Planning for the future of PPC automation

Episode Description

Every day, Google puts more roadblocks in front of PPC marketers.

You need to be ready to jump these hurdles from the most influential entity in search advertising. We speak to two clever and shrewd PPC experts on what the next phase of search marketing and automation could look like.

This panel covers:

• Managing accounts with less access to search terms data
• Finding the geo data that Google removed from its system in September 2020
• Optimizing ads now that Google has de-prioritized Expanded Text Ads
• Other data-limiting changes to be expected from Google in 2021
• How to keep your PPC career alive when Google seems to be automating it all

Episode Takeaways

Managing Accounts with Less Access to Search Terms Data

• Use Ngrams to analyze trends within the accessible search terms to infer patterns in obscured data.
• Shift focus towards broader keyword management strategies and rely on Google’s machine learning to match intent effectively.

Finding the Geo Data Removed by Google

• Although Google has made it harder to directly access user location data, it can still be found under custom reports in the interface.
• Advertisers should use this data to adjust strategies based on actual user locations, particularly for campaigns targeting specific geographical areas.

Optimizing Ads with De-prioritized Expanded Text Ads

• Embrace responsive search ads (RSAs) that allow Google to dynamically adjust ad content to match queries.
• Continue to use ETAs strategically until they are completely phased out, ensuring that messaging aligns with campaign goals.

Data-Limiting Changes Expected from Google in 2021

• Prepare for further restrictions on data visibility, requiring a shift towards more generalized and automated campaign management.
• Focus on setting clear conversion goals and refining input data to improve the performance of automated bidding strategies.

Keeping Your PPC Career Alive Amid Automation

• Evolve from hands-on keyword and bid management to strategic roles that focus on overall campaign performance, optimization beyond the keyword level, and client relationship management.
• Develop expertise in areas less likely to be automated soon, such as conversion rate optimization and client-specific strategy formulation.

Episode Transcript

Frederick Vallaeys: Hello and welcome to another episode of PPC Town Hall. My name is Fred Vallaeys, I’m co founder and CEO at Optmyzr and I’m your host for today’s session. So we’re going to go back to one of our favorite topics on PPC Town Hall. It’s about data and automation. So you know, every morning I wake up and I go on Twitter to see what crazy news the president has put out there.

That’s what I used to do. But nowadays it’s like I go on Twitter and I look for what crazy announcements does Google have? Like how much more can they automate and how much more can it change the way that we as PPC professionals go about business? So every day is a little bit stressful because things are changing very, very fast in PPC.

And so that’s why I thought. It’d be good to have another session and talk about, well, what’s the future, right? When Google keeps doing this, when Google takes away search query data, when Google makes it harder to make expanded text ads, and it’s giving preference to responsive search ads, and when Google is doing more and more automated bidding, and when Google is saying you can do shopping campaigns without even having conversion tracking.

Like what’s the picture on the wall here? Right. What is that saying about our future and what do we do? So I see a lot of accounts in my work at Optmyzr, but I brought in two panelists who also see a lot of accounts that are different. And I really wanted to get their perspective on what we do when Google keeps automating and taking away our data.

So that’s the topic for today. So let’s start rolling.

All right. And welcome to my guests for today’s session. We have a returning. Martin Rutgerding. How’s it going, Martin? Hi. Very nice. Very good. Where are you calling us from today?

Martin Röttgerding: I’m from Germany, still in Germany. It’s six o’clock here. It’s getting dark soon. It’s the end of my day.

Frederick Vallaeys: Yes. Well, thank you for joining us and I hope you have a glass of wine ready to go.

I think Brady and I were both on the coffee side. So, hey, Brady. Are you in California?

Brady Cramm: I am. I’m in California and I’m not drinking wine yet. It’s 9 a. m. So I’m having my cold coffee.

Frederick Vallaeys: Okay. Well, you’re not that stressed out that you have to have wine at 9 a. m.

Brady Cramm: I hope it never gets to that point.

Frederick Vallaeys: Okay.

We’re in California.

Brady Cramm: I’m in Irvine. So Southern California.

Frederick Vallaeys: Nice. I’m in Northern California, Los Altos, a little bit of a rivalry between those two, although nowadays I think we’re all like unified with the fires. It’s like we got this crazy stuff, right?

Brady Cramm: I know it’s crazy.

Frederick Vallaeys: Hey, but Brady, you haven’t joined us on a PPC town hall before.

So tell people a little bit about yourself and what you do.

Brady Cramm: Yeah. So I am a PPC expert, so relevant to this conversation, but I’m the director of PPC at directive consulting. And so we are PPC and SEO for B2B SAS companies. And so I look over the entire team, very involved in strategies and got my roots in medical skincare device B2B sales doing PPC.

Frederick Vallaeys: Interesting. Okay. So it

Brady Cramm: was fascinating, fun place to start.

Frederick Vallaeys: And then it says, so you were doing the PPC for the skincare and then you decided to take your skills to more accounts and go the agency route.

Brady Cramm: Yeah, it was a fun way to start. You know, whenever I interview a PPC, I ask, how did you fall into it?

Right. You didn’t major in it. You didn’t get an internship. So my manager handed me an AdWords account back in the day. Because I brought them from Dreamweaver to Mailchimp. And so she trusted me with some marketing tech and the rest is history. I, I fell in love with the platform and saw that it could drive results.

And so wanting to do it for more companies.

Frederick Vallaeys: Interesting. I I got, I fell into PPC selling VHS video cassettes that I would pick up at Blockbuster after they were used. And I would sell them to people who did not have a Blockbuster in town. And I was young, so I made a lot of money, what I thought was a lot of money at the time, but it was a fun hobby.

Martin Röttgerding: Nice.

Frederick Vallaeys: Martin, how did you fall into it?

Martin Röttgerding: Started as an intern, my first internship 15 years ago, and I stayed at that exact company. Now I’m Co ceo own part of the company and that’s that

Frederick Vallaeys: basically worked away from the bottom and took over

Martin Röttgerding: Yeah, you could say that. I mean we were two people at the time at the beginning roughly two maybe three, sometimes and I stayed others didn’t but many came along and Yeah, now we are like, about 30 people here Well, not really here because everyone’s working from home.

Everyone except me. Yeah, I started as an SEO actually wanted to do the cool stuff and then I had to go the PPC route because we kind of divided the fields between us and Well, my boss was there first and he was the SEO expert, so I had to become the PPC expert and here I am.

Frederick Vallaeys: Well, maybe you thought you got stuck with it in the beginning, but now I’m sure they’re all envious of Yep, you got that part of the business.

Well, good. So yeah, let’s talk about automation And we got a couple of topics lined up here, but anyone who’s watching us live we are live streaming So please do ask questions and weigh in Through the comment section that’s on youtube as well as on facebook if you put those comments in we’ll see him a couple seconds later and if anyone wants to tell us hello from where they’re located, it’s always nice to see Where people are watching from so thanks for watching anyway So why don’t we start with an article that martin wrote fairly recently here about some new google tax And i’m going to share that on screen here.

So Google tax martin tell us more

Martin Röttgerding: the google text. Well, it’s not actually a google tax. Yeah I got some pushback on that because someone said it’s not Google who, who can raise taxes. But the thing is that some countries they want to text Google and they started to add, well, Austria, Austria and Turkey, I think they added 5 percent on what you pay to Google.

Yeah. What, what Google builds you Google has to pay 5 percent in taxes for that. And our Google says no, we don’t want to pay that. We pass that on to advertisers. They also do this in the UK where it’s only 2%. And I guess other countries are to follow because France has this but it’s held up because I think the U S administration administration put some pressure, pressure on France and the same in Spain.

And so they went back on it, but I think it’s coming to other countries as well.

Frederick Vallaeys: Interesting. And so for us based in the United States, like how does that work? Do you actually. Does the CPC on the back end get increased? And so the tax is built in or do you get a charge afterwards in a separate invoice?

Martin Röttgerding: Well, I think if it would be factored in, which would be, I think the right thing to do in theory, then no one would be paying it because in Google ads, you control your bits, you control your budgets. And if you do this and you have like, I don’t know, you have a bit of a dollar and now 5 percent of those of those 100 cents goes to the government and 95 go to Google, nothing would change for you.

You don’t care if Google gets your money or the government gets your money. It’s just gone. So what Google does is they add this to the invoice. So after everything is said and done at the end of the month. You get your invoice and there’s like the five percent charge for those countries or maybe two percent so google doesn’t really

Frederick Vallaeys: Nobody likes paying taxes.

Let’s put it that way. Nobody likes paying taxes. So what are the optimization strategies to? Yeah, I mean, I guess you can’t say the words right now pay taxes, but how do you optimize your taxes?

Martin Röttgerding: Yeah, usually You already know how much you want to bid and you know how much you want to pay. And if you have to pay five percent more than okay, you will dial everything back like five percent So it comes back the same you mean I mean you already have your bids, you know, what’s profitable for you You usually can’t afford to just add five percent or at least you never wanted to so you just go a bit back I mean, your, your bids, they have a reason why, why those

Frederick Vallaeys: are back.

So, so, I mean, are you seeing any automated, and you also said it’s different in different countries, right? So are people using scripts to automatically reduce their bids to, to then not pay too much in ROAS and CPA, for example,

Martin Röttgerding: well, if you don’t really look at it, nothing for you will change in the accounts.

It’s just on the invoice. Right, so you won’t even feel that the text is there because it’s not in your KPIs. It’s not in your cost it’s just on your invoice. You have to pay it, but it’s not there in your KPIs for the

Frederick Vallaeys: campaign That’s the problem, right? So if you have a ROAS targeted campaign and you’re like sweet we exactly hit the ROAS target that we needed but then at the end of the month you get a surprise bill five percent for austria two percent for france or the uk But you’re like, okay, well And all of a sudden you’ve missed your raw ass target.

Martin Röttgerding: Yeah, and maybe you will even well you will miss that this is on the invoice because at least what i’ve seen the invoice usually goes to bookkeeping bookkeeping and Away from the the people who actually manage the accounts. They never see the invoices someone pays them Their bosses pay them. They don’t know that it’s there right and I mean bookkeeping They don’t know what this is about.

They just they know they have to pay this So this can go undetected for a while, and it, you know, Could go on for years that you actually, you reported your, your numbers up the chain. So this is my cost. This is what we’ve spent great results. And then at some point someone learns that, Hey, there’s, there’s this this 5 percent charge on top.

Oh no, we’ve always been paying this. We never knew this. And you’ve been reporting the wrong numbers all of the time.

Frederick Vallaeys: Now, in a way, let me play devil’s advocate here. So isn’t that a little bit the same as the agency tax? So if a client works with an agency and they get billed an extra, say 10% usually the numbers they report are like, here’s your CPCs.

Here’s your role as a chief on Google. Oh, and then by the way, here’s our invoice for managing it.

Martin Röttgerding: Sure. You could say that, but I think that this is not a hidden cost. Right. I mean, there are different ways to build this. We could have a fixed fee. We could have a variable fee where it’s 10 percent or something like that.

Okay. And some agencies report this too, and others don’t. But in any case, this is not, not something that’s hidden from anyone. Whereas google is trying to Do this in a way that you can’t react to it because you don’t know it

Frederick Vallaeys: Right. So let’s talk about that bit a piece that piece a bit more. So the session topic today is google hiding more data So I think this relates to what you’re saying about the taxes and that google has removed some geo reports and made it a bit More difficult to find them.

So how are those two things connected?

Martin Röttgerding: Well, taxes depend on where you are or where, not where you are, but where the, the people are who see your advertising and where they click, where the cost occurs. So if you’re in the U S and you advertise worldwide, I don’t know, you may be advertising in the UK. Then the people who click the ads in the UK there’s some CPC involved there and 2 percent are added to that CPC.

Okay. Right. So it matters where someone is who sees your ad. And you can find out usually by going through the location reports and just look at the user location. And user location may be different from the targeting location. Because you may target some place but if you target a place that can mean that you will get people who are, for example, regularly on in the U S who have expressed some search intent that tells Google they are In some way interested in the U S or in, in some other location in the U S for example, but their physical location is still in the UK and the physical location is what matters and what, what what will make this more expensive for you.

And it used to have a report where you can just look at user location and it used to be more or less transparent that there is a difference between targeted location and. And now Google has removed those reports. There are still ways to get the data, but it’s not right there front and front and center.

It’s no very easy way. You can

Frederick Vallaeys: talk us through a little bit how you can still find that data, but it’s also we had the episode last week with David Sattel and one of his tips was make sure that you look at your geo Targeting and you are aware of the difference between physical location and location of interest.

Cause that does make a big difference. But then it was funny because he makes that, he gives that advice, but then on the back end, it’s now become so much harder to actually measure when Google is doing one versus the other. So Brady walk us through how you guys as an agency could still see that.

Brady Cramm: Yeah, it’s interesting just combining the taxes to this. This article is using an example around running shoes in New York, but being in the U S where that tax isn’t happening yet, I think where the industry, I always like to think what industry is in the most trouble. And the first one that came to mind was New York pizza, right?

That’s a, that’s a style of pizza where anyone around the world could just be wanting to see what it is. Right. And in these taxed countries, now these local shops actually in New York, they’re capturing people around the world, looking up New York pizza and the location settings directly in the unifi Yeah, user face.

They just don’t show that that’s happening s O. I can see a lot of New York style pizza places in the U. S. Now getting U. K. taxed and they have no idea why. But to find that information like we mentioned, it’s not in the locations directly in the user interface anymore. So you have to go to reports.

in the top right corner of the interface. And then go to custom and build your own report. And if you go in the search bar of that report and you actually type in user location, you will then have options to put in your rows that will show the user’s actual location. So it does still exist. Any advertiser can still see, Hey, Am I a New York local pizza shop in the US getting traffic from the UK?

Because in the interface, if you just go into locations, it will say New York as, I believe it’s called a matched location. Right. And so if someone in the UK, if they looked up New York style pizza, they would still come in in the interface as New York because it was a matched location. And so you can easily think none of your traffic is coming from the UK when in reality it is.

Frederick Vallaeys: Good. So that data is still there. It’s just become a little bit harder to find it. Do you guys think this is a first step towards kind of like getting rid of that data, kind of like to do with some other things?

Martin Röttgerding: I wouldn’t think so. Because, well, I think Google is just making it harder. They are just trying to delay people.

From being able to, to work with this. I think it’s kind of strange that they do this. I think it’s a mistake to do it this way. Because I mean, it’s so, so obvious. I mean, you just had the screenshot with the old interface where you could, with the old reports selection, where there were predefined reports and now they are just gone and there’s no value to removing.

These predefined reports,

Frederick Vallaeys: right? It’s not saving any space, right? It’s yeah, and there’s no like privacy concerns. So it’s it’s a little weird. Why? Why you’re doing this? Yeah, but I guess the good news is the data continues to live in the API. It’s still in the report. So for those of us who want it, it’s there.

I mean, and I think this is the broader theme as well is that Google seems to be trying to make things much easier for the novice advertiser

Martin Röttgerding: and

Frederick Vallaeys: The average advertiser, but then us who have agencies or just do this day in, day out, some of the things that we’ve built best practices around are being either moved to the background or worse, completely taken away.

And so that kind of leads us, I think, into the next topic, but search query data. Right. So we’ve talked about this before for the listeners, but Google in September said we’re going to remove a bunch of search query data from the reports that you see. And we’re doing this for privacy reasons. I’m still trying to puzzle together exactly how that privacy component works.

And I am an ex Googler. So I do believe that when they say these things, they probably mean these things. And there’s probably some like weird corner case where somebody can construct personal data from a whole bunch of queries. But the pain that we’re seeing as advertisers is we have so many best practices around search query management and looking at your search terms reports.

And so what you see up on the screen here, this was Matt Umbrough, he was looking at what percentage of his search queries became basically invisible. And it went from 4 percent for one account in August to 37%. Right, so one third of his traffic all of a sudden. Became invisible and that’s based on clicks.

That’s not impressions. That’s clicks that he actually pays for and he sees that he did five accounts All of them are you know as high as 37 Brady, what are you guys seeing?

Brady Cramm: Yeah, you know, I recently saw a screenshot where one conversion out of 23 Was reported on in the search terms. So very extreme data.

But yeah, we’re seeing the same thing. We’re seeing this being a struggle for low volume accounts. So those accounts where you can actually spend every day in the search term reports and you can read the search terms and you can make your moves based off of that. A lot of those accounts, I believe, are struggling with it.

I think when it comes to high volume, it makes something like Ngrams even more relevant, right? If we don’t have a lot of data there, we have to find the trends within it. And we have to collect the data for searches using the word free, assuming a lot of the searches we aren’t seeing them are probably using it too and making those decisions.

And having some visibility to what we’re currently blind to. So I do think it makes sense. Things like Ngrams more relevant, but when it comes to low volume accounts I do see it as a major struggle and really tough to have a solution outside of being patient and keyword stuffing exact match, which I don’t think is the direction we want to go in.

Frederick Vallaeys: Yeah, exactly. Because exact matches and even exact match. But that’s really interesting that you say that Ngrams is sort of the way forward, because I was thinking about that too. And it’s like Ngrams, the whole purpose is to see deep into the long tail and what are those common terms that might appear very low volume queries, but in aggregate have a big impact.

And it’s, it’s oftentimes that really long tail traffic that’s becoming invisible to us, right? But you’re saying that. I guess based on statistics, if you see a pattern in the head of all the searches, that pattern probably exists in detail as well.

Brady Cramm: Yeah, if you see a pattern within the searches you have access to you can assume that still shows up in the searches you’re blind to and make decisions that are impacting based off of that data that we don’t have.

Frederick Vallaeys: Right. And so then when you use ngrams, is it now more about putting in negatives or do you still use that to then also say, Hey, well, we should probably add a few keywords. We don’t know exactly what the query might have been, but at least these words should have been in it.

Brady Cramm: Yeah, I think that’s where maybe a three gram could be relevant, right?

If you have that data and you can drop in a phrase match or a broad mod off of it. But I think mostly for negatives and to my route that I don’t think is ideal earlier, maybe you could start collecting exact match, but I think that’s where the actual search term comes in handy and not necessarily an Ngram doing exact matches off of what you see in there.

Frederick Vallaeys: And I want to talk a bit more about some other tools you use to do like keywords given the lack of search terms But martin, I want to go to you first. What are you seeing in terms of the search terms data changing?

Martin Röttgerding: Well in the post you showed before I argued that this is kind of a diversion. Don’t go I don’t want to go full conspiracy here but I think this is the timing is perfect because Because everyone is talking about this and no one’s talking about the text thing Which I understand actually I haven’t done that many big analysis in our accounts because everyone else has.

And I think it’s kind of the thing that we won’t change. It’s like, it is what it is. Google won’t go back on this because this is done officially for privacy reasons. And now there are petitions to. Get Google to roll this back, but they don’t go they won’t go back on privacy, I think. And I think what we should better push for would be some alternatives.

And what would be really great would be to have Ngram data from Google, right? Because yes, Google doesn’t want to give us this specific data, an impression here, this is a search term. Okay. That’s okay. But if Google could give us, give us Ngrams aggregated data. Where privacy is not an issue anymore, that would have helped very much.

Frederick Vallaeys: Right. So do the data obfuscation or the privacy obfuscation for us. Interesting. Have you talked to anyone at Google about that?

Martin Röttgerding: No, because, you know, of course we have agency reps, but They have different problems. You know, sometimes you talk to them about this and they have never even heard about that problem.

They don’t know about search query data. I mean, they don’t know it internally until we are in other agencies start to complain, but for them, this is, this is not that big an issue.

Frederick Vallaeys: Right. And let’s talk about that disconnect for a second. Right. So when you work with a Google rep or really anyone who’s newer in the PPC space and who maybe didn’t grow up with some of the best practices that we’ve gone through and who comes into a world where machine learning and AI is like the buzzword of the decade, I just say, well, yeah, I mean, sure, machine learning can figure that out better.

Like, why would a human manage search terms? I mean, our kind of argument would be, listen, we still know something about the businesses that we work with and we can take that nuance and we can use that to basically unlevel the playing field and take some advantage of the system and win a couple more auctions and have a slightly better, Lower cost per acquisition, slightly better return on ad spend.

And that is why we want this data. That’s why we want these levers, right? But how do you make that argument to someone who maybe hasn’t been there and, and believes the AI story so much? And how do you make this argument to someone who has no way to change anything, who won’t even talk to their bosses about this because no one cares.

Martin Röttgerding: And that’s the disconnect that we see in our reality. Right. I get that. Right. I mean, that’s disconnected, but at the same time, like, listen, we’re out here educating and the more that we can see, the problem is sometimes a change gets made. And advertisers rise up and revolt. And they say, I mean, this is a horrible change and Google is evil.

Frederick Vallaeys: And Google’s trying to like screw with the advertisers. Most likely not what Google’s trying to do. It’s just that there’s a lack of information there. Right? So if you can go to them and say, well, here’s actually what we did and here’s the actual impact on our business. Right. And so, and, and that’s what makes it a little tough too, because what I would love to be able to go and do is say, listen, Google, if I had had the search the search terms, here’s all the new keywords I would have added.

Right. And I would have maybe bit more for these keywords. Yeah.

Martin Röttgerding: Maybe excluded some of them,

Frederick Vallaeys: obviously people would exclude them as well. Right. And Google is going to look at the big picture then, right? And the problem again becomes they look at averages. So does the average advertiser do better? And that’s kind of their metric because they can bring more novices on board. And if that means it comes at the cost of the experts having a little bit less of a good result probably a trade off they’re willing to make.

Martin Röttgerding: I think in this case, we have several things that are in Google’s interest. One is privacy and they can’t. Go back on that. And privacy is always a reason for them. The other one is they can prevent us from excluding things, which is something they want, or they don’t want us to exclude things. And also they don’t want us to spend too much time managing this stuff.

They don’t have anything against us doing it, except that this makes the whole system more, more expensive because someone has to pay the people who do this. And I think there’s this mindset, maybe a Silicon Valley thing where I’d say Google thinks they know better. And I think they have this, this trust in the technology and that they can, can make it better.

And yes, they look at averages and on average, many of these things may be true and we look at our, our own cases and we don’t believe it. Even though some of us are certainly wrong about this. I don’t know. In this case, in other cases, Google has changed a lot over the years, which we didn’t like. In some cases I think it was okay.

And probably a good thing, even though people revolted.

Frederick Vallaeys: And maybe it becomes a political discussion now. Right. But it’s like, is Google trying to democratize the access to successful pay PPC? We can all acknowledge, I think that PPC has become really difficult. I mean, just the number of buttons and dials that you have in the Google ads account nowadays, like if I was looking at it for the first time, I’d be completely overwhelmed, right?

So right. So, but is that kind of the socialist approach where it comes at the cost of you know, maybe the people at the high end who were really successful at Setting all these dials the right way to get the most success. And now, you know, there’s kind of a penalty on them to help support the novices.

Martin Röttgerding: I don’t think that’s the intention.

Brady Cramm: No, I, I don’t either. I think, you know, making it easy to opt into the display network when building a search campaign, I don’t think having the search network or the display network hard to opt into was, you know, a challenge for novice people, right? So moves like that, you know, everything we’re talking about, I’m so curious, What the financial model meeting is behind these decisions, if that does exist right now, I’m kind of being a conspiracy theorist myself.

But I always think too, what is the revenue model of making opting into display network the automatic setting when building a search campaign and having a click away to remove it. Even search for reports. I’ve always been fascinated by the fact that it’s buttons away. Versus keywords being right there in front of you.

Frederick Vallaeys: Right. Well, and it’s also interesting when you compare and contrast Facebook and Google ads. So Facebook came in when machine learning was much further along. So they were able to build a system where it was sort of assumed that a lot of the decisions would be made by the system. Whereas in Google ads, we grew up in a world where we had to set CPCs manually.

We didn’t have conversion tracking, like a thing like CPA just didn’t exist because we couldn’t measure it. And then it started migrating towards where we are today. And at every stage, it’s like a little bit less control for the sake of, you know, letting the machine help you maybe find new things. I was actually there when the search network was in sorry, display network was opted in by default.

That’s widely acknowledged as a mistake by Google and one day undid. Hey, we were young. We didn’t know any better. But, but, but so I mean, it’s interesting too, right? Because I think at the end of the day, a lot of advertisers, why do they go to agencies? They go to agencies because they want more conversions.

They want lower cost per acquisition. And I, I think Google is making an argument that the display network. There’s certainly potential there, right? I mean, we can’t say that there’s no conversions coming off of that. It’s just like, what’s the cost of those conversions? And if the machine can help you find those conversions, then great.

But I think what it. Requires is that advertisers almost have to shift to saying like, listen, everything’s automated because when we’re talking about managing search terms, but we’re still manually managing bids you know, that’s why we need to manage search terms. Because if we have manual bids and there’s some search term that comes in, that’s kind of, kind of related to what we sell, but not exactly.

While we want to be able to bid lower for that. Right. I think what Google is saying, listen, give us, you Responsive search ads do automated bidding and let us figure out the targeting and that combination of things If you’ve told us exactly what a cost per acquisition is worth the computer can probably do a pretty good job But you have to kind of like be all into all of the automation and if you use it piecemeal That’s when things start falling apart a bit.

Brady Cramm: I’m not fully against it. I mean, we, we’ve run a campaign across our portfolio and once again, it’s, it’s mostly B2B software but it’s full broad keyword targeting paired with target CPA and it does fairly well. And looking into the search terms, we see some of them are non branded solution based terms.

Some of them are. Comparing our client versus a competitor. And some of them are comparing a competitor versus another competitor. But in the end of the day, the cost per conversion and even cost per MQL it seems pretty good. So it’s one of those things where, you know, even myself, I was, Really against it.

Right. A broad keyword. I’m never going to have that in my account and pair with target CPA. And you still have to look into search terms and add negatives, but, you know, I am seeing Google machine learning, you know, work fairly well when giving it, Hey, here’s the freedom on the keyword and search term level.

But here’s the rule that you have to abide to, which is the target CPA. And I think the biggest mistake. And I know we’re going to talk about more of the black box strategies and smart display is just what you’re telling Google is success and in B2B telling Google just a form fill itself is what we’re looking for that can get very dangerous when dealing with smart display campaigns.

Frederick Vallaeys: Yeah. And talking more about the MQL. So the marketing qualified lead. So basically you’re saying somebody fills out the lead form and then there’s some process that happens with the sales team, I guess, to qualify those, or is that even automated too?

Brady Cramm: That’s dependent on client, but I think to my point of like feeding Google the right information, measuring it on the backend is one thing, but we strive for offline conversion tracking with all of our clients.

And so once again, just telling Google, Hey, a form fill is what I want. That’s just telling Google, whether it’s a human or a robot, as long as these fields are filled out and a button is clicked, it’s what we want. And you know, that’s not really a good business KPI to, to go after versus e com the credit card work, they went to the thank you page, you got the money and you can be fairly confident that that’s success for your business.

Frederick Vallaeys: Yeah. So would it be fair to say that as a, as a team, as an agency, you’re spending more time on helping people implement conversion tracking offline conversion tracking, then managing some of the details, or that at least there’s a shift towards doing more of that?

Brady Cramm: It’s definitely the first step, right?

We can take over accounts that, you know, we look into the conversion tracking and it’s forms across the entire site are going to one. Contact us and it’s tough to make decisions. It’s tough to opt into machine learning It’s tough to manually optimize based off that information And so it does start with the conversion tracking but from there, you know There’s a lot of strategies we can put in place once we get that information whether it’s with smart bidding or without it

Frederick Vallaeys: Martin, what about you guys?

Do you see any shifts in the agency as far as the type of work you do? Given the automation.

Martin Röttgerding: Well, we do mostly e commerce. So What you just described is true for us. If you have the money, if you have the credit card and everything, that’s already very good. But we are shifting towards going through going towards you know, tracking margins instead of revenue.

And then we want to move on to customer lifetime value. And I think the third thing to go after is incrementality, which is nothing that you can easily track, but is an issue that you have to look at. And. Margins and customer lifetime value. I think those are things where you can work with the machines to give them right data, and this is going to be an increasing part of our jobs.

So I think it’s kind of removed from, from what Google does. We just give them something. It kind of feels like Google hasn’t gone there yet. So we go there and we do it until Google gets there and can do it themselves. I don’t know if that will ever happen. But this is certainly something that we can still improve on and which will be more important in the future.

Frederick Vallaeys: Right. It almost feels like, you know, when you think about the Google ecosystem and Google ads and what we used to manage was like at the core of that system, which was the exact targeting, the exact bidding, but now we’re getting more and more to the edges and like the Google system manages itself, but what we manage is inputs into it as far as what is a marketing qualified lead what kind of attribution model do we need to put in place so that we correctly value You know the whole funnel, and brady you were talking about that too a bit right like how we’re going from just Being the ppc experts putting in keywords and ad texts to maybe having to think a little bit more broadly about campaigns and traditional marketing campaigns

Brady Cramm: Yeah, it’s interesting.

Even what Martin brought up, you know, when it comes to B2B software marketing, we’re really looking at doing it on an LTV CAC model. And so modeling that out, not just for Google ads, but for every channel we’re dealing with and really helping the client on that level. And so, yeah, it’s less like, Hey.

You’re paying us this retainer because we know the keyword match types, right? And we know all these different alpha beta structures. It really is, you know, we’re bringing ourselves up the ladder in terms of the decision making process and the modeling process of these programs that we’re running.

And, you know, Google ads is one of many online advertising programs that our clients are using. So we do find ourselves having more conversations like that. And then the post click. You know, when it comes to landing page optimization and A and B testing, what offers are out there keeping a pulse on the competition?

What are they telling users? You know, we have more time to look there as well. And to me, that’s exciting, right? We get to spend our time looking at the competition, checking out their page, reading it word for word. What is their call to action? Putting it up against ours, reading ours against the scene.

You know, is this the decision I would make? Because that’s where we’re seeing a lot of our results come from especially as we’re losing control of the levers. And, you know, I tell my team, I think the white space of Google is gone, right? There was a heyday where you could find that keyword. You could find that time of day and it makes you a lot of money.

And so you put those settings in place. And I think that white space is gone. I think you have to fight at the top of Google and accept that cost per click and the optimization beyond that and the modeling. Has to make sense.

Frederick Vallaeys: Exactly. And that’s why it’s so important to have a great conversion converting landing page because we can debate like being, well, and it’s not even 5 cents cost per click more, but in many cases it’s like an extra 10 cost per click if you don’t optimize, but you could actually derive more value from just doing some conversion rate optimization.

No, just smart and curious. You know, maybe let’s talk about the whole black box thing. Right. So I think Brady’s making the point that you go full funnel, you do more competitor research. Like how do you see one company being more competitive when everyone’s using the same black box automation?

Martin Röttgerding: Yeah. I think a bit big part of this comes down to what you put into the machine. Like in terms of conversion data, in terms of signals not just, you know, measuring the conversion post click, but also how can we give it more signals, micro conversions, whatever to, to just give it more and more to work with.

The other thing is in terms of, you know, product data, for example or, or, or maybe in terms of if we talk about DSAs dynamic search, as well as about your website that Google has to understand. So if we can give Google better input there then Google, you know, Might be able to get us better results.

I think that’s kind of not really satisfying if you are used to being able to control everything, everything. Now you can just have someone else do everything and you just have to, you know, work for them. But this is certainly one thing that you can do. The other thing is not everyone is using the same like boxes.

I mean, there are still differences. And what Brady said that that traditional marketing, like you know, looking, what’s your, your KPI, what’s, what are your USPs? What, what’s the messaging, for example, what’s the right messaging. This is still something that you as a human has, has have to do. Yes, there are responsive search ads that mix the messages up.

But this is only as good as what someone put in there. And I think it’s a mistake to think that the machines can do everything by just mixing things and, you know, throw a lot of things against the wall and see what, what sticks, I think is the impression you know, and just see what works. I think it’s.

If there’s a human who can, can relate to the audience and who know, know what their, their worries might be and what they really, what the selling points might be they will always do better than the machines alone.

Frederick Vallaeys: Right. Now you’re a big fan of. Scripts and automation, right? Ad customizers and scripts.

So what, what does that mean?

Martin Röttgerding: Are really one of my things, my favorite things about Google ads and yeah, I fear that things like these are going away because right now we can also set ourselves apart by using ad customizers, for example, and some other stuff. And if Google automates everything, I think what, what I, I’d like to have would be a world where Google.

Gives people a lot of automation so that they can, that everyone can use Google ads and use it reasonably well, but the experts, the people who, who are willing to, to do a little more that they can still do a little more. I hope this is the world we are going to, but I’m not really sure because ad customizers aren’t compatible with RSAs. So I’m not sure it’s an advanced feature. But it looks like it’s, I don’t know. Is it going away? I don’t know. It hasn’t been updated in a while.

Frederick Vallaeys: Yeah. But then I guess a lot of stuff hasn’t been updated in a while, partly because the API issues they were having. So they were transitioning to a new API that was supposed to happen in late 2018.

They launched it. It was broken. They went back and spent more than a year fixing it. It just now came out. So yeah, like scripts has fallen really far behind in terms of development, because I think everything shifted to fixing the new API and now that, because that’s kind of the core of the system.

Right. And now that that’s fixed, hopefully we start to see new capabilities.

Martin Röttgerding: But about that we’ve, for some reason, we have been using the old API which is the AdWords API, still not the Google ads API from 2018 is the last update and I just realized that I’ve, there’s nothing I miss. I mean, Google has evolved, but there’s nothing I miss because.

Google, the involvement they have done is usually around stuff like discovery campaigns or local campaigns. So all these black boxes where we can’t do much anyway. And all the things I like to automate is still about shopping about search campaigns and everything’s there in the old API. But not

Frederick Vallaeys: everything, right?

I mean, so that’s sometimes the mistake too, is that people say everything’s there and then Google’s like, Oh, great. We’ve done our work, but even like competitor data and auction insights, I don’t believe that’s in the API and that would be tremendously useful.

Martin Röttgerding: No, no. Okay. But that’s not, it’s never been in the API and I think it’s not in the new API.

Maybe I haven’t.

Frederick Vallaeys: No, no. Yeah, you’re right. It’s not in any of the APIs, but like if we, I mean, our customers, right. They look at why is my cost rising. And if we can explain, well, it’s not because of the agency, it’s because the competitor came in while, okay. That justifies a lot. So now. The agency doesn’t have to fight for their, to keep that client and they can actually focus on strategy.

So I think there is a lot of value that could be gained from Google, exposing some of these things and making them more absolutely assuming, but okay. I get your point too, that you’re saying all of these new things are basically black box. And so when, when it’s all black box, are you saying you’ll just set them up into Google ads interface and then there’s nothing to manage.

So who cares? We care. Yeah.

Martin Röttgerding: Yeah. I mean, this is why I haven’t missed any functionality from the API, I think. I mean, there is just nothing new where I thought that this is something that I’d like to manage. Maybe I’m just out of date. Could be true, too.

Frederick Vallaeys: No, I don’t think any of us are out of date. Hey, so Brady, you had talked about alpha beta structure and Alex is asking about quality score improvement.

I think those two sometimes go hand in hand, but let’s talk about quality score and like campaign structures for a second. And, and even something like single keyword ad groups or skags, is that still something you do? Do you still worry about quality score? And how can you. Maybe automate improving it.

Brady Cramm: Yeah. I think it’s definitely something we still focus on. We never want to make a decision strictly based off of it though. We, some of our lowest quality scores are our most profitable keywords. Right. And so we still want to have that mindset. But you know, where my mind goes with it, talking about whether it be structure or, you know, dynamic search ads responsive search ads, you know, I guess not to be a conspiracy theorist again, but, you know, you know, The more you give Google the decision, the more they’d be hypocritical to say it’s a low ad score in, in my opinion.

And so that’s the way I see responsive search ads. You know, you give it 30 headlines, you give it multiple descriptions. Google’s kind of hypocritical. If they think I expect a low click through rate when I’m the one forming the ad. And so I do think automation does go hand in hand. With quality score. But I do also think the SKAG structure allows you to control it.

I just think it’s going away. You know, I think the SKAG structure with expanded text ads, when. Let’s say exact match was not close variance. You know, it made perfect sense. You knew what the searches were going to be. You could control the headlines, but we’re just moving further and further away from that to where us being able to control it on that level is going away, but I think there’s still a solution and I do think that is accepting the automation that is really taking your time on responsive search ads, knowing your queries are going to be unique.

And you have an ad that can form around them and be exactly what the user wants. I think the gap right now is the page routing. Right? We can, we have very broad, unique queries. We have ads that can dynamically form to the query, all going to a page. And that’s where, you know, I think we’re going to talk about it maybe later, but that’s where my thoughts around DSAs.

Being the future is more on the page routing level. We’re still kind of constrained there. The more broad we get with targeting and more responsive we can get with ads still going to one place. And I think that’s where it breaks. And I think Google’s moving in that direction,

Frederick Vallaeys: right? And again, maybe it’s the point I was making earlier that if you really want success with automation, you’ve got to automate everything.

And if there’s even one piece that’s not in line, like you’re saying with the page routing, then it breaks apart at that last stage. And then you know. Low conversion rate, even though the bit might’ve been right. And the keyword might’ve been right. But now we complain about the performance, right? And that’s the tricky thing because at the end of the day, like Google might be listening and it’s like, Oh yeah, there’s another thing we should automate and give them less control over what we do, we’ll complain more.

Right. But I think it’s fundamentally again about like, how do we optimize at the edges at the boundaries of the Google ad system? And so that’s landing page optimization that might be for shopping campaigns. How do you optimize your feed, right? Because you don’t have an ad necessarily, but you still have the feed, which has the title, which has the description, which then goes into the ad.

What about price optimization? People are super price sensitive. So if you can have a product that’s 5 cents less than the competitor, as opposed to bidding 5 cents differently, that might be your better give you better results,

Martin Röttgerding: which is really a field of its own.

Frederick Vallaeys: Exactly. And that’s, what’s so fascinating, right?

So how do we, having worked inside of Google ads for so long, how do we shift our roles to be on the boundaries of it? Because I think there’s still a lot of optimization that’s going to be possible, right? I’m still happy with the company name that I, that I picked Optmyzr, right?

Martin Röttgerding: We’ll work forever.

Frederick Vallaeys: Exactly. We’re always going to need optimization. Just might be a little bit different. So Brady, you wanted to talk more about DSAs, I think, or you said we would cover it later. So I want to make sure you cause we’re getting closer towards the end here. I want to make sure that anything that you wanted to cover, we had talked about.

Brady Cramm: Yeah, the only I just we were talking about, you know, the future of Google and what to expect in 2021. And I think we are backing our way into DSAs. And I know I mentioned to you both when we think about the future. With any topic, we think net new, we think, okay, the next iPhone is going to have 10 cameras, right?

It’s going to be something new, but When it comes to the future of Google, I do think there’s obviously going to be new features involved, but it seems like we’re backing into something that’s already existed, which is DSAs, which is Google having control of the search terms, the ads and the pages in a sense, right?

We can choose it on a page level but it just seems like that’s where we’re going. I think the future of Google. Already exists. I don’t think it’s going to be as extreme as Google Express or AdWords Express. But it definitely seems like we’re moving into something that’s already there.

Frederick Vallaeys: But let me bring up one thing we haven’t really covered today at audiences, right?

So I think there’s a lot of restricting of data, but audiences is that one thing that’s actually new and more data is coming out about it. And more capabilities. And now we have GA4 tying more closely into the new ad system. But, and I said, I think especially in B2B, like audiences are so important because you distinguish between that person searching for a printer for their house versus for a big office.

So how do you guys think about audiences in that whole mix?

Brady Cramm: Yeah, they are crucial. You know, when you’re dealing with B2B cloud storage, it gets a little tricky. Right. Versus B2B. The millions of people trying to back up their phones. So audiences are definitely necessary. And even talking about the black boxes and how do you compete?

I think building your own audiences and what are your efforts before Google to create brand awareness to educate people on the problem that your solution solves, how are you building those audiences and having campaigns in Google specifically for them? Because yeah, if you’re, let’s say the third ad for that primary query, You’re the company that they learned the problem from.

You’re the company that they already trust and it disrupts any algorithm that the competition is running off of to position themselves for that same person. So I think the prebuilt audiences themselves and market audiences, if any audiences are very valuable and they can definitely weed out the wrong people.

You know, unfortunately what we’ve found is the B2B specific audiences around company size, anything like that, that I believe might still be in beta. The. Quality is just not there yet. It’s not like Microsoft’s partnership with LinkedIn. We, we measure it one on one and the data is just not accurate.

But I do think focusing on the audiences you’re building yourself, your first party audiences, is a crucial part of the strategy to compete when we don’t really have all those historical levers at hand.

Frederick Vallaeys: Right, and maybe the part of that point is that Google uses machine learning to make all these decisions for us.

And you know, the word learning is in there so that we can teach it ahead of time and give it a really good starting point that prevents a lot of the waste upfront where it would just be throwing stuff at the wall. And like Martin said, see what sticks, right. But a lot of that stuff wouldn’t stick.

And that’s how it learns. If we can actually say, well, here’s an audience that we’ve already. identified as being really high value, very likely prospects that learning becomes more confined. And there’s less wasted spend.

Brady Cramm: Yeah, a hundred percent. I still don’t think, you know, a feature like similar audiences, it’s just not there yet for B2B and the type of people and the situations we’re trying to get our advertisement in front of.

It seems more built for the consumer market. Right. And that’s interesting because you’re making that point and you’re in the United States. We, we often hear that point from smaller markets where there’s just fewer search queries. In that specific language. So if you’re complaining about that in English, then it’s probably a big problem in Polish, for example.

Yeah. I mean, we’ll find ourselves where the total addressable market is 30, 000 people in the U S like that’s it. It’s something like a similar audience, which is going to build a list of a million people. At least it just isn’t relevant to what we’re trying to achieve.

Frederick Vallaeys: And by definition, you know, it’s it’s incorrect. Martin, what about you? Any topics we haven’t covered that you think are important here?

Martin Röttgerding: Probably. I will think of them later. I think about audience, I’d like to say to add that in B2C, audiences are kind of strange. I’d say because they are on one hand, they are important to have, and you should add them to your campaigns.

But on the other hand, they’re just used. If you, if you’re using target rowers or target CPA, they’re just used by Google to, to refine your bids. And you don’t really know what’s happening there. You get the data, or you get only data for the audiences that you’ve added. So yeah, you’re better off if you add a lot of them.

So we get a lot of data, but besides that. I don’t know what they’re good for because if I don’t change my messaging for the audience then yeah, I mean that’s the, the black box of Google’s bid optimization and that’s about it. And what I know if is if Google needs those in the in the ad group, for example.

So do I need to add a thousand different audiences for Google to be able to look at the audiences and refine their bids or don’t I.

Frederick Vallaeys: That’s a great question, right? Because you sort of assumed that Google already has a lot of audience signals. And so what’s important to them is to start to connect what is a conversion and how does that connect to those audience signals? So by us giving it different audiences, it’s just like different groupings and I suppose the different value we can assign to them. As a way to tweak maybe the target cost per acquisition, the target ROAS. But in terms of literally saying like, here’s an audience that is different from this audience.

I think Google’s machine learning probably has got that one figured out long before we do

Martin Röttgerding: probably. So I don’t really know what it’s. Good foreign b2c. I mean, I know that you can do things with it But mostly it’s just there and it’s kind of like hidden away keywords are important audiences Are also there

Frederick Vallaeys: but at the very least I mean, I would say do it for observation purposes

Martin Röttgerding: Yeah,

Frederick Vallaeys: right.

So just put them in. And then if you start to see that, Hey, actually, there is a difference in conversion rate and row ass between this audience and that audience done. Yeah, but I didn’t change your targets, right? And that’s where it’s always so fascinating. We go from these really detailed structures like single keyword ad groups and alpha beta structures.

And then Google says, roll it back up. Make it more generic. And then, and then we started looking at audiences and then we’re like, Oh, actually we do see differences and we do want to have different attacks. So we do need to have different ad groups and we do need to have different. And so the structure now builds out again.

Right. So it’s, it’s this endless cycle of,

Martin Röttgerding: yeah, those things are really conflicting and I’m not really sure where Google wants us to go. I think ultimately Google wants to do everything they can, they want to give, get us to give them a blank check here, do whatever you can, and, and we’re happy with the results.

Results may be good.

Frederick Vallaeys: Eric Schmidt actually did say that at a internal meeting. Way, way back when he was like really the vision of advertising was that someone should be able to write a blank check and say, what is your business goal? But I think where it always breaks down is, you know, as an agency, you talk to your client and it’s like, what’s your goal?

So many clients cannot formulate what their actual goal is, or they state conflicting goals, or they don’t know how to measure that goal. And until you have the day when that’s solved, like Google just cannot take your blank check and do a good job for you because they’re going to do what they thought was the good job.

And then you look at it and you’re like, well, actually, yeah, actually you told us wrong what to do. Right. And that’s the problem with machine learning until you tell it what the actual goal is and the actual value it’s hands are bound.

Martin Röttgerding: So this is something that we as agencies can still do.

Frederick Vallaeys: I think we should.

Right. I mean, that’s the value that we bring is helping the client. Understand the system that we’re playing with and what that system requires to successfully achieve the outcome and defining what the outcome is that you want. Is it profits? Is it revenue maximization? Is it new customer acquisition and fine if it’s at a loss?

Right. But you have to be very clear about that before the machine can do its thing. And honestly, I mean, even if it’s not the machine doing its thing, like if it was us humans managing it, We still need to know that too. I mean, I was just talking to an agency last week and they were like, well, we’re not real happy with the results that we’re seeing in this one campaign.

I was like, okay, well, let me take a look at it. And right. And yeah, and maybe the conversion tracking was broken for a few days last week, but we’re not, we’re not a hundred percent sure. And it’s like, well, what do you do then? Right. You can’t figure anything out if you don’t know the data’s correct

Martin Röttgerding: on

Frederick Vallaeys: that.

Happy note. I mean, are you guys optimistic or pessimistic about the future here, given all the automation?

Martin Röttgerding: Torn really because I like some things to be able to do them manually or to automate them myself. And that going away is something that I don’t want, but that may be just a personal passion of mine. If the results are good, who am I to argue with that?

Brady Cramm: Yeah, I think for me, if I were more of a PPC technician that lived in the spreadsheets and did the data exports all day, I’d be a bit confused, but my passion is in advertisement.

And so, you know, I get happy sometimes when I think less time in the platform and more time thinking about the call to action, the offer, you know, how do you create excitement for the free trial versus just having a free trial button, the strategies like that, where we still see incremental growth from implementing it.

That’s where it makes me happy. And so I’m not too concerned. I just don’t want to waste my client’s money. That’s where. The fear comes from is losing the freedom, losing the control, losing the visibility. Are we wasting money? Are there a bunch of irrelevant search terms that would never become customers, clients in the future?

That’s my biggest concern is not knowing that and having a strong belief that it might be happening. Well, those are great points. Thank you both for having been on this week and thanks everyone for listening. We’ll be back. Next week or the week after we haven’t quite decided yet, but we’re going to talk about marketing in a VUCA world, which means volatile, uncertain and then I forget the other two, but it’s, it’s basically, there’s a new research report that’s out and it’s going to tell us how we as marketers can thrive in this very uncertain world.

Frederick Vallaeys: And I think part of that is the fact that automation is taking over more. There’s a lot happening in the world. So thank you both for putting in your viewpoints and for being on the show and we’ll see you next week.

Brady Cramm: Thank you.