing strategies and maintaining a holistic approach enhances the effectiveness of smart bidding by informing Google’s algorithms more comprehensively about desired outcomes.

Practical Tips and Troubleshooting

• When smart bidding does not meet expectations, consider adjusting or revisiting campaign settings, such as conversion goals and bid adjustments.
• Utilize ECPC (Enhanced Cost Per Click) as a hybrid approach if transitioning from manual bidding to fully automated bidding.
• Experts emphasize the importance of understanding the underlying factors driving bid adjustments and recommend structuring campaigns to better inform smart bidding systems about business-specific goals and values.

Episode Transcript

Frederick Vallaeys: Hello and welcome to another episode of PPC Town Hall. My name is Fred Vallaeys. I’m co founder at Optmyzr and I’m your host for today’s session. So we wanted to talk about smart bidding and automation, and I know it’s a topic we’ve talked about quite a bit, but the feedback we’re hearing is Our customers, our viewers, they want to hear about tactical advice about automation and how to have your own career in PPC when so many things are becoming automated.

Now, obviously the driver of a lot of this automation is Google, and Google is a technology company. They’re a very good, they have great engineers, they really know their technology, they’re best in the world at machine learning, artificial intelligence. So it kind of makes sense that a lot of that change is being driven by Google.

So for that reason, I thought it’d be really great to bring in two Google experts to help explain how smart bidding works, how you can optimize it. If you’re already using it, or if you’re not using it yet, how you can start and actually get the results that some advertisers are seeing, which are really good. And then also open it up to questions from everyone, right? Some people here may have tried it, may have seen success, may have seen some failures. So let’s talk about it. Let’s figure out what makes campaigns on smart bidding successful. And so for that, we we’re bringing in two Google experts and let’s get rolling with this.

All right. So my two guests today, I have been on PPC time hall before. So welcome back, Emmy and Peter. How are you guys today?

Emi Wayner: Very good. Thank you. How are you, Fred?

Frederick Vallaeys: Very good as well. And you may hear some kids in the background, but you know, that’s how things are these days, right? Everybody from home.

So so Emmy, tell us a little bit about yourself.

Emi Wayner: Sure. Hi everyone. My name is Amy. I’m a platform partner lead at Google from channel partnership team. We focus on my tech and platform partners like Optmyzr and we really excited for the partnership with Fred today and really looking forward to the conversation.

Frederick Vallaeys: Yeah. Thanks for coming back on the show. I mean,

Emi Wayner: yeah.

Frederick Vallaeys: And Peter, tell us where you’re calling us from and how things are with you.

Peter Oliveira: Yeah, thanks, Fred. Doing good. Thanks for having us here again. So I work very closely with me on partnerships with some top PPC optimization platforms like Optmyzr. And my focus has been really to work with these companies to build third party products on their platform that will make our products both easier to use and a bit more scalable, right?

So a lot of what we’ve been focusing on together has been building bid optimization tools or tools that have really great synergies with smart bidding. Thanks, Fred. Which is what we’re here to talk to you all about today.

Frederick Vallaeys: Very good. And so just I know a lot of people watching have watched us before, but just to explain the the way that we take questions on this live stream is by putting them in the comments.

So if you’re watching on YouTube, just put in a comment with your your question and your name will show up and I will just put it on the screen. And to get us all warmed up, usually we ask, Hey, where are people calling us from today? So if people want to start putting that into the comment section.

Then then we’ll show you on the screen and Pedro, hello. Don’t know where you’re calling from today, but but that’s exactly how we show the questions on the screen. Okay. So so yeah, the topic today is smart bidding, right? And there’s this whole notion of the wave of automation. I was actually thinking about it recently, how in the mid 2010s, every year was like, this is the year of mobile search.

This is the year of mobile search. And I feel like nowadays in the late 2010s, we’re, we’re kind of in that phase and then the 2020s, I guess, yeah, just 2020 year you want to forget about, right. But but now we’re in that phase where every year seems to be the year of automation. And it’s really being driven by technology, and Google is such a leader in that.

So how is that changing things when it comes to how search marketing is done?

Emi Wayner: Yeah, I can jump in. I think what we see today is really the evolution of machine learning. I think machine learning is getting smarter and smarter, and people are more comfortable about using the machine learning and then leveraging that to to the next level.

So that’s what we’ve seen. But you really want to step back a little bit and share one slide with everyone today. If you can share one on my screen. Yeah, thank you. And this is really to tell you like, let’s think about what we do every day as a consumer. We search so differently compared to three years ago, right?

15 percent of queries on google. com I knew every day you That means there are tons of queries that we’re missing as well. So we really have to depend on the combination of machine learning and our knowledge and skills to really to go to the next level. So that’s that’s one insight that I want to share.

And the second slide that I want to share with everyone before going to the more tactics is this. We had an executive summit earlier, and when you think about the online advertising as a whole the machine learning is getting smarter. And the data that you provide to machine learning is going to be the really key to be competitive and to go to the next level and, and achieve your goals.

So that’s what something we’re really looking at this year, not just from CPA perspective, But also broader profitability and lifetime value of your customers. So we’re going to talk a little bit more about that in a later slide. But that’s as a thought starter, I wanted to just bring it up to the attention.

Frederick Vallaeys: Exactly. And I think you guys call it modern search, right? So basically the point is we’ve Been doing PPC now for, you know, myself for over 20 years. And we’ve got all these like best practices and we get single keyword ad groups and we manage our bids and we do bid adjustments. But, but I guess the point that you’re making is the types of things we need to focus on today, given smart bidding are a little bit different than what we may be used to.

Right. And so I still want to set the framework here a little bit before we get deep into the tactics. But Peter, maybe tell us a bit about, like, we call this smart bidding, we call it smart campaigns, like, and then we also have automated bidding. So how are smart bidding and automated bidding different or are they the same thing?

Peter Oliveira: Yeah, so I think a lot of times people can use the two terms interchangeably, but here at Google, we do have a pretty clear distinction between the two. Usually when we’re saying smart bidding, we’re typically referring to fully automated. Auction time conversion based bidding strategies. And those nowadays will consist of target CPA, target ROAS, maximize conversions and maximize conversion value.

And we probably wouldn’t use the term smart bidding as much anymore to refer to like a semi automated bid strategy like ECBC or non conversion based non auction time bid strategies like target impression share or maximize conversions. So that’s what we’re referring to really the conversion based auction time strategies.

Frederick Vallaeys: All right. And so, you know, I’ve long, I’m an ex Googler, right? So, you know, some people say I’ve drank the Kool Aid, but I fundamentally do believe that there’s a real benefit to smart bidding and, and, and using Google to do your bidding and basically when you think about it this way, right? So automated bidding, there’s a lot of tool vendors that say we automate your bidding for you.

But what they do is kind of what Google used to do. They would say on a daily basis. What was what was the conversion rate like yesterday? How much did we sell yesterday? How much did the clicks cost and then based on those signals? They make an estimation of what’s going to happen tomorrow for the next 24 hours and they set a bid that seems relevant For that next period of time but that one bid which you have to set at the keyword level It sits there for every query that happens.

Every search term that comes in every geography, every device type, like there’s no differentiation, right? So it’s still kind of a sledgehammer approach to doing bid management. Whereas the thing that’s so unique about Google and smart bidding is that literally at every single moment when a search happens, and I don’t know if you guys know the latest number of how many searches happen, but it’s like billions a day, right?

So for every single search that happens. In real time, it’s calculating what is the correct bid based on all the factors pertaining to this auction. And I know you guys have some numbers on that, right? Like, is it, how many, how many signals are being looked at in smart bidding?

Peter Oliveira: Yeah, I think we’re looking at billions of signals to set the right bid for each and every auction.

And I think, do we have a slide on that too? I mean, we’re looking at a bunch of different factors. Let me share this with you.

Frederick Vallaeys: Yeah. And I mean, if you go into presentation mode on that, we’ll see it a little bit bigger.

Peter Oliveira: Yeah. So like, as you can see, like, we’re looking at a bunch of different things in those 70 million signals.

And we’re not just looking at. The signals themselves, but also like the intersection of all these different signals, right? A lot of the things that you use to optimize manually with like devices, they have weak time, location, all the different keywords, right? All of those are sort of taking care of you for smart bidding, but there’s still, there still are some tactics that you can use in order to like inform smart bidding of what the right way it can treat some of these things are, which we’ll go deeper into later in our talk.

Frederick Vallaeys: All right. And so I think an interesting point is that there may not be. Billions of specific signals to look at, but it’s sort of the intersection of the signals and in the book that I wrote, I made the point that even if you looked at something as simple as audiences and like demographic age ranges and two or three fairly basic signals that most people would agree on.

May have some impact on the likelihood of a conversion to to happen for a single keyword, you could very easily get to 180, 000 different combinations of age range plus gender, plus something like household income, right? If you just multiply them out, that’s a single keyword. Right. And so that kind of alludes to the complexity.

Of doing this sort of you know, bid management, if you will. Sure. I mean, I guess you could do it manually, but you’d be spending more time than it’s really worth. Right. And so that’s the whole point of why smart bidding is different from automated bidding and why it works pretty well. So like Emmy was saying, you really, it’s not set it and forget it.

There are still inputs that you can give into smart bidding. And so we’ll talk in just a minute about how you set it up and how you optimize it. But I think you guys also have a slide from some Optmyzr data, right? So Google took a look at all of the the accounts with an Optmyzr that are using smart bidding and how those were doing.

And so I thought that’s what that was interesting. So let’s take a look at that.

Emi Wayner: Yeah, let me share this with you. So this is a recent performance review and it’s a selected samples focusing on U. S. accounts. And when we looked at this that smart bidding outperformed manual bidding across most of the spent buckets in on optimize our account, right?

When you look at on the right hand side, a lot of people should advantage should take advantage of smart bidding if you would regardless of the budget. So when you look at on the right hand side, under 500 is a 500 budget dollar budget. And we see significant performance, even smaller budget as well, all the way to above 50, 000.

This was a very interesting data set for us to really think, okay, what can we do for, for that users and smart bidding users as well. To provide better solutions and provide even build a better solutions to optimize marketing with Optmyzr.

Frederick Vallaeys: So to make it very clear here, the point is Optmyzr is not a bit management company, right?

So we’re not comparing doing your bids through Optmyzr or doing your bids through Google. I think this is the holistic picture of Optmyzr can be used to manage budgets, to manage your search terms, find negative opportunities, help you Excluding geographic areas that might not be performing quite as well.

Right? So when you layer all of that together with smart bidding which is the most sophisticated bidding system in the world, and if you use Optmyzr to change your targets and connect your targets to your business data, that’s when you can start seeing really nice performance, I think the other point that I took from this slide was on the higher end, right?

If you’re spending more money, you probably just have more data points for the machine to. Analyze and optimize with and if you’re on the lower end of spend, well, then even if you’re managed through an agency, the agency probably not paying as much attention to your small account as it might be to the medium accounts.

Right. And so that’s, again, where smart bidding just far outperforms anything that you would do manually, but probably not quite as actively as you should. And then sort of that middle of the range that we’re, that’s where we see Humans being relatively active and having kind of using their personal insights.

And in, in some cases that might trump the machine’s prediction because there’s just not that much data for the machine to make a prediction to it. So that’s kind of how i’m reading this chart. Yeah. So we, we have some questions. So let’s hop into that. So Peter, this one might be a good one for you, but that’s Peter from Peter.

So Peter’s asking what about audience lists should we put into the system and how are those considered in that real time analysis of billions of signals?

Peter Oliveira: Yeah. And I, I saw that question come in and I just dropped in a slide to the last one on there to, to have a visual, to talk a bit about this. Essentially the audience lists that are taken into account. Are going to be your your remarketing, your customer match, and your similar audience lists other audience lists like in market or the detailed demographics that you split up or affinity audiences. If you add those smart bidding won’t necessarily take those into account, right?

Smart bidding is already taking a lot of different like factors and signals into account, which are included into those audiences, but it won’t necessarily change the performance per se, but a benefit of continuing to add those will be so that you can learn which audience are performing better or worse.

And then in the future, if you want to break out a separate campaign for certain audience types, you’ll have an idea of the different ROAS and CPAs that those are performing at.

Frederick Vallaeys: Right. And so a lot of these things, it’s also helpful just to get the measurement, right? So some things are highly defined, like geographic locations, you don’t need to attach locations to a campaign for Google to be able to give you data on how things perform in different cities.

But when it comes to audiences, that’s where you can get pretty crafty and do that definition yourself. Okay. Well, we’re getting more questions and these are kind of leading into what we’re going to talk about anyway. So why don’t we go with the questions? Okay. So Adrian’s asking us about discovery campaigns.

And so these are more automated campaign types that combine automated bidding, but also automated targeting, automated messaging. And unfortunately didn’t didn’t get any conversions for a good amount of spend. And I guess the question more broadly than is how long do you test something for before.

Deciding if it’s working or not.

Emi Wayner: I think Peter is frozen.

Frederick Vallaeys: Yeah, Peter froze. That question just literally froze him. So Emi, do you want to take it? Yeah

Emi Wayner: I’ll cover it until Peter comes back in. But we’ve seen actually a pretty good performance and then use cases with Discovery Campaign. Even for the first time in a midsize accounts.

So I I would encourage to continue looking into why it’s not delivering a performance. But Peter maybe can tell you more in terms of the how long we look at, we usually test in a pilot for a week and then we Regroup with the customers and then we reset and then regroup again. So one week at a time, but Peter can maybe jump in to give you more tactical advice too.

Frederick Vallaeys: Well, I would also, while Peter kind of warms up to this question again, I would also make the point that discovery campaigns, kind of as the name implies, is probably at a different stage of the consumer journey. So it’s not quite as close to the ultimate conversion as you may be used to. So as you use these campaigns, make sure you see its impact on what else is happening across your campaign.

Think about it more in terms of how you may have been advertising on Instagram, on Facebook. You know, in the beginning of Facebook ads, just literally taking your Google ads and putting them on Facebook was basically. A great way to waste your money. Like it didn’t work. And it took a while for people to figure out, well, how is it different?

How does it work on there? And so again, I don’t know exactly how you ran those campaigns, Adrian, but I think those would be some of the questions I would ask.

Emi Wayner: Peter, anything? Yeah. Yeah.

Peter Oliveira: Yeah. And I think it’s discovery campaigns. I’ve seen like some very strong results for, for certain advertisers. Like I work, I worked with a real estate partner for a while and they were even seeing some better results on discovery ads.

Then they did on search even which isn’t something that I’d go ahead and like promise to start out with. Right. But we can see some pretty strong performance. I’d say if you’re having issues with, with the return and with getting conversions, make sure that we’re setting our CPAs properly. Right.

And that we’re not setting unrealistic CPAs in that sense. And that, you know, those represent your goals. And then there’s also the flip side, right? If you look at our. Recommended audience so far, the audience that we feel comfortable recommending discovery ads for there’s usually the daily budget we usually recommend is about like 20 times the target CPA.

But if you’re really on a tighter budget, maybe about 10 times the target CPA. That being said, I have seen success at lower CPAs, but it’s, if you’re, if you don’t have that kind of budget, it’s not something that we’re going out and proactively recommending at this point. Wasn’t it also related to remarketing as well?

Emi Wayner: So like a leverage of remarketing with the discovery ads and a search altogether. Yes, absolutely. If, if you layer on audience lists like remarketing and customer match, you may be able to see some more successful results. Right. And so that goes back to the audience question. So Peter had a follow up as far as how to segment out the time, any recommendation on that?

Peter Oliveira: Segment out the, the audiences over time.

Frederick Vallaeys: Yeah. So he asked the audience question, right? So he’s saying, should he segment the list based on 30, 60, 90 days, or just one big 540 day list?

Peter Oliveira: We can check in what the latest best practices are for the audiences. I don’t have our latest on top of my head.

Frederick Vallaeys: So I’ll give my take on that.

And I think it depends for every advertiser. So, you know, if you have a very seasonal business say you sell Christmas trees, then maybe it doesn’t make a ton of sense to run. A list that looks back into summer when maybe right now things are very different as people get ready for Christmas. So that’d be one point to it.

But, but then I think the other point kind of to what I said earlier is with locations, for example, you can always, you can six months from now, you can say, Hey, well, I ran a campaign to the whole U S like, how did I do in New York? And Google has that data, right? And then you can make a decision on that data.

You can say, maybe I should have excluded New York when it comes to audiences. So it’s like, The answer here is that you, you may not know, like, is there a difference between a 30 day and a 60 day and a 90 day audience? And so add those lists, but Adam, for observation purposes, and then see, is there a difference, right?

And if you see there’s a fairly significant difference, then yes, maybe keep them segmented that way. And either you optimize with it or you let Google optimize with it. If you see that it’s basically no difference in conversion rate. CPA, ROAS then maybe you’re just doing extra work, maintaining several lists, so then you can go back to a single list.

That’s how I would approach it. Yeah. That goes back to the customer value conversation that we had

and so, and that’s where Optmyzr can then come in and help you maintain some of those lists so we can do your custom customer match audiences. We can automatically update those for you.

We also have a nice little. Tool that if you’re a customer, you may not have seen it in the menu. It’s kind of a secret tool, but we have a quick linker so that you can attach a lot of audiences in bulk to ad groups, campaigns, et cetera. So you can just email support at Optmyzr. com and they can point you to that feature. All right. So let’s let’s continue here with the agenda we had. And we’ll come back to some of these other great questions in a minute. But okay. So. Smart bidding whether we want to do it or we don’t want to do it, it’s kind of the wave of the future. Some people are clearly seeing success, even if not everybody’s seeing success.

So what can we do to at least set ourselves up to have a better shot at success, right? So how do you, what are the prerequisites? Let’s talk about those. Peter, you want to kick us off with maybe like your top or say even like, what are the biggest mistakes people make when they turn on smart bidding?

Peter Oliveira: Yeah, for sure. So, I mean, there’s so many different things that I could call out here. I think to, to start really making sure that you know what goal smart bidding is optimizing towards and that that goal aligns with what you’re trying to achieve. Right. I’ve seen so many times where like someone’s maybe set like a maximized conversion campaigns on a maximized conversion strategy on a campaign that was maybe.

Spending half of its budget, and then we didn’t see positive results. CPA wise, right? That’s because maximize conversion isn’t necessarily looking at to optimize your CPA to a given point, right? It’s like, you know, get you as much convergence as possible within your budget. Right. And if you’re only spending half of your budget before.

Opting into maximize conversions. We don’t necessarily know whether that incremental conversion will come at a CPA that aligns with your historical ones or if it doesn’t, right? So that’s a common mistake that I see as well. The other thing is the target that we set. Ida look into our optimization score recommendations in the front end to see what we’re recommending there.

But you probably wanna start by setting a target that’s in line with what you’re historically performing at. If you’re historically performing at a CPA of $100 and you try to switch that down to 50, we can massively end up throttling the amount of convergence, the amount of sales that we’re getting from our advertising, right?

Frederick Vallaeys: The point there is this is not the time to be aspirational and like, tell Google what you really want to hit. When you start with something, it’s better to not shock the system. And I think that goes to a lot of the points Google makes to like, don’t change your target too much. Don’t do this during a time of heavy flux and conversion rate because machine learning, right?

It’s learned something based on how things have historically worked pretty consistently. And if you didn’t say, well, Hey, great. What you’ve been doing in the past, but like now double my conversions. It may take a while to get there. I like. Think about a gradual approach as opposed to shocking the system.

Peter Oliveira: Yeah, for sure. And you know, just because we don’t want to necessarily start by telling smart bidding the exact target CPA that you want, doesn’t mean that we can’t eventually like work towards it over time. Right. I say, start with what, what you’re historically performing at. And if that’s not where you want to be, let’s slowly start adjusting so we can get there, right?

If you want to eventually get your 100 CPA to target a 50, let’s try 90 and then let’s give it some time Let’s try 80 and eventually work that down and then you can sort of weigh out with the trade offs, right? If When you go down from 90 to 80, you end up losing half of your conversion volume. Maybe you want to like re evaluate getting it back down to 50, right?

So those are things that you want to take into account as the target CPA setting and or the target ROAS setting can be a very iterative process.

Frederick Vallaeys: Yeah. And so that’s where I wanted to show some of the capabilities and Optmyzr too. And this is a generic town hall, right? So I don’t want to talk too much about Optmyzr, but we wanted to get tactical.

And so one of the things we do on the Optmyzr is we have this rule engine capability. So I’m going to put that in the stream here. But this rule engine basically lets you do exactly what Peter was just saying. So you can say if my bid strategy right now happens to be target ROAS or target CPA and I’m meeting certain criteria for conversion volume, for performance levels, then I can make increases or decreases to my targets.

One thing that always confuses me is the target CPA and target ROAS, they move in opposite directions, right? So I always have to think really hard about which way do I move it. So be careful with that. But this is exactly one way that you can do it and you can do it gradually. And then because this is an automated rule through rule engine, you can say, okay, one week later, we come back to it again and we actually analyze like, are we still happy with the conversion volume that we saw?

Is it still above some threshold that we wanted? Even though we may have told Google to. Get us cheaper conversions. And so long as that keeps working, you know, you drop that bit a little bit more, and then once it stops, you hit that inflection point. That’s when that automated rule can stop lowering your target.

And now you’ve kind of found maybe your sweet spot. So it’s, it’s basically an optimization curve discovery type system. So that’s the one way to do it. There’s another question, maybe a good time to bring it in now, but what is the preferred campaign structure? Right. So as we’re talking about setting up campaigns and giving you the best probability for success.

And kind of the question here that may be underlying this is. Now, okay. Google calls it modern search, right? So I don’t know what we call the thing we did before old fashioned search, but old fashioned search was pretty granular. We do things like skag single keyword ad groups. We’d be replicating campaigns for mobile and desktop devices.

We’d be doing all these splits and making massive, massive accounts with lots of segmentation. When it comes to automated bidding and smart campaigns and smart bidding, Google often says, well, don’t worry about doing that granularity. For me, the question is then really like, is that granularity hurt you, or is it just that it’s no longer quite as necessary as it used to be?

Emi Wayner: Yeah, that’s so, you know, after this call, we’re going to share the white paper and then there’s a very detailed instruction How to structure your campaign for smart bidding just for everyone.

Peter Oliveira: Yeah, and To your question, Fred, does it does it hurt to segment? Not necessarily. When we have a, when we segment our campaigns smart bidding actually can take data and learn from performance that’s happening throughout the entire account, right?

So while you don’t need to simply, you don’t need to break everything up. If you do end up breaking up your campaigns, smart bidding, we’ll be able to learn from performance from the other campaign. The only thing I’d caution about when we’re breaking it up too many ways is if you break your campaigns down to the point where there aren’t, there isn’t a lot of conversion volume in each individual campaign and you have you know, target CPA or target row as targets on those that your campaigns may not be able to meet.

That’s something to watch out for because the more constraints that you place on the targets, the more you segment the campaign. And add different constraints, you could end up like throttling your overall performance

Frederick Vallaeys: To be very clear here. Right. So to, to kind of like fundamentally how does the system work?

So Google looks at conversions on the one side, it looks like queries more or less on the other side. And so it doesn’t really care where those queries are coming from. So long as they relate to a specific type of conversion within smart bidding, each smart bidding is sort of its own conversion tracking and prediction mechanism.

And so the point is that If you have five campaigns, but they all happen to be using the same bidding automation and bid strategy, that’s fine. They can share data between each other. But if like Peter says, you have 50 campaigns and each one has a slightly different target then they may not be sharing data quite as much, right?

And that’s where you need to maintain some consistency of conversion volume. Now one thing I find fascinating, I don’t know if advertisers have noticed this, but it used to be that for smart shopping campaigns to be set up. You had to have conversion tracking enabled because obviously like we’ve been saying smart bidding is conversion based real time bidding, but nowadays that’s no longer a requirement, right?

So it kind of leads to the assumption that Google is basically able to figure out, Oh, what products are similar. from other advertisers across the whole ecosystem. And let’s make some predictions based off of that. I don’t know if you want to comment on that, but what I’m seeing is machine learning is very good at picking up commonalities across the whole system and using those to still make good decisions even when there’s a lack of information from your specific account.

Peter Oliveira: Yeah, and that’s also part of the reason why we’ve over time been able to go from like strict conversion requirements on our smart bidding solutions, right, even on search before there used to be requirements for target CPA and so on. And now we’ve brought a lot of those down to zero. A lot of people wonder, okay, how can our smart bidding accounts perform when there’s No conversions beforehand, right?

Smart bidding is looking at conversions and then adjusting over time. It’s because we have contacts from, from, from other auctions and we’re able to identify similar ones to yours and use that to calculate a conversion likelihood, even in the absence of historical data from your specific campaign.

Frederick Vallaeys: So I mean, to answer that question then about campaign structure, right?

I know we’ve haven’t really given you the answer. And I think the reality is there is not necessarily one right answer. The way that I think about it again is And let’s maybe simplify it, right? So say that you have Legion and you have different types of leads that you’re generating that have a different value, right?

So you want to have a different target cost per acquisition for each. Then it makes sense to split that out so that you can have different targets. Or obviously you can set those targets at different ad groups, right? So your structure should support your ability to set your goals and your goals, by the way, should be fundamentally related to your business goals.

Okay. Right. Nobody really cares about target CPA. Nobody really cares about target ROAS. What you really care about is revenue and profit and probably growing both of those. And these metrics do relate. To the metrics that Google gives you. Right? So if you want to have a break even profit and you have a 33 percent margin on the products we sell, then you have to have a 300 percent target ROAS, right?

So you kind of do the inverse there. So all of these things are connected. So that is really the first level of structure that I think people should put in. And then you can start measuring things and you can say, well, listen, again, maybe we found that there is a huge difference between how things perform in different cities.

And actually, maybe it’s not so much that the bidding is wrong by Google, but we would have done much better by having a different message for the East coast of the U S versus the West coast of the U S. And we can’t do that unless we have multiple campaigns, right? So then it makes sense. Put in multiple campaigns.

To optimize those campaigns and then the bid management from google is automatically going to set the right bids And still get you the benefits and then hit your t row as your tcpa.

All right

Emi Wayner: We have lots of questions coming in fred.

Frederick Vallaeys: Yeah good. So let’s let’s keep going on those questions but before we do that, so, Peter, when we had talked about like setting up your campaigns the right way. So no, so notes that I had taken before was understand your goals, attribution model selection.

I don’t think we’ve covered that. We want to quickly cover that one.

Peter Oliveira: Yeah, for sure. Attribution model is extremely important for smart bidding. And we do highly, highly recommend a non last click attribution model whenever you’re on smart bidding. The reason for this is that a non last click attribution model essentially gives smart bidding the transparency and allows it to take into account.

Every touch point that happens along the path to conversion. Right. And by doing a last click, you’re essentially restricting it from taking into consideration the other data. Why is that an issue with smart bidding? And why is it even more so of an issue with smart bidding than it was with manual bidding before?

With manual bidding as a PPC expert or someone who’s doing this every day, you might be able to look at keywords. There’s certain segments of traffic that aren’t converting that, right. That have lower conversion volumes overall, but you as a human, right. You can say to yourself, well, this isn’t converting because We know that it’s, it’s converting along the way, right?

It’s maybe not the last click conversion, but we know that it’s bringing a lot of value. Smart bidding won’t be able to do that as effectively. If you’re on last click, it’ll see that certain things are converting and certain things aren’t. And for the things that are converting, we should signal that as a bit increase.

And for things that aren’t converting, we should signal signal that as a bit decrease, right? So in order long last click attribution is crucial so that it has access to all of those data points. And since we’re dealing with machine learning here, right? The more data that it has to learn from, the better, the better it will perform.

Frederick Vallaeys: And that’s what it leads nicely into this next question. Which is how much time should you give automated strategies? And if you find that one is not performing after whatever amount of time should you just switch between strategies?

Peter Oliveira: Yeah, for sure. So we definitely want to give it a couple weeks at first so it can get out of the learning period, and then the exact amount will depend on whatever your conversion lag is, right?

Some, some companies, some industries may take a long time from click to conversion, and some others may be a lot quicker. So typically the sweet spot will be like a few weeks. I know it’s not like a very clear, clear answer, right. But maybe three weeks to six weeks. And yeah, if it’s not, if it’s not achieving at schools properly, then it might be time to switch over.

But you ask yourself first, is it achieving the goal that I’ve set out for it? Right. If I’m putting it on maximized conversions, is it giving me more conversions? If I put it on target CPA, is it achieving getting more conversions at the CPA that I set? Right. And also make sure that we’re using a lot of the best practices, right?

Like the optimization score recommendation is a good point for that. But there’s obviously so much more that you can do, right? Leverage broad match keywords. Are you leveraging a non less like attribution model? Are we potentially using this in conjunction with something like RSA? If you can check off like all the best practice that it’s still not performing well.

Yeah, you shouldn’t continue doing like it. Yeah. A certain strategy that isn’t working well for your business.

Frederick Vallaeys: Right. I said, so one thing I want to show here is again, how would you kind of take what Peter was saying and put that into practice and not spend so much time, like constantly checking stuff.

So the rule engine and Optmyzr, we we have these prebuilt strategies. So this one is related to automated bidding. This first one is audit, maximize conversions. And what you can do through here is basically look at conversion volume. So I’m going to give full control so you can see exactly what this rule is going to do.

But you’ll see the logic here, right? So if your conversions volume for the campaign is exceeding some threshold that you set, and you can pick this threshold. So you can say, I really only feel comfortable when there’s a ton of conversions. Cause that’s what I think Google is going to have the best performance.

So we’ve set it at 50 here, but if you say, well, I just, I want five conversions. That’s fine too. You can set five conversions. And then you can look at stuff like, have we fully depleted the the budget roughly. Right. So has my cost been greater than 30. 4 times my daily budget. Right. So so our cost is, so we’re basically budget capped.

We have tons of conversions. So then we can say about maybe instead of just doing a maximize Conversions strategy. Let’s go to something a little bit more restrictive where we also bring in goals like TCPA and TROAS. And you can do the flip side as well. So if you have fewer conversions than what you would like, then you can put a label on that campaign that says, Hey, maybe we should revert it back to manual bidding because it’s not going to perform so well, given that for whatever reason, conversions have dropped.

By the way, if conversions have dropped. It’s definitely good to know, right? You, you should look at it. Is it smart bidding that’s broken or did your landing page break or did something again with 2020 being kind of a crap year, did something else happen that we should pay attention to? And maybe we should change the messaging.

Maybe we should just pause that campaign. So these types of rules really. Get that daily report that says here’s what’s ready to move to the next stage of bidding. And here are the things that may be, should be brought back and re evaluated manually first. So yeah, that’s how you we, we, we came here from attribution models and sort of a roundabout way, I think. I mean, let’s talk maybe a bit about. Marketing is a holistic picture, right? So I get customers coming to Optmyzr and they’re like, Hey guys Google does smart bidding and automated bidding. Like, what do I do in Optmyzr? And I have an answer for that, but like, talk about marketing as a holistic picture and why it’s not just about bidding.

Emi Wayner: Yeah, so we think a marketer can play a bigger role than than just strategy for CPC and PPC marketing. So what that mean is we look at marketer as growth engine for the company. And this is a great opportunity with machine learning and what Optmyzr is going to offer for everyone. is that you can be that really gross engine of your company.

Really thinking about profitability and the value of the customers. Those are the strategic part. And that’s going to bring the value for, for your work and for your business. So really want to encourage everyone to think, not just PPC, but How are you delivering the value to the company in a business overall?

That’s kind of high level that we’re thinking about.

Frederick Vallaeys: Right. And I would say the way that you deliver a lot of that value is again, don’t just think about bidding as one thing that you do by itself, but think about it as part of that bigger marketing strategy. And so let me give you an example, right?

So say you’re doing manual bidding. And you have selected your exact match keywords. Remember an exact match keyword stopped being exact match many years ago. Now there’s close variance. So Google is going to show that keyword for potentially some search terms that it thinks are quote unquote, same meaning.

Some of these are great. Some of these are not so great. They, they are kind of in the same realm, but they’re definitely not going to convert at the same rate that your exact match keyword was going to convert. If you’re on manual bidding, here’s the problem. You’re going to pay the same amount for that low quality, kind of less relevant search term than you do for the one, the keyword that you actually chose.

So if you think about these things independently, then yes, you are going to run into issues. But if you think about putting together smart bidding, responsive search ads, And still doing some search terms management and looking at your close variants, all things you can very easily do an Optmyzr. Now you’re looking at it holistically, right?

You’re letting all of the automation systems interact with each other. And now what happens is yes, Google shows you for a close variant. That’s not so related. But it knows it’s not so related and it’s going to put in a lower bid and it may still get you that additional conversion. And that’s kind of what Emmy’s talking about.

It’s the, it’s the incrementality that’s being driven and that’s facilitated by you putting all the pieces of the puzzle together. Right. And now you can look at that. Through Optmyzr and you can say oh interesting There’s a search term here that I I thought it was never going to convert. I was going to be bad But actually it did convert just at a much lower cpc Could we do something could we boost the quality score on that by putting it as a keyword by writing a better ad text?

And the ad text by the way, that’s like rsa optimization, right optimizing responsive search ads These are all the pieces of the puzzle if you keep doing that Well, you’re You’re going to be better than the average advertiser, right? And I think fundamentally where we’re moving, and I’d love your take on that too, Peter and Emmy, but Google puts in place all these automations to make it easy for anyone to be an average advertiser, but if you’re listening to this and spending an hour with us, thank you first of all, but, but you’re not here to be an average advertiser, right?

You’re looking for a way to beat the system and there are ways to beat the system by putting in that extra effort and putting all the pieces in the right order.

Emi Wayner: That’s it. Yeah. That’s great. Some are yes.

Frederick Vallaeys: Very good. Okay. So the let’s say now let’s move on topics here. Right? So say that you are currently running smart bidding and maybe you’re not seeing the performance.

And so we have a couple of questions around that. So let’s jump into those. So what do you do from a specific optimization perspective? Once you are using smart bidding I had a question here. Where’s that question? I mean, so many questions coming in, but somebody was asking the question. About that

Emi Wayner: performance.

Frederick Vallaeys: Okay. I don’t see it anymore. Oh, here we go. So Adrian is asking

so again, in terms of timing, he’s asking, like, if it’s not working that well, do you move it to a semi automated management? And that’s an interesting question, Adrian. So you actually do have a choice, right? So you have manual CPC on the one end. By the way, somebody is asking what’s going to happen to manual CPC.

Any speculation on that? Before we go to Adrian’s or Pedro’s ask. Yeah,

Peter Oliveira: I don’t think we could speculate on that specifically. Right. We are seeing more and more advertisers seeing success on smart bidding. We obviously can’t speak to the future of the product as a whole. Obviously we don’t work on the product specifically.

On building it. But you’ll probably continue to see more and more advertisers transitioning over to these conversion based strategies because a lot of the skepticism that has been so prevalent in the industry in the past is slowly being lifted as everyone was starting to see some better results with this.

Right? So while I can’t comment on the product, I do think you’ll start seeing more and more advertisers on smart bidding and more and more advertisers. More importantly, you’ll Learning how to get smart bidding to optimize to their goals more effectively. Right. And learning how they can control smart bidding to work for them better.

Frederick Vallaeys: And I think that’s the key point. It’s like CPC manual CPC exists today because people haven’t figured out how to use smart bidding the right way. And they haven’t acknowledged necessarily that it’s not just set it and forget it, but there are ways to optimize it. And so that’s why we’re doing the session today.

We’ll continue to do more of these. But yeah, back to Adrian’s question. Right. So Is there sort of that hybrid where you don’t go all the way back to manual bidding, but you still use ecpc Peter, you want to like shed some light on ECPC? Like, not everyone on the call may be super familiar with it.

Peter Oliveira: Yeah, I think, I mean, ECPC used to be a big, like, compromising point for us at Google. A lot of times, right? You do sometimes want to use like the best of Google’s machine learning. However, you might still want to control your, your bids individually. I think we’ve shifted the narrative a bit, like slightly away from ECPC, not necessarily because the product isn’t working overall, right.

Just because we’re seeing. Much better results on our fully automated bid strategies. Right. So I wouldn’t necessarily recommend ECPC over the other ones, but if, if you’re stuck, you’re doing manual bidding but you still want to add a little bit of, of our smart bidding technology into there, then yeah, that would be a, definitely a good compromise.

In answering your question on, you know, What should you do if a certain ad group isn’t performing? Well, I mean, look at it holistically. Is the campaign overall meeting the targets that we set for it? If it is, and you know, it may come from certain ad groups overperforming and certain ad groups underperforming.

If you’re not okay with that certain ad group underperforming, you might want to, to pause that or break it out into a different strategy. But just know that that could come at the result of like decreased conversions or decreased sales overall. So that’s something you want to play around with and see how, how it affects your bottom line.

Frederick Vallaeys: Right. And again, make sure that you have your attribution model correct, because if it’s on last click attribution and you see one ad group that’s not performing well, well, it could actually be contributing to the success of other ad groups down the line. So if you have these features in place, and then Peter is also making the point that you know, are you okay with subsidizing, having your winner subsidize your losers to have sort of a portfolio strategy?

And that’s okay in many cases, but, but it’s not okay. If you’re taking a thousand different products with different margins and just dumping them into a single campaign, a single ad group in my opinion, then it’s not okay to have the winner subsidize the losers, right? Cause it’s not going to maximize your profit that way.

Sort of the analogy that I. I’m liking quite a bit right now, so I’ll tell you, but it’s it’s basically think of smart bidding as this fancy new washing machine that you just got. And up until today, you’ve been washing your clothes manually and Google comes along, gives you. A washing machine.

And so you’re like, well, I got some walls and I got some cottons and I got some synthetic material. And I watched those manually at different temperatures, right? Because they have different margins, different profitabilities. And they got this new washing machine. Let me throw them all in. And I’m going to set the average temperature from all of my.

Manual washing that I used to do, right? What’s going to happen? While your wool sweater is going to shrink and you’re not going to be happy. And that’s kind of the same thing with structure and targets, right? Don’t set an average target when you have very different things, services, products, different margins right?

These things should inform your structure so that you can have one bucket of smart bidding. That has a goal that actually makes sense for that bucket of things, whatever that bucket happens to be. So that also talks about, you know, if you have an ad group that’s underperforming, but your goals are correct, your attribution model is correct, then think about it more broadly, right?

Is it still helping that whole campaign? If it’s just in the wrong campaign and that’s why it’s performing poorly, pull it out, put it in the right place.

All right. Since we’re talking about optimizing your existing smart bidding, right? One lever that people might still have is quality score. One of you guys want to talk about quality score?

Peter Oliveira: I mean, I’ll add some comments, Fred, but I know you’re, you’re pretty familiar with quality score as well. I think quality score still brings us a lot of like great insights on, on the keyword and the queries that it’s bringing in, right?

I think looking at the relevancy of, of your ads, looking at the quality of your landing page, right? Looking at the expected click through rate. All of those are still like extremely relevant metrics for you to look at. I wouldn’t say it’s like the end all be all right. If your keywords have a lower quality score, but they’re bringing in good results, then, you know, it may be working for you, but it definitely.

Puts up some flags for us that, you know, you shouldn’t, you should investigate and you should look into and Fred, feel free to comment more on that.

Frederick Vallaeys: Yeah. I mean, so again, upleveling here for a second, right? Like the way that the auction works, the ad auction, it’s still taking a cost per click bid. Even though that’s now been informed by your target CPA, your target ROAS, but it sends a CPC max CPC into the auction.

That max CPC is combined with all of the quality score signals, like the probability of getting a click on that ad, which is how Google makes money. It’s not just by having people put high CPCs. It’s by some advertisers actually paying those CPCs. They said they’re willing to pay for. And those two things combine.

Right? So if you’re able to boost your quality score by having a better ad, a better special offer, just more relevance in your campaign, that is a signal that’s going to boost your position. And so then Google’s automated bidding system might say. We actually have to pay less To achieve that same position that we wanted to be in with this ad or we can get you into a higher position And higher positions will get more clicks and more clicks will lead to more conversions because google data has shown That conversion rate tends to be fairly steady across positions on the page Right.

So by moving up on the page, you just get more clicks. They convert at the same rate That’s more conversions. That’s more sales for you. Hopefully more profit if you’ve set the right targets. But yeah, that’s why quality score is still so important. Right. And I think there was actually this flow and quality score people weren’t paying as much attention to it.

But now that much of the bidding is becoming automated and easy. It’s like, we can actually go back and figure out how to make more relevant ads. So I think it’s, it’s still super important.

All right. We have some other good questions. So people are curious how to learn more. You know, obviously Google is not going to disclose the algorithm behind it. I would say that reading search engine land certainly is a good place. And Jim is talking about an article on search engine land that says broad maps, suggestions will soon surface. Do you guys have any advice on Like how do the most sophisticated advertisers get to that level of sophistication? Is it do they read a secret book somewhere?

Peter Oliveira: Yeah, so obviously search nglm can be a tremendous resource for that, right? It’s not like google’s official opinion in a lot of cases, but we have some tremendous Experts in the community that are you know, updating you on the latest and greatest there So that that’s a good point to check.

Other than that i’d recommend like our our training now the skill shop That should be able to like offer you like some, some insights into different things that you wouldn’t necessarily get directly from the support article. Those are recently revamped I think sometime in the past year. So there’s a lot of good stuff there that you know, wasn’t available in our previous trainings.

So look through that. I think the Google ads channel as well on YouTube does post a lot of videos on that as well. So I think that’s a good resource too.

Emi Wayner: That, yeah, that is true. To add we’re thinking about adding more webinars like this. To share more tactical advice with the audience. So this is something that you can leverage as well.

Frederick Vallaeys: And then, so, I mean, the question was really kind of like get into the algorithms and I think another question here from cozyaholic is asking for work around, so how do you still do bid adjustments when you’re on smart bidding? And first of all, I wanted to show, we do have a chart that we put up on the Optmyzr blog.

So this shows the comp compatibility of different bid management types with the different bid adjustments. So one thing a lot of advertisers don’t realize is that even though most of them are not compatible, they just, they simply get ignored when you set a device bit adjustment in conjunction with a target CPA goal, it actually changes the target between those device types.

So that that’s kind of a fascinating Insight that I, for a long time, didn’t know until somebody smart at a, at an SMX conference actually kind of pointed it out and then I did some research and found that that was true. Right. But, but, but the point I guess is. Don’t try to game the algorithm, the smart bidding algorithm.

Think more about how do you give it the right information to make the right decision on your behalf. Right? So don’t assume that it’s broken. And also, and again, as an ex Googler, you take this however you want it, but don’t assume Google to be nefarious and what it’s trying to achieve. Like, listen, I was the AdWords evangelist for a long time and I had to get up on stage and convince people that Google was doing things for the right reason.

You know what? The way that Google makes money. Is by having a lot of advertisers and a lot of advertising revenue And one way to make a lot of money is to screw all your advertisers And you’re going to make money for one or two quarters and then all your advertisers are going to leave your business is done So what google really wants to do is just match goals with what advertisers want and provide a long term sustainable Ecosystem for those results to keep happening, right?

And so that’s really fundamentally What smart bidding is about. It’s about matching your goals to how do Google system works. So I think, and let’s talk about this more, but how do you really help inform Google what you truly care about? Right. Because too many of us, I think, have conversions that we call conversions, but it’s not really what we want to have happen, right?

It might be somebody filling out a lead gen form that we call that a conversion. But what we really want to do is get that person to buy something from us. Just me getting the lead gen form. That’s not enough. So how do we inform Google about what we really want? So

Emi Wayner: that comes to, I guess, Peter, you can jump in, but

Peter Oliveira: it’s helpful to think about yeah, go ahead, Peter.

I think for a second.

Yeah. I think I had a lag. Sorry, everyone. I think it’s, it’s helpful to think about why you’re doing the bit adjustments in the first place, right? Are you adjusting bids? Because you know, you’re trying to get closer to the goal that smart bidding is achieving, right? Are you adjusting bids? Because.

You believe that this place is getting more revenue, like this location is getting more revenue or that this device would lead to more conversions or CPA, or are you doing it because there’s something that smart bidding doesn’t know about, right? Are you trying to adjust because there’s higher margins from a certain place or because there’s a higher customer lifetime value coming from a different device?

If that’s the case, I think we want to do as best of a job as possible of informing smart bidding of those things, right? We can either do that through campaign segmentation, right? Segmentation, segmenting groups that have a higher margin to a separate campaign. And then maybe we’ll set more aggressive we’ll bid more aggressively on that through our ROAS target, or we can set that by modifying our conversion values, right?

So we can use those tools, campaign segmentation. And conversion values to segment certain portions of our traffic. So that’s smart bidding knows which ones it wants to bid on more. If you’re just trying to achieve like higher revenue, higher conversions, we think smart bidding does a pretty good job involving that, but usually.

The need to use the device adjustment come because there’s something that’s marketing doesn’t know. So let’s use these tools to like inform smart bidding of.

Frederick Vallaeys: Right. And so things like offline conversion updates and after the lead form has been filled, informing Google about what happened with those leads afterwards, like did they actually turn into customers and updating the value to be higher and if they didn’t convert, maybe you’re reducing the value.

So Google gets you less of that type of traffic. When it comes to e commerce things about, you know, submitting. Profit or margins as opposed to basket value. After your returns have come back, update the value. So Google knows that yes, you may have sold 200 worth of stuff, but half of it was returned, so it was only worth a hundred dollars.

So all of these things can be done through offline conversion updates and reporting which takes a bit of time. And it’s something Optmyzr’s thinking about helping you with. But that’s really how you help inform the system a bit better. So we’re actually, you know, pretty close to the hour here and we have tons of questions left.

So if we see a bunch of good questions, we’ll try to answer those offline. But you can also email support at Optmyzr. com. Just tell them you were watching this. And to get in touch with me and we’ll do our best to answer those questions. Don’t expect the support team to know exactly what we talked about.

We’re not all watching this right now, but here’s another question that maybe we can have as one of the final ones, but when setting up standard shopping campaigns, would you still segment those or Well, smart bidding, just figure it out.

Peter Oliveira: Yeah. I think we don’t need as much of a granular segmentation as you would, if you didn’t have smart bidding, right?

So if you didn’t have smart bidding, maybe you want to break off like every single category of every single product group that you would like to break up differently and you would have to sort of hyper segment it, we don’t need to do it that way, but for example, like we were talking about, if we do have different margins or if we want to set different ROAS targets for different segments and different products, then that’s when the segmentation would, would come in handy.

Frederick Vallaeys: Yeah, and then Jordan is asking are you guys seeing smart shopping campaigns outperforming traditional from a romance or CPA perspective?

Peter Oliveira: Yeah. So typically with smart shopping, we really use the robust and conversion value metrics and yeah, we’re certainly seeing it outperform traditional shopping.

I don’t want to say that the specific percentage right now, because I think it can be anywhere from like 10 to 30%, but we are seeing outperforming pretty much across the board, be it for smaller advertisers or, or for, for larger advertisers as well, but also at the same time, maybe This is just on average in aggregate, potentially a sophisticated advertiser using traditional shopping on smart bidding could see some stronger results as well.

Right.

Frederick Vallaeys: Right. And keep in mind that those smart shopping campaigns, they bring in remarketing and they kind of bucket that together in terms of the performance. So that’s kind of the unknown, how much of that performance is on new customers that you’re bringing in versus ones that you’ve already kind of captured through other means like organically.

For example and that’s kind of what we see. So we, we see a mix. I mean, smart shopping campaigns are certainly very good if you weren’t managing your regular shopping campaigns very actively then they’re certainly going to be better, but if you have very specific structures in mind where you want to have individual.

Products and in separate groups then standard shopping campaigns still work work very well. So ultimately, I guess the one takeaway we can have after the session today is you know, sometimes these things work better. Sometimes they don’t. It really, it’s always dependent on you as the advertiser and you having set up all the prerequisites.

We haven’t really talked about experiments, but experiments are quite a good way to test it out. Do a 50 50 split test on trying a new bid strategy type and see what happens before you fully commit to it. So and that’s Google technology. So you can just set that up as an experiment and a draft campaign first and then turn it on.

So with that, we’re at the top of the hour. Our Google experts have to go on to their next meetings, but if you wanted to get a copy of the smart bidding guide, we’ll be sending to that early next week. If you haven’t registered, just put your email and your name on that page and we’ll send it out to you.

Thank you so much for watching. We’re going to be back in two weeks with some of our favorite speakers from PPC town hall. In this year. So we’ll have certainly very lively sessions coming up and we’re going to keep them tactical and full of practical advice. So tell your friends, subscribe to the channel, Emmy, Peter, thank you for joining us today.

It was great having you answering lots of questions. Really awesome. So thanks again. And thanks everyone for watching. Take care.

Emi Wayner: Thank you for having us.