Marketing in a VUCA World

Episode Description

Today’s world is more volatile, uncertain, complex, and ambiguous than ever.

Search marketing is not immune to this VUCA syndrome, so Anders and Lukas dissect some of the findings from their latest research report. With interviews from experts and a survey of award-winning digital marketers, it features actionable tips to find a way forward.

This panel covers:

• How to market in a volatile, uncertain, complex, ambiguous world
• The rise of AI and the race for data density
• Advanced attribution to look beyond brand conversions
• How the 2020 health crisis has influenced digital marketing
• How AI-based bidding is coping with external factors
• Preparing for the end of cookies
• Paths to firm ground in shaky times

Episode Takeaways

Digital Marketing in a VUCA World

• The VUCA (Volatile, Unstable, Complex, Ambiguous) concept is increasingly relevant due to ongoing global instabilities like COVID-19, affecting digital strategies.
• Agencies and marketers must adapt to continuous changes, leveraging automation to handle the dynamic market effectively.

Impact of Automation on Agencies

• The shift towards automation is prominent, with many agencies embracing machine learning for bidding and reducing dependency on data analysts.
• Automation helps manage the workload but requires strategic oversight to ensure that marketing goals align with business objectives.

The Role of Data in Decision-Making

• Access to detailed data is diminishing, making it challenging to perform in-depth analysis.
• Marketers need to focus on understanding the inputs and outputs of automated systems to optimize strategies effectively.

Preparing for the End of Cookies

• The anticipated removal of third-party cookies will significantly impact tracking and ad personalization.
• Companies should prepare by developing direct relationships with customers and collecting first-party data to remain resilient in a post-cookie environment.

Episode Transcript

Frederick Vallaeys: Hello and welcome to another episode of PPC Town Hall. My name is Fred Vallaeys. I’m your host. I’m also the co founder of Optmyzr. And last week we had a great session. There was a lot of talk about automation and how Google is changing things. Really well attended session. And we thought we’d continue in that vein and talk to two people who’ve just come up with a new report that’s talking about digital marketing in a VUCA world.

And VUCA stands for Volatile Unstable, complex and ambiguous, and I think we can all certainly agree that 2020 has been exactly that kind of VUCA year. So so what does that mean? So we’re going to talk about some of the Unstableness and the constant change that’s happening, but we’re also going to talk about automation.

And this is a trend that we’ve long been seeing where the engines are just moving more and more towards automating the tasks that we’ve done. The authors of the report talked to a lot of agencies, a lot of experts. So what they have to share, I think is going to be really amazing. And as a special, thank you for watching the session with us live today.

We’re going to have a couple of we, we will give you that report for free for the first 20 people who go to the website with a code that we’ll be sharing with you. So stay tuned. We’ll have that code towards the end, but let’s go talk about digital marketing in the VUCA world.

Look at others. Welcome. Thanks for joining us on the show today.

Lukas Adamec: Thank you.

Frederick Vallaeys: Thanks Fred. Where are you guys calling us from today? I’m in California. Yeah.

Lukas Adamec: I’ll start. I’m in London, which has been quite a sunny day so far, but it’s literally just changed to quite a storm. So I’m hoping there’s not going to be any background noises, but it’s been quite interesting in the last 10 minutes.

Okay. Well,

Frederick Vallaeys: welcome from London. Good to hear that you finally don’t have rain for a change. In other words, we had you on a couple of weeks ago, right? Are you still in Paris?

Andres Hjorth: Still in Paris. And and we’re expecting announcements from the president in a couple of hours about, you know, how VUCA the coming weeks are going to be.

It’s probably locked down again. So yeah. Tough times these days. Yeah. So I’m calling in from from Paris. I’m, I’m Danish but I’ve been in France for more years than I lived in Denmark. So quite the, quite European

Frederick Vallaeys: Well, good. Yeah. I speak a little bit of French because I’m Belgian. It’s one of the languages we’re required to learn originally, so, but but good. Yeah. So VUCA world. So tell us a little bit about the report before we jump in. Why did you decide to write this report?

Andres Hjorth: The sort of the, the, the report idea started out two and a half, three years ago, I sold my, my last agency back then I it was bought up by Accenture, but you know, the, what was left of it was bought up by, by Accenture after having been acquired by another company.

And it was my time to move out of the agency business and, and. Do something else. And what I really wanted to do was to find and share digital marketing insights. As a as a as an agency leader and agency founder, I was always looking out. I was going to conferences. I was listening to you, Fred. I was looking for insight.

I was looking for digital marketing inside. What’s gonna happen next year? What works? It’s this year. What about all these things that the engines are coming out with? with, you know, all the media owners are suggesting all the technology owners are suggesting what, how do you, how do you sort through that and find what is good and what is bad?

So that was the original idea. I was always also judging search awards, other digital marketing awards. And I saw so much good stuff that I wanted to share. And basically, you know, if you’ve ever been to an award ceremony, it’s like, well, and the winner is, and you get a name and it was I feel that frustrating.

I want to share stuff.

Frederick Vallaeys: Yeah, exactly. How did they win? Right? Sorry. How did they win? What’s the secret of how do I get there?

Andres Hjorth: Absolutely. I see. So recently I see, you know voices saying, oh you know, you just have to pay and then you’ll win an award. You know, as a judge, I totally don’t agree with that.

So I’m taking fights there. But anyway, I wanted to say, I

Frederick Vallaeys: think it’s a little bit, it is a weird space, right? Because the award ceremonies, they make money by selling tickets to the award show. So I think most people, and actually correct me on this, like does everyone who nominate themselves, if they’re like, if they have a pulse, do you get shortlisted or is there even like sort of a cutoff at that first level?

Andres Hjorth: So, so there is a cutoff, but it depends how many entries there are. So, yeah. So many categories. So it is not really difficult to be nominated somewhere. So what I really thought nominations are self nominated, right? So so there’s a you know Like if you look at categories like best ppc campaign for example, there will be you know I don’t know how many entries but typically maybe 25 30 50 entries and there’s only like a short list of Maximum 10.

So, so they will be the best of that, those 50. And then the winner will really be, you know, the best in the, within those 10, but in, in other ca categories, that can be less entries. So there’ll be less competition to be shortlisted. So I wasn’t, we weren’t looking at shortlisted, we were looking at award win.

So what we do in the research is, yeah,

Frederick Vallaeys: I, I’m, I’m

Andres Hjorth: happy to hear it. Nice.

Frederick Vallaeys: Be one one right there.

Andres Hjorth: Yeah, I’m not, I don’t, I, so I don’t even remember if I, if I judged that one. I’m not sure. I probably read the entry. Do we catch it? It’s just like this straight away. Do we want to screen? Why this name VUCA?

Frederick Vallaeys: Oh, yeah. Okay. Let me put the okay. So great question from the audience. So why VUCA, right? So let’s talk about VUCA. So

Andres Hjorth: Lost her. But so the story of it is look into insights, find experts, survey award winning teams. So that, you know, the best way to find changes to say, okay, who won and what, and then ask them what’s going on, what are the trends?

So why VUCA? So when we started out looking at this report, actually we started out, Looking at PPC trends in the market, right, Lucas? That was the original idea. And and this ended up being, you know, not just PPC, but digital marketing overall. And in the VUCA world, because the major, major trend that we were seeing was that everything was, you know, chaotic.

Volatile. Uncertain, complex, ambiguous. When I first saw this, these four words there, I was like, well, this really describes the world that we’re living in. And we were interested in, you know, what does it do to the data? What does it do to this already, you know, constantly changing sector of digital marketing for that for going in that direction?

Frederick Vallaeys: And so when you put together that report, you talked to a number of people, right? So and Lucas actually tell us a little bit about the role that you played. We haven’t heard much from you yet.

Lukas Adamec: Absolutely. So Anders actually contacted me about the report and he contacted me initially about the first sort of version of this report of a similar report from 2018 and The one reason where it really resonated with me was the specifics of knowledge sharing because I you know, I work as a, as a freelancer.

I work with a lot of agencies here in the UK. I also have my own agency that I provide expertise for clients. And one of the main things that you always look for the likes of benchmarks trends, you know, anything that you can basically take away and more easily explain some of the craziness happening online every now and again.

And so when Anders contacted me and, you know, said that He would like to cooperate on this and collaborate on it. And you know, actually be able to share this information with a lot of people. I was on board pretty much right away from that one call that we had. And since then it, it grew and you know, we focused on our own kind of key strengths when it comes to, for example, Me looking at you know, handling the data of the of the responses and kind of helping out with that more.

Because that’s kind of where some of my technical background is as well. And, you know, basically syncing this together with frequent calls with Anders that we had making sure that we synced our ideas together as well, and our views on the data that is there, because as you can imagine you can extrapolate data in almost any way.

You know, you would like it to look when we had the raw data, we really wanted to make sure that, you know, what we were looking at, we were describing it as impartially as possible. So that’s where a lot of back and forth was happening even up until the last few days, actually. So you know, it’s been, it’s been a very cool ride writing this report.

Frederick Vallaeys: Nice. So let’s talk about the report a bit. So I think one of the first points about VUCA and the world being so volatile and what happened in 2020 and what’s still happening in 2020, But big impact on digital agencies, right? So what are some of the numbers that you saw?

Andres Hjorth: Yeah, so I think the first big one, so you see it on the left side there, we asked the 20 agencies that participated in the survey, we asked them, how bad were you hit?

So you got about I want to say 15%. That was no, no change plus minus 10%. You got the huge chunk 50 percent saying between 10, 10 to 40 percent reduction media spend, and you’ve got a fairly large chunk there saying more than that. So. 40 to 60 percent reduction in media spend. We also had other options there.

We removed them because there was nobody responding. So, so we’re looking at agencies suddenly having 40 percent of their media spend disappear. And as we know, a lot of the economic models are tied in some way into the spend, not 100 percent but there are very few are on a retainer basis or on consultancy basis.

So this means they lose a lot of money. They lose a lot of activity. And then the other thing is that the rest of the activity is, you know, had to be changed. You had to do something like everybody’s in panic. Yeah. What do you do?

Frederick Vallaeys: And Lucas, what did you see from working with so many agencies? Did they actually just like stop working with those clients because it didn’t make financial sense?

Or did they put in the. Place plans to sort of transition to, to almost bridge with the expectation that things would be okay in a couple of months.

Lukas Adamec: So in certain cases, that’s actually been the case that you know, there are some of the agency client relationship has sort of ended because of that. Now bearing in mind that the vast majority of agencies, particularly the big ones in the UK, and they do have very different clients in their portfolio.

So that’s where, you know, the impact was basically mostly in the sort 40 to 60 percent range. But yes, I, I definitely know of instances where clients basically have stopped their relationships with the agencies because they were the most hit, particularly around tourism, you know, hotels and some retail as well.

Although as we will see later on as well, you know, retail was. One of the areas particularly for consumers that also thrived in this in this time.

Frederick Vallaeys: In other words, what did you see?

So, so what did you see anything different to have any, let’s talk about maybe the third wave, right? So I think some companies that we saw were trying to bridge the gap to when things would get better, but now we’re in that third wave. And like you’re saying, you’re going to find out any minute whether France has additional lockdowns, which might be, and I hear restaurants and bars largely are having to close more again. So are we seeing anything different or is this just kind of the same thing again?

Andres Hjorth: The big movements is that, you know, so the agencies, what they saw straight away was you know, a number of advertisers would just cut spend. What happens in marketing as soon as there’s uncertainty is that you reduce risk.

As soon as you’re into performance, you’ll say, Okay, I don’t know how this gonna perform. I’m gonna I’m gonna reduce risk. I’m going to do less. So everybody was doing less. And then the ones that were maintaining. So there’s very few. There are examples of growth, but they’re very they’re very niche related.

So what you see is that A lot of people resettling. Resettling means you got to rethink everything. So in the case of one of the things that we saw that we asked the question about smart, smart bidding overall. What did you do when when suddenly user behavior totally changed? Did you? continue as it was before.

And, and you had like, so I actually thought it was fairly low a fifth of people or a fifth of projects sort of on an average would have a totally changed bidding strategy on that basis. I would have expected. more like 75 actually. But this all depends on the nature of the business that you’re running and whether you’re impacted or not.

It’s so, it’s so varied the way the way the lockdown would impact people. Of course, you know, anything related with travel would just shut down. And I

Frederick Vallaeys: think one of the big parts of the report is about the shift towards automation, and automation obviously is very helpful in these uncertain times.

But one of the points that you made, and let’s talk about this, but if you put the wrong data in, the wrong prediction will come out. And I think part of the concern and what you guys were trying to investigate was has there been so much volatility in the current situation that the machine learning model is actually helping us?

Breaking and is not able to deal with this. What’s happening right now. And by the way, I don’t want to say the word too much because YouTube doesn’t like it when we do. And then the video gets, so let’s not say the word, but we all know, right. So, so but yeah, what did, what did you guys see in terms of the volatility caused by 2020 and how did the machine learning models respond to that?

Andres Hjorth: So, so you know, just to, to, to, to reframe this a little bit when we talk about VUCA, so of course there’s COVID and there’s lockdown but there’s also, you know, a couple of other things like terrorism like the end of cookies. And, and, And there are and sort of and more and more ai and and also less data sometimes so so so it’s actually a number of things that we’re seeing we’re putting all these together under under the under the vuca, Into the vuca framework the direct impact is of course easily more easily measurable on the covid because it’s such a such an abrupt abrupt change and you know And it was less visible than we than we actually expected.

We see people adopt automated or machine learning based bidding massively. huge increase in, in, in adoption. We saw them. So we, we got, we got feedback from from agencies saying we use data analysts and in comparably less cases than we did before. Which is sort of that’s

Frederick Vallaeys: right because so what’s going on are they figuring the machine is just taking care of everything and you no longer need a data analyst or are you think that’s what this

Andres Hjorth: is one of the places where we spend a lot of hours actually discussing

Lukas Adamec: it actually felt like kind of the outcome is sort of It’s almost you know two sides to a coin where on one hand there is less of an involvement of data analysts because there is more reliance on AI, but on the other side as well it’s not about the position of the data analyst.

It’s more that. That insight into performance is actually being put on the marketer’s shoulders So where previously you would have a dedicated analyst that would be able to support you with some of these. You know Analyses and basically looking at different trends looking at performance, with different impacts you now have this almost becoming a standard part of a marketer’s job instead of focusing on the you know, main platform changes and trends and you know creatives

Frederick Vallaeys: And so what i’m showing here on the screen is the I guess the answer to the question.

Do you have a dedicated? Analyst function attached to your projects We also see the distribution on the left, right?

Lukas Adamec: Yes. So in in 2018, that pie chart was Basically the other way around where you know over 60 percent of respondents said that yes, they had a dedicated data analyst on their team attached to their projects.

You Obviously, this could also mean that, you know, the nature of the projects has changed a little bit. Maybe some of the you know, agencies don’t really have such high profile projects as much anymore. You know, could be that some of that got watered down a little bit, if you will. But it still suggests that there is quite a notable shift to how specifically data within paid search is being managed.

Andres Hjorth: And there was a, there was an interesting quote from from Aaron Aaron Levy from Tenuti saying, you know, everybody wants data science, but nobody wants to pay for it, which is, which sort of ties back to, you know, that question of maybe advertiser agency relationship, which is not always easy.

So when do you pay for what? And yeah, there you have the, the quote. So they look at the data they can see, but with tunnel vision And I think we’ve got another another approach on this one saying, well there are what we call black boxes, or, you know, with the input, as you mentioned earlier, the input from one side, and then the prediction that comes out on the other side, and we don’t really know what happens in between.

But should we not be curious about it? And, and we should, and we should have data analysts. So yeah, it’s a, it’s a, it’s a, it’s a, it’s a question we couldn’t Thanks. Immediately say, okay, we’re going to we’re going to look at the data and say in 2018 there were more data analysts or make more people saying we have a data dedicated analyst and in 2020 less people saying we have a dedicated analyst.

We weren’t happy saying, well, that means there’s less data analysis because we think. It’s the opposite. So it’s probably a changing, changing roles within agencies or changing roles between advertisers, agencies and maybe consultancies. At least we know that there’s a lot of things happening.

Frederick Vallaeys: And I want to bring up this slide here, right? So this is later in the report, but basically I think you’re making the point that Digital marketing is becoming much more of a black box, at least within the engines, right? As the Google, for example, automates bidding, automates ads through responsive search ads.

We know a little bit less about what’s happening. And so maybe the assumption is, well, if we don’t see the data, we don’t get the metrics. We don’t get the optimization levers the way that we did before. Maybe there’s not so much of a point for having a data analyst to do that work for us. But then I suppose the point you’re making here and the kind of point we make an Optmyzr to is that.

Yes, the role is shifting and you’re not managing within that tiny bubble of Google ads, but you’re managing at the periphery of the black box of the machine learning system. And you know, things happen in the black box, but if you don’t like what’s happening in the black box, we want to make it a tiny bit better.

You still have control over the inputs to that black box, right? And, and that also then goes to your point of bad data in bad results out.

Andres Hjorth: Absolutely. And, and so an interesting take on this was, so we’ve been discussing a lot on on social media about this, you know, as a warm up for the, for the recent report and I got one feedback saying, Hey you know, it’s we, we tried, we try out the the smart bidding or the, the, the automated solution and and if it doesn’t work, we’re going to try another one.

And it’s like, you know, it used to be we test hypotheses, and now it’s we test algorithms. And then if one doesn’t work, we’ll try another one, and we’ll try another one, and we’ll try another one. So, so the testing has moved level, right? It used to be we test an idea, now it’s we test a tool, we test, test an algorithm, which is an interesting insight.

I thought that was really interesting to see that type of, that type of approach, yeah.

Frederick Vallaeys: Yeah. And so when I wrote about it, it’s like it’s the PPC role is becoming that of the PPC doctor. And basically think about prescribing the right medicine to solve your problem. And even if you said, well, you should do automated bidding.

Well, there’s like nine types of automated bidding just from Google alone. Right? So it’s not flipping the switch. It’s about, okay, test this one. And how is this one better than that one? And really understanding and picking your automation, first of all, and then within the scope of the automation that you’ve picked, how do you inform it about what success means to your business?

What is a conversion? You also have some questions about lifetime value, right? And So what is the right way to help these automated systems drive better results?

Lukas Adamec: I, I actually kind of wanted to slightly come back to your point about you know, being a paid search doctor in a way, because the, the one challenge that you have with the black boxes within AI and, you know, trusting the inputs and outputs is that if you do go to a doctor, you have to tell him what your symptoms are so that you can actually analyze what is going on.

And this is one of the biggest challenges. With using automation as it is at the moment and as it’s becoming a bit more opaque Is that you cannot actually see some of those symptoms. You cannot really analyze them the same way So your input may not have changed at all, but suddenly your output may be changing Because of something you cannot actually fully identify You know, it starts by Removing some search terms.

I mean we have known on social media where people are sharing their accounts You know, the impact is up to 50 of their search terms in terms of their cost Is basically, you know dark they cannot they cannot identify where that’s going and you know, it goes to average positions, as well, you know That’s been a changed a little while ago now But people are still kind of double checking and wondering where their Ads are actually still appearing right now.

Because if you don’t have much competition on Certainly still paying for them. It just,

Frederick Vallaeys: it’s a mystery. What

Lukas Adamec: exactly, exactly. And, and, you know, when you don’t have much competition on some of your results, position one could be above the results. It could be below the results and there’s very different traffic volume for, for those two instances.

So you can find yourself in a position where suddenly you don’t really know why suddenly. All your traffic has disappeared or why suddenly, you know, you have to pay more without really knowing, you know, or haven’t really changed anything on the input side of things.

Frederick Vallaeys: So Julien you seem French by that last name, but it’s a good question right there.

So is the problem not the fact that we have automation, but who’s pushing the automation and that the incentives might be misaligned? I have an opinion on that, so I’ll definitely answer it after. Yeah, I think there might be three opinions on this one. You guys start. Lucas? Yeah, I think we will definitely have a very similar opinion Anders and I as we found from, you know, working on the report.

Lukas Adamec: We are very much on the same page. I think for me, it’s basically just trying to understand, you know, You know, some of the logic behind what’s going on. I don’t think that any particular platform has any reason you know, necessarily to basically just do something that wouldn’t necessarily be amazing.

For marketers. But there’s also, you know, some some thoughts behind that. That’s not the case. I think for me basically the, the whole idea of paid search in history and up until now has been that everything is so transparent. You can track everything down to a single penny you spend and, you know, you will be able to show your results.

This is now basically changing the fundamentals of what paid search would be for me, because I cannot explain everything.

Andres Hjorth: So very similar. We we’re moving away from raw data. I mean, it’s always been like this with the platform. So if we look at not just, you know, the search engines, but look at Facebook, Amazon, whatever. They are platforms where basically no, if you have a website, you own the website, you pay for the hosting, you are the owner of the data.

That sort of runs through that platform. If you have a Facebook page, you are not the owner of the page of the data of anything. You, you are, you sign a contract with Mark. Or you could go on Amazon and well, basically all the dates. So they’re very, Generous in Amazon. They give you a lot of data.

Facebook is very very good designers. They give you beautiful graphs back, right? That’s the inside you get. But the raw data, you don’t get it. And it’s like, we’ve forgotten this because we’ve just gotten used to it. It’s like, well, okay, on Facebook, you can’t track, you know, And why is that? Well, it’s because it’s Facebook.

And okay, yes, on my website, I can on Facebook. I can on Amazon. I can get a lot of data, but not necessarily something that corresponds to what I get in other places. And it’s like, so that’s one side of this. We don’t get access to the raw data anymore. We get the data they want to give us. So that’s one thing.

The other thing when going back to search something quite interesting. I think this was again, this was Aaron again. He said, search isn’t growing anymore. He said, Well, yeah, of course, you look at the number of searches, whatever the nature is changing. There’s more time. There’s more other stuff. So the volume of searches that people do is not growing anymore.

So that’s one thing on the other side. And you can tell me more about this, Fred. But I don’t think the CPCs are increasing anymore. I think, you know, we’ve really reached some kind of seats, some kind of ceiling or saturation level where if the CPC, the cost per click goes any higher, nobody’s going to go there because they just can’t find any profitability in, in, in doing ads on a CPC basis.

Right. And I

Frederick Vallaeys: think that’s why, and you guys pointed out in the report, but quality score is the original machine learning system. And that’s why Google has it. Because. In an ideal world, every single search that happens leads to the most expensive and profitable click. And yes, you’re right. You’re going to be capped at some level on how high that CPC can go, right?

So what Google is ultimately trying to do is like, okay, who’s that consumer who’s actually willing to convert and who’s going to click? And that gets combined with the price. And so Google doesn’t really think about CPCs. And that’s the weird thing. Like the analysts, the business analysts are always like, where are CPCs going?

It’s like Google doesn’t care where the CPCs are going. They care about CPM. I mean, because that’s the one thing they control. They can have like, okay, if we have a better search engine that we can get more searches, then we’ll make more money because we have a good CPM CPC. Nobody cares because we only get paid if somebody clicks.

And if we have, I mean, we could have million dollars CPCs and nobody clicks on an ad and nobody makes any money. Right.

Andres Hjorth: Right. And reach that and to reach that goal. I mean, if, if if, if the, the Optmyzr. Oh, that’s, you know, the data marketer is is focused on the CPC and the keyword and trying to optimize that.

But what the engine wants to do is the CPM. They have this incentive of not giving you all the raw data and then, you know, stuffing in whatever. CPM no more volume into, in, into, into your coverage. Right? And, and so there’s sort of a, there’s sort of a move, a shift from, well, it’s not volume anymore.

It’s not CPC anymore. It’s sort of the density of what you can get out of that data. And, and so that’s my interpretation as well. The, the less we can let people look into this, the better we can monetize it.

Frederick Vallaeys: I think, in my opinion, it’s also to do with the fact that. You have very, you have many, many advertisers, right?

And very few of them are true experts like the three of us might be into people watching this call. So a lot of people are going to make really bad decisions if you give them too much data. So average position case in point, I mean, everybody was like, so focused on position one quality score.

Everybody’s focused on having a super high quality score and then they forget that. Yeah. Maybe your quality score for keyword. Is actually one of your most profitable. And so I think that’s what Google is fighting against. And you made the point that Facebook is much more about pretty graphs, but Facebook also came into play when machine learning was at a far much further along and they were able to do some things where they didn’t have to ask the advertiser for every single little detail.

When AdWords, like there was no conversion tracking. There was no quality score. There was, I mean, so many things that we take for granted. There was no search terms report. I mean, we’re complaining now about losing some of our search terms. When I started doing advertising, there were zero. We did not know we bought a keyword, we paid a CPC for that.

And that was it. We do nothing. We just kind of looked at the bottom line of the business and did we have more money in the bank at the end of the month or not? And that was the gauge of success, right? And so I think in a way, Google ads and AdWords put themselves in that position where it’s very hard to now start reducing the data, even if they think that that would actually be the better way.

To get more advertisers to be successful, but granted super frustrating to experts like ourselves, because that’s the data that we’ve used to optimize. And I think the point you’re making is data is the new oil in the world, right? Nothing is more valuable than a, it used to be oil. Now, now it’s all about data.

And so you need to have that raw data rather than the process data with the results already, like fed to you. We we actually ended up coming up with our own quote. We, so we used a lot of quotes. Every time we, there’s something we want to, we want to sort of express. We use the quotes from all the experts, but we had to put it in our own quote at one point, which was, which is can, can you monetize, can you, can you extract more value?

Andres Hjorth: from your data. And if you give it to a platform, it’s sort of this, this, this question of who controls it and should you keep it or should you give it away to whatever Google, Facebook, Amazon. I think that illustrates sort of the importance is one of the, it’s one of the sort of the, the orientations that we think are are really important.

To start controlling your own data, just start protecting your own data. It doesn’t mean that you shouldn’t, you know make it flow. Blowing data between platforms is extremely important to get better insights. But you need to consider every time you do that and consider what beta am I using?

What am I giving away? Is there anything that I could keep on on my end instead of letting somebody else monetize it? I think that’s quite, quite an interesting point and quite an important one that we’re recommending at the in the last piloting report.

Frederick Vallaeys: And then there’s always the speculation like Jitin is mentioning here, so Google said reducing the data was due to privacy concerns.

Yeah. But was that really the case or are they trying to to hide something? Listen, I’m ex Google Maybe I’ve drank the Kool Aid, but I do believe that when they say it’s for privacy reasons, there’s probably at least some level of truth to that. And when you think about it really hard, you could actually start to figure out cases where you could construct sort of an interesting picture of a user that gets into personal details, like what diseases they might have, what their financial situation might be.

It’s not easy to do, but it can be done. And I think that’s where Google is coming from. What I wish though, is that Google would actually respond by giving us, maybe like Facebook does give us more of the process data, right? If you’re not going to give us the raw insight, then at least help us figure out the search term commonalities where the privacy.

Concerns have been addressed, but you can tell us, well, okay. We kind of think that maybe saw an example that logo design and trademarking are close variants of one another logo design and trademarking very different, right? Like that’s the type of thing we’d want to know so that we can make it a negative keyword.

Andres Hjorth: So Fred to that, I think this is, I think this is really interesting and a really interesting question. I, and I came to this you know, I think this was in relation With Amazon again, I, there was this there were figures out on the market saying that 42 percent of view reviews on Amazon or fake right.

Okay. Wild, crazy. Yeah. So, so we have these big engines and they’re capable. And you said it in the beginning, they’re capable of finding people who have an intent. Yeah, so they’re capable of finding people who are ready to buy ready to whatever, but it seems that they’re not as clever at finding out people who have a bad intent to people who, who, who leave fake reviews or people who who use the data in the wrong way.

It seems that they are reluctant. Of course, it’s not profitable, is it? But I do think they have some kind of responsibility with the data to, you know. To weigh those two sides equally. One is okay for commercial purposes and for protective purposes, but is the right thing then to cut off the data or is it to analyze the data like they do on the other side?

I’d go into the second direction and say, well, you know, guys, you are the cleverest engines in the world. Apply your knowledge to the bad stuff as well as the good stuff.

Frederick Vallaeys: Yeah.

Andres Hjorth: Apply the, apply what you know how to do to, to your own processes as well.

Frederick Vallaeys: All right. Good honors. I know you wanted to share some slides and let’s maybe get some takeaways.

Like, what is it based on the report and what you’ve seen that you think agencies and PPC advertisers can do? To put themselves in a great position to thrive in a VUCA world.

Andres Hjorth: Yeah, I I wanted to give sort of a, a, a couple of a couple of views of of what’s inside. And so you know, what we have in the center, there is it’s the coronavirus, isn’t it?

It’s sort of without scary stuff here. A couple of extracts from the report we asked, we asked questions that were sometimes for fun. There was a question in the beginning. What is volatile? It’s, or what is, what is VUCA? It’s, it’s not just COVID. It’s lockdown. Yes, of course. It’s also face masks.

Yeah. It’s like, well, you know, human behavior is changing. It’s like you don’t book a ticket in the same timeframe as you did before when you’re going somewhere because you don’t know if you’re allowed to. It is also protest movements. It’s terrorism. It’s economic recession. We’re going to see this coming.

You know, I think, yeah, I think it’s, it’s, it’s showing in various places. All of this makes things. More volatile. So all of these things influence what we’re doing as as digital marketers. A little bit of look in digital COVID 19 end of cookies and AI. Sort of in a summary of the three big things that we’re looking at and then how to hold to build firm ground on below you.

So we saw this already. I think you showed it just before that the hit digital marketing agencies or the ones that we surveyed took a during the sanitary crisis or the lockdown. And then this big question. How do you think I will be performing your current job in the future? 90 percent 90 percent of them of the survey of the respondents said no, we don’t think so.

That’s really interesting,

Frederick Vallaeys: right? What’s going on there? I mean, people don’t think automation is going to do the work they do today, or they think

Andres Hjorth: I saw what I Yeah, no, I think I think they’re quite pragmatic about well, the machines have been there all the time, and they’re coming all the time. And we’re still we don’t, we still haven’t lost our jobs.

On the contrary, we’ve actually got better jobs. I think that’s the that’s the reason why they respond in this way. But I also think that they’re very aware of the fact that, you know, the scope as we talk about https: otter. ai is moving into the scope of work of the human, right? So, so the role is different like like you mentioned in, in your book.

So quite an interesting insight. I wanted to sort of quantify this as well and, and and give it some some depth. So this is one for you Fred The smart bidding the end of bit tools. Yeah you know, I I was also Well, well, let me

Frederick Vallaeys: answer

Andres Hjorth: that

Frederick Vallaeys: I’ll answer that. Yeah, tell me so, I mean, I think it’s very simple, right?

So people have this misconception And they think automated bidding is set it and forget it by no means The only thing that automated bidding does is it takes your bid Goal, which is usually ROAS or CPA, which by the way, they’re not very good goals. Like your actual goal should be profit, but whatever.

Google gives you TROAS and TCPA. Okay. So now you’ve got these two goals and all Google does in its automation is calculate the predicted conversion or the predicted click rates, and they calculate the CPC from that. It’s something that we used to do on spreadsheets. What they’re not looking at is how do things outside of you know, what, what’s happening inside Google.

So how does the outside world, how do my promotions, how do those things change what’s happening and what the prediction should be. So to me, it’s kind of crazy when I see advertisers come in and they set a target CPA and then they walk away and two months later they come back and they check it out again.

No, like your targets should still evolve just as much as they used to in the past. Yeah. But it’s you know, it’s just Google does part of it. You still need to manipulate the the inputs. And I think it’s very much to the point that you made too. Right. You got a black box. That’s doing some calculation.

It doesn’t mean you don’t have control over inputs. It’s just that the inputs have shifted out to a little bit further from where they used to be. And thanks for the comment, Arjit, and you you kind of made me laugh, but once a Googler, always a Googler. That’s also the point, by the way, why Optmyzr has never been a bid management company.

We’ve never positioned ourselves as a bid management company, because I fundamentally believe that Google is a Is the only one in the world that can do the auction time bidding. And the, the only ones that have as much data are the only ones with machine learning capabilities that they do. And for any other company to try to fight that is just crazy.

You gotta, you gotta figure out how to take the best of what Google does, but then still control it. Right. How do you control your inputs and how do you measure the outputs to make sure that the machine is not doing something that’s not great for your business?

Lukas Adamec: Personally, I find that there’s a huge level of sort of complacency when it comes to automated bidding, especially in a lot of agencies.

Where you will actually find that, you know, 90 percent of your time, you can almost set the AI bidding and let it do its job and it will still kind of do a good performance for you. But it’s, it’s those times when something goes wrong or, you know, when you have a promotion that suddenly skews the data, whether it’s a promotion before you started applying automated bidding or a promotion within suddenly as soon as you have an impulse that.

Basically, the tool cannot really account for directly. That’s where everything kind of falls apart sometimes. So this is where I even feel like sometimes the way that some of the platforms sell automated bidding, where they sell you know, where they say things like, Oh, the tool is learning for two weeks.

So you shouldn’t really touch it for two weeks. I understand there is some sort of analysis. Time, you know, that time period for, for this to happen, but you should never really be, especially within paid search in a position that you cannot really change.

And

Frederick Vallaeys: let me jump on that. Right. So we’re getting very tactical here. The reason that Google says don’t touch stuff is at some level, not because they’re scared that you’re going to make smart tweaks based on what’s happening in your business, but it’s because when you change your targets, you might become eligible for fewer.

or more auctions. And now the search terms mix, the query mix is changing. And as the query mix changes, those might be new searches that Google has never really seen before. And so they might drag down the performance. And then the fear from Google is that this is hard to explain to a user that You’ve changed your target stuff is moving.

And like, there’s some portion that Google has already learned about. They know the, the, the queries that are staying the same, they’re going to continue to perform at the same level they were before, no matter how much you change your target, but it’s all that new stuff, right? And now the user sitting there looking at their overall reports and they see, Oh my God, my CPA is way up, freak out.

And they like pull back on the automated bidding. When in fact it was like one really small portion of the automated bidding that was. Now learning again. And so that that’s what’s so difficult from Google’s perspective is how do you communicate that? And how do you get the average advertiser to make a reasonable decision based on what’s happening?

Lukas Adamec: But it’s it’s exactly kind of this level of you know Not knowing that does feed the black box as well and not just from you know, the optimization point of view But even from actually being able to make the correct decision I think you kind of hit the nail on the head before when you said that it’s not necessarily about not having access to all the data you know, we have had less data before we’ve had more data before but it’s more about Understanding why some of these changes are happening as well that will basically set an advertiser’s mind at ease you know rather than going to crazy conspiracy theories that basically platforms want to make more money

Frederick Vallaeys: Well, I sort of knowing what to plan for the future, right?

So if we see, I mean, RSA versus ETA is a really good topic here. So responsive search ads, that’s kind of the way forward. But Google hasn’t really officially announced the deprecation of expanded text ads, however, in the interface. Now, if you want to make a new ad, RSA is the default. And in fact, it’s very difficult to find how to do an ETA.

You still can, but Google really disincentivizes it at some level. And so now what we’re doing is like, Oh my God, is Google going to pull this? Is it going to be this year? Is it going to be next year? Like, or is nothing else going to change? Is it just always still going to be there? And we just have two options.

And it’s that. Volatility and the uncertainty, like you say, in your report. That’s what makes it tough because how do I plan for my business when I don’t know if, and by the way, search terms reports and that change was basically announced on the day that it happened. Yes. So everybody’s scrambling, right?

Nobody likes to scramble this, especially when we’re already in 2020 and every day feels like a scramble. So give us some heads up, right? That is, that is entirely true. Yeah. And you know, this is, I think this is the one point where marketers, no matter how long you’ve been doing this job, you just feel extra uncertain about, because we are never in a position where we can easily make you know where we can easily understand why a change has happened where we’re not privy to some of the thinking that the platforms are doing behind these changes and you always find yourself in the middle between a client and a platform, you know, one of them changes something and then you still need to make sense of it.

Lukas Adamec: This kind of feeds. Our view that we’ve got in the report as well, that basically, especially when it comes to AI, you need to continually test and retest and reset your tests and, you know, basically never really stop. Where previously you wouldn’t necessarily have to do it as often as you might do now.

But yeah, it’s, it’s just one of those things where I think extra information. Would really be helpful for a lot of advertisers.

Frederick Vallaeys: Yeah. Hey, let me switch topics here for a minute And we’ll go back to others and report but some of the viewers are saying make this actionable So so here’s the script that you can download it’s a geo anomaly detector, right?

So we’re talking about uncertainty fluctuation One thing that’s been very challenging for people is that in 2020 The world has completely unflattened. And the point that I make is that when I used to go to conferences, when I could still get on a plane every city in the world that I go to, they have a Starbucks guaranteed.

So if you were Starbucks, it was really easy. You could just market to all these places and basically know that. There’s going to be a Starbucks no matter where people are searching from. But now there might be a Starbucks, but it could be closed or it could be very limited in store capacity, or it might be to go orders only.

Right. So how do you stay on top of that? Because I live in California. I have no clue what’s happening in New York. I have even less of a clue what’s happening in Paris and London where you guys live. Right. So but the numbers will actually tell me that story. And so here’s a script that you can download and it’ll basically pick up and say, Oh, there seems to be anomalous behavior in a certain specific location.

And that could be a country, could be a state, could be a city. And now, you know, to go and investigate that, right? So so that’s a script that can really take some of that uncertainty and that volatility that we’ve talked about, and at least give you more insight into where that volatility is happening so you can respond to it.

Anyway, that’s a little sidetrack.

Andres Hjorth: It’s great because it’s the kind of thing when we, you know, we discussed COVID as well as an opportunity, Lucas, we felt that this is a great opportunity. occasion for people to, you know, take a step back and look at the bigger picture, unplug the data, rethink a bit, and then plug it back in afterwards.

Yeah. Because most of the time when you’re in, you know, you’re optimizing. So this whole business of you’re optimizing, optimizing, optimizing, you’re starting somewhere, but maybe you started in the wrong place. And when do you have the opportunity actually to step. Step back and look at the bigger picture again.

And then, and then, and then try out, which is something, you know, that people need to do. And so also something that we’re actually recommending in the report to sort of make cycles, optimization cycles that have an end. Yeah. So just don’t just carry on with the same thing you already did you know, unplug sometimes and.

So a script like that to alert you to, okay, here’s something, you know, maybe you need to just to step back and, and, and have a, have a broader view of the, of the data. It’s interesting.

Frederick Vallaeys: Well, and since we’re talking about scripts, I’ll put one more up on the screen. This is actually from Niels Roimans. But we were talking about search terms, right?

So this is an anomaly detector for search terms, basically rising and declining search terms, all nicely put onto a spreadsheet. Hint, hint, you can also do this in Optmyzr without a script, but here’s a script if you want to do it for free. But yeah, so that’s another one that helps people deal with volatility.

All right, so we’re only about 10 minutes left, so Anders, show us maybe some more of the slides that you did. Yeah, maybe

Andres Hjorth: I’ll, I’ll, I’ll, I’ll run it. And I’ll give people the link for that report. This so we asked people about so one of these big things that is that’s happening is you know, there’s ITP in Safari that came, you know years ago already So all the Safari browsers would start, you know We’re moving cookies and they started with the third party cookies And then they actually went into the the first party cookies as well at one point and We so ITP has been updated.

We called I think in the report, we got a little funny and called the last update, the Dory update. You remember Dory from Finding Nemo? Because fish, you know, forgets everything every 30 seconds. Like if you have a website and somebody comes to see you and say, Oh, hello, welcome to my website.

And then how are you today? And then, you know. Next day, they come back and say, Oh, hello. How are you? Welcome to my website. I’ve never seen you before. Like, I’m just Purchasing the data every day because the cookies are not sticking in the browser. It’s just a funny Image of this but the trend goes on and Google have announced that they will remove third party cookies from the browsers I think they extended the date a bit.

I think we’re talking about 2022 doesn’t really matter. The end of cookies is arriving. And this will have an impact on digital marketing to, you know, big time. And we’ve got. So we ran this. We ran this as a survey as well. LinkedIn just a couple of days ago and and Now, in this case, you’re seeing 35 percent of people saying, Yes, we feel that we’re prepared for this.

You know, we didn’t, we didn’t dig in. How did you do this? Because, you know, there isn’t much you can actually do, you know, you don’t know what’s going to happen. Nobody knows what’s going to be possible or not. On LinkedIn, we had a 25 percent I think in the in the poll, we had 25 percent saying where we feel ready for this.

But a lot of people like, well, okay, we, you know, we just gonna have to wait and see what, what becomes possible and impossible, but it’s going to change the whole nature of the digital marketing ecosystem. So, that’s one of the biggies that people need to think about in the future. Let’s move on to the next few here.

Have you started planning for this? Well, yes, a lot of people are still planning, but they still don’t feel prepared for it. Very obviously, will it impact your choice of technology stack? Certainly for a third of them. I mean, I think there’s a lot of technology vendors out there who are going to say, Wow, this is a good time for us.

Volatile times are good for business, are good for making people change their minds. It’s you know, a lot of uncertainty about how this is going to pan out. I mean, you know, So we’re feeling, you know, we’re giving directions on certain things, but, but, you know people are lost. I think in this setting, we tried to chat out some of the things.

I think what the first thing here is we, we try to give dimensions to, to some of these things that we’re, we’re seeing happening. There are things that have to do with the human versus the machine versus the data. Some of them are more human and data. How do we, you know, how do we work? Some of them are more human versus machine.

What kind of technology are we using? We try to describe these six things. Biggest single recommendation to prepare for the end of cookies. Oh, my God. I think we’re going to get back to this. Yeah, I’ll ask Lucas. So we came up with with six things that we find important. We talked about. We talked about one of them already.

The one called deciphering black box was basically saying, Okay, this is black box. I don’t know what’s happening inside. But I want to try to understand as much as I can about it. So that’s one. We talked a little bit about strategy was performance, which sort of a big challenge. They’re saying, Okay, am I looking long term, maybe, or just short term?

And how can I make these things actually work together? The variable technology? Am I going to have to update my technology on the basis of the things that are happening? And there was this one you know, this is a funny one, the death by optimization. It’s sort of a one that you can run into if you optimize too hard, and you get too short term, basically.

Where, you know, if you, if you don’t, If you don’t sort of step back sometimes, purge the data take that broader view you’re going to optimize yourself into a very tiny business amounts. Yeah. And finally, you know, the, the one up there sort of very much the, the human dimension of remotely anywhere or working from working from anywhere, which is.

Becoming something. So we, we in the survey, it’s quite interesting. There’s a, there was 20 agencies responding and we had a, I think we had a two to four of them either already going 100 percent remote, no offices or in the process of going there, which is sort of very, very big steps in terms of organization.

I mean, how you organize a business to not have an office anymore, crazy stuff. So so these are these are sort of the big dimensions that we that we touch upon at the at the end of the report. So As we’re almost at the end I think I think this is there’s the final start. I think you, you showed the the discount code code already, didn’t you?

Right on the screen, you did

Frederick Vallaeys: discount code. Awesome. And so Lucas while people can Download the report here. Let’s talk about those two tactical questions. What is your single biggest recommendations about the end of cookies? And then also asking how will remarketing shape the VUCA world without cookies?

Lukas Adamec: Yeah, so

That’s basically where the biggest impact will be because without cookies you get basically less data so your remarketing will then be You know Therefore less efficient. Now I’m very pragmatic in, in my view of the sort of whole cookie removal in that. Yes, there will not be any cookies, but the platforms will still be able to track users and what they do online.

And what this means is that as an advertiser It’s it’s almost a little bit of a wait and see game from my point of view because I need to see what the new developments will be that will basically you know depreciate cookies, but then it will sprout new tracking methods

Frederick Vallaeys: Like the high level it just means that if you have a website like really find Reasons to encourage users to be logged in give them benefit being logged in because that is the thing that enables you to track without a cookie.

And that enables you to track across sessions. I mean, tracking within sessions very easy. Just put stuff in the URL. There’s no cookie needed for that. But across sessions, that’s where you want to have that logged in user. So I mean, think about those types of scenarios. And, and it does mean maybe installing a little bit of a different technology stack and thinking about tracking in a different way.

But it’s certainly not undoable. And, and the other thing, I mean, we’re already looking at the removal of third party cookies and those are usually. More of the sketchy cookies, right? To be honest, like I think a lot of people don’t really have that big of a problem at first party cookies because they Are the thing that helps make for a better experience on individual websites.

It’s more that okay. I just visited You know a financial website and now Facebook got that data Like I did not actually go to Facebook and like why did that happen? And that’s what confuses users. That’s why it’s A bit nefarious in some cases. And watch the movie, the social dilemma on Netflix. If you don’t understand what’s happening with all of that.

Fantastic movie, The Social Dilemma. Right. But yeah, users don’t want to be manipulated by the artificial intelligence machine to buy stuff they don’t need.

Lukas Adamec: I think interestingly as well, we kind of get a little bit of a visibility or more visibility on you know, some of these cookie changes here in Europe.

Because the changes to GDPR, which is basically data protection means that you now as a website owner have to show exactly which cookies you want users to sign up for and basically allow you to you know, track them across your website journey. So you’ve got your necessary ones for the running of the websites.

And then you’ve got all of the additional ones such as all the social cookies and everything else, and you can opt in or opt out individually. Between those if you’re willing to invest that time the vast majority of users still just click on accept all but I think generally speaking There’s about a 20 drop off in terms of you know, the data that you will be able to track Just by asking whether people will accept those cookies or not.

Andres Hjorth: Yeah, I want to give my very quick Answer to those two questions. The first question was will remarketing be in trouble and the answer is yes and you know, hopefully all the bad stuff will die off and then there’ll be some intelligent use of user data And the second one was what is the one biggest recommendation?

In the sort of in the in the end of cookies is well get direct relationships with your end clients via Email or some other, you know way of having that of owning that data of your relationship with clients if if you only have Access to your clients via, I don’t know, via notifications or via audiences or via Amazon or via some other intermediary.

You’re probably going to lose a lot of that access. So build your proprietary data on, on use and get direct relationship with them is my one piece of big advice there.

Frederick Vallaeys: Good advice. Well, gentlemen, thank you for writing the report. Thanks for offering a bunch of copies for free to our audience. Thanks for joining me.

And we will be back with PPC Town Hall taking a few weeks off, but November 18th, we’ll be back with a smart bidding session. We’re going to get philosophical, but also very tactical. We’re going to have Google on the call to tell us what they think about smart bidding. And I will tell you given what Google believes we should do.

How you can still manipulate it, still optimize it using other tools. So definitely a topic that I know is close to a lot of people’s hearts. So keep watching. Thank you for watching today and have a great rest of your day.

Andres Hjorth: Thanks Fred.