Google Ads AI Isn’t Strategic — You Still Need to Be ft Mike Ryan

Anu Adegbola (00:01.623)
Hello, Mike Ryan. Welcome to PPC Live the podcast.

Mike Ryan (00:05.619)
Hey Anu it's a long time coming. Way overdue actually.

Anu Adegbola (00:10.135)
Absolutely, absolutely way overdue. I just feel like Mike, you know, we've been at, you know, several events, especially the the the ones with our good friend Sandra Finlay at, you know, whether it was like out in Germany at her SMX Advance events. And I just feel we we're always just bumping to each other talk. I'm always seeing like your great posts on LinkedIn. And I feel like like I just kept going on and trying, and each week I'll be like, my God, who should I do next? And I feel like this last week I was like, my god, Mike has not been on the podcast yet. How is that?

Mike Ryan (00:36.142)
Yeah.

Anu Adegbola (00:39.743)
How is that working out? so yeah, thank you so much. That on I as as usual, my guests know now I just give my guests short notice on like, yeah, can you be on the podcast in two days' time? Because I want this to be a very like raw conversation, not nothing like too polished, something that just really shows what we're doing and the and the mistakes that we've come about. yeah, but before we get into the mistakes, because you know, one thing that's a big point for

For me to reiterate for our audience, for our listeners, is that you know, I bring on people who are doing well in their career, who are doing amazing stuff now. But, you know, we'll then get to the stories about the F ups that left that did, but the F ups that still didn't turn, turn around. So who's Mike? So Mike is someone who's who's been like he's the head of e-commerce insights, that's correct, at Smarter, at Smarter e-commerce or SMEC as sometimes it's called. I always see a S E C. So if you see that.

Mike Ryan (01:22.947)
Yeah.

Mike Ryan (01:32.163)
Yeah.

Anu Adegbola (01:36.354)
He's he's he does that with smarter e-commerce and they turn millions of data points into practical strategies, especially across Google Shopping, Performance Max, AI powered advertising, of which there've been lots of that lately, and the wider retail ad ecosystem. He's got a background spanning both retail operations and advertising tech and brings a unique focus on what really matters, and that's protecting margins and driving profitable, scalable growth.

And I think one way to really I I that might just have sound like of like a string of words of what Mike does. But for me, my biggest thing of seeing what Mike does is you're someone who does a lot of research and brings it to the stage and breaks it out for us. So be like, okay, this new thing, we've tested it, and this is what's going on. And I think it's really good to get that that that perspective from.

Mike Ryan (02:18.688)
I mean.

Anu Adegbola (02:31.713)
From experts, you get a lot of people just sharing, like, yeah, we won this client. We, you know, we do 10 million in revenue, blah, blah, blah. Actually, you know, be more generous. And I I think you come from a very like you give a very educational perspective to your talks, which I really love. but yeah, thanks, thanks Mike for for joining us. And hopefully we'll get one of those educational talks here again.

Mike Ryan (02:51.832)
Well Yeah, yeah. Thanks for having me and thanks for the for the great intro. I mean it's the same thing, like like the thing is that i there's this bias of what you see online and stuff like that. That is that's the times where I succeeded and I found something and I made sense or tried to make sense of it, but what you don't see are all the dead ends, all the times I explore something and I can't get an answer or

it's just not interesting. So there's its own kind of survivorship or whatever going on there with the the social media and stuff like that. But

Anu Adegbola (03:25.853)
Yeah. Yeah, the social media aspects of it. And in fact, I think even one of your stories will even touch touch on this, like the social media aspect of what people want to share and what people are willing to share. And even if they've got it wrong, how they spin it so that they still look like the good guy in the story. I'm like, guys, it's interesting too. We want to hear the mess. Well, we want to hear a bit of chaos. We want to hear a bit of like, my god, that went totally wrong. But hey.

Mike Ryan (03:48.354)
Yes.

Anu Adegbola (03:54.316)
I still survived it. So that's what I try to bring to our industry through the podcast of like let's let's hear about the chaotic moments. but before we get before we get into the chaotic moments, I like to do this with my guests. Fun facts, any interesting facts about you wanna share that people might not know about you?

Mike Ryan (04:11.31)
let's see. Fun fact, I am an identical twin. So there's two of me running around out there. yeah. Yeah. So yeah, there's my twin my my twin brother Phil. And yeah, I'm all I'm getting like invited by Google to their headquarters and stuff like that. Phil gets invited to Salesforce headquarters. So I'm a Google guy, he's a Salesforce guy. And there you have it.

Anu Adegbola (04:16.411)
well. my god, I don't even think I knew that. that's so cool. well.

Anu Adegbola (04:32.957)
Right, okay. He's a Salesforce guy. Okay. But I guess that there is there's a little bit of a bridge. You guys understand each other's languages, I feel, 'cause you get you need you need data. Data for Salesforce, data for digital marketing, all of that kind of stuff.

Mike Ryan (04:38.243)
Yeah.

Mike Ryan (04:42.03)
Yeah, that's the

Yeah, we're both big nerds, so

Anu Adegbola (04:50.333)
Nice. that is so cool. Yeah. I mean, I've got twin cousins. I was gonna be like, yeah, there's and in fact, I think, yeah, my auntie has twin girls and twin boys. And it's like, so I four. And I was like, wow, okay. yeah, so twins, twins rung strong in my family as well, which is real cool. Anyway, let's let's we are not here to talk about family histories and family trees. We're here to talk about you know, f ups. And I I I couldn't be more delighted to have

Mike Ryan (05:03.96)
That's wild.

Anu Adegbola (05:20.083)
Mike or that's the latest on a long string of like brave, you know, open, vulnerable people who are literally willing to be like, Hey, I've I've messed up before. Even though things are going well, Help, I've messed up before, because we all do, and those are where the biggest learnings are. So passing it on to Mike now. Mike, what's the F up you'd like to share with us today?

Mike Ryan (05:40.323)
Sure, yeah. love love to talk about my mistakes. There's more more of those than than successes. No. one thing so just j just to explain quick a little bit about my my role, that mouthful there, head of e commerce insights. you know, my my role is probably that job title is different than probably a lot of people listening to this podcast.

I d I do not do anymore a lot of stuff where I'm just like hands-in operating the ad accounts and the campaigns directly. It's now my bit to like I basically research and learn about campaign technology, platform trends, like what's going on with cost per clicks and stuff like this, market and competition trends, and then I bring this stuff.

Anu Adegbola (06:08.191)
Mike Ryan (06:27.724)
to our client teams to inform them, to our our software developers, to our clients directly and QBRs and stuff like that. And then of course also stuff on the market like when I'm giving a presentation at a conference or whatever the case might be. But so having said that it like I'm I'm gonna sit I'm gonna safely go back about a year or over a year from now.

when AI Max for search was a lot newer, and now I have the comfort of of that distance of time. But like at the start, I I I work with a lot of like big data across hundreds of advertising accounts and and try to learn and draw conclusions. But when a technology is new, you don't always have that amount of data and stuff like that. And I get also called in for

Anu Adegbola (07:01.792)
Yeah, yeah.

Mike Ryan (07:24.684)
Early opinions and sort of running like due diligence in a way like it does is this technology credible? Should we be using it? And these are calls that it can be hard to make early on. So where I messed up, there's I had a couple of mess-ups in the early days with with AI Max. There's one where a client team brought me in to kind of they were launching AI Max in some single campaigns.

in their account. They weren't launching across the whole account or anything like that. but they wanted just my read on it and what do I think about this so far. And the thing was that the performance, like AI Max Research, was generating a modest uplift of impressions, and the performance looked really solid. And I guess maybe a little lack of curiosity on my part, or I was looking, I was like, okay, it just looks

Fine. Actually, it looks really fine. It doesn't look like there's anything to see here. I didn't dig a lot deeper than that. It's like, yeah, go for it, keep going. And then I I checked in again on it pretty short time later, but it w it the amount of impressions was scaling dramatically. So this me went from let's say an interesting, you know, modest, like a pretty good like 10%.

Anu Adegbola (08:27.551)
Okay. Okay.

Mike Ryan (08:50.422)
More impressions or something like that, good performance. And now it was dominating some of these entire campaigns. It was, it was providing the vast majority, like 80% of all impressions in these campaigns. And they were growing a lot. and the thing was that the performance still looked really, really good. But at that point, it's like, okay, well, what is actually happening here? Like

Anu Adegbola (09:04.533)
Okay.

Anu Adegbola (09:09.719)
Yeah.

Anu Adegbola (09:14.751)
Exactly. Dig in. We need to dig in. Yeah.

Mike Ryan (09:17.27)
This this looks too good to be true. Like it these were not these were like mature search campaigns that it it didn't make sense that they could just suddenly be that much unmet demand that AI Max could find for them. And so yeah, then I was digging into the search terms more and which I should have done from the start and I found that

Anu Adegbola (09:24.727)
Mm.

Anu Adegbola (09:38.729)
Right. gosh. What did you find?

Mike Ryan (09:42.979)
I it's tricky to illustrate this too specifically because I I can't reveal like who this client was, but they're like they're like a a market leader. Like, you know, here in Europe, they are they're big in their market in their category. but they were advertising massively against kind of the global category leader, so a direct competitor of theirs. AI Max was just massively scaling into

Anu Adegbola (09:47.903)
Yeah, yeah.

Mike Ryan (10:09.998)
into these competitors' terms. Yeah. And you know, there was a reason why that wasn't happening before, and it's because they didn't want to do that. Because they don't want to initiate like a bidding war. This is a much bigger advertiser than them. and so that was kind of not good. and you know that in fairness, I I mean AI Max

Anu Adegbola (10:13.095)
It's oof.

Anu Adegbola (10:20.895)
Rights. Yeah. Yeah.

Anu Adegbola (10:31.711)
No. No.

Mike Ryan (10:37.772)
That's the thing about AI MX. Like it was doing its job. It's not strategic. Like its job is to find pools of demand that you are not currently meeting with your keyword targeting and stuff like that. And it and they were not targeting their competitor terms because they didn't want to. And AI Max was like, hey, here's this massive volume of queries with that perform really well. and yeah, so that was then I had to get back involved there.

Anu Adegbola (10:41.11)
Right.

Anu Adegbola (10:49.493)
Yeah. Yeah.

Anu Adegbola (11:01.857)
Yeah.

Anu Adegbola (11:07.625)
Yeah, yeah, of course. I mean we could have a whole whole episode about you know, how AI Max chases for, you know, different kind of terms and whether it should do it in that way and whether it's the advertiser's fault or Google you know, the the tool is not mature enough for it to be doing that way. But you know, yeah, that's again, that's that's a different kind of conversation. But let's let's focus on like what you could have done differently. Like what so what was

What was the pain and and so this is a c a client's account. Did they did they so so how did the conversation with the client go and yeah, were they happy, upset? How did it yeah, how did that go?

Mike Ryan (11:41.25)
Yeah.

Mike Ryan (11:47.407)
Yeah, well, I mean, I think that they were I mean, I was happy that like they were basically pretty grateful that we realized that this was going on and not too much time had passed and stuff like this. so it was a relatively it was an early enough intervention, which is great. but you know, I the thing is that

People have a certain level of trust in me and stuff like that. They and they it's like, you know what? Mike gives it the green light, fine, very good. Let's carry on. And that that

Anu Adegbola (12:15.734)
Yeah.

Anu Adegbola (12:23.841)
I mean, Mike, I think to be fair, there are loads of people in the evening just in the general industry that just do that. I see people's decks. So it's like Mike said this, so it's true. And it's and obviously a lot of the time it it is. But yeah, it's a big it's a big responsibility, isn't it?

Mike Ryan (12:32.302)
I don't know.

Mike Ryan (12:39.118)
Well yeah, because that's 'cause, you know, this extends to the industry level and I have another F up related to AI Max as well. But yeah, I mean ultimately it it was manageable, but I I think also it was important for like I didn't know that AI Max was gonna behave this way and none of us did. So it was also a good lesson that came out of that. We start we understood a lot better kind of how AI Max thinks on the basis of that.

Anu Adegbola (13:07.189)
Yeah. Yeah. Yeah.

Mike Ryan (13:09.43)
And that was that was important. But you know, would it have killed me to look at the search term report back on day one? No, it wouldn't have. And this was would have been because I did go back and look and this was instantly recognizable from day one. That is what AI Max immediately s it immediately started doing this behavior. And I could have caught it back then, but I just basically thought, well, nothing really to see here. And it was a little bit of a lack of curiosity or that extra dil diligence.

Anu Adegbola (13:23.477)
Yeah. Wow.

Anu Adegbola (13:39.212)
Yeah, yeah. It's I d and I Jeff definitely think with all the whole like, you know, the new updates that are coming out every week, every month, and new ways and even like how the tool is evolving. So yeah, even AI Max, it be one thing like you know, AI Max has come out like and every month they're doing like new okay when you can now do this, you can help these you can make these. You really need to have that due diligence after setting the campaign live of like what should we check. Yeah.

Mike Ryan (13:46.862)
Yeah.

Mike Ryan (14:01.678)
Yeah, you know, the the it's this is trust but verify. Like you you you you really it's a it's a brand new technology and you have to go and take a look and understand it best you can from from from early on. Yeah.

Anu Adegbola (14:17.407)
Yeah, yeah, absolutely, absolutely. And so this is a client I'm guessing you're still working with. Yeah.

Mike Ryan (14:23.724)
Yeah, we're still working we're still working with them and they're investing in AI Max still, but with guardrails in place, you know. yeah. 'Cause y you can you can in AI Max's defense, usually if you find a problem with it, you're able to fix it or rectify it. so there is that. like another thing so this is this was something I then started looking into AI Max with a bit more like we there were

Anu Adegbola (14:30.015)
Wow, yes, yes, yeah.

Anu Adegbola (14:41.685)
Yeah. Yeah. Yeah.

Mike Ryan (14:53.1)
some more campaigns, but still we didn't have a huge amount of data on it yet. So this was all anecdotal, let's say I started jumping around and I I I found a couple of campaigns that had huge amounts of search partner network in there and the performance was horrible. And I was like, okay, here's the next thing. Great. I found out that it it it really likes brand traffic, which I think is

Anu Adegbola (15:10.135)
Mm.

Mike Ryan (15:22.338)
generally true. It it it'll it'll go after a competitor brand in a heartbeat if you let it, which could be fine with you or not, depending on your strategy and stuff like this. but then I was like, here's the next thing, it's going after Search Partner Network. and then I raised a flag about that on LinkedIn and you know, the post went kind of viral or within our niche as these things as these things do. And

Anu Adegbola (15:26.549)
Yeah.

Anu Adegbola (15:29.985)
Right.

Anu Adegbola (15:35.723)
Yeah.

Anu Adegbola (15:43.135)
Yeah. Yeah, yeah.

Mike Ryan (15:48.451)
You know, I I later heard from Google directly got involved. Ginny Marvin was like, no, no, the AI Max does not like Search Partner Network. and as time went by, I revisited this because I know what I saw. but it's really true. AI AI Max by d there it has in general, it does not like to go after Search Partner Network. It I have seen I've I have seen cases where

Anu Adegbola (15:52.374)
Yeah.

Anu Adegbola (15:56.479)
Okay.

Anu Adegbola (16:03.211)
Yeah, exactly.

Mike Ryan (16:17.398)
And I don't know how to get to the bottom of this. And I ask Google, like, okay, I agree with you, this isn't generally true, but sometimes it is. Can we figure out why this is happening? And I and I never got to the bottom of that. But the problem is with something like that, you know, I really have to be careful with statements that I make. I feel more pressure about this than I used to because there are more people following and stuff like that. where

Anu Adegbola (16:23.167)
Hmm, yeah, yeah.

Anu Adegbola (16:38.625)
And

Yeah.

Mike Ryan (16:44.716)
Like it was actually a false generalization and I can't easily take that back. There's no easy way for me to fix that because people read that post and if I go and publish a redaction or something, what am I gonna do? Publish a follow up post. It won't go viral because it's boring. No one cares.

Anu Adegbola (16:47.637)
Right. Yeah. Yeah.

Anu Adegbola (16:57.769)
Yeah. No. Yeah. Yeah, exactly. Yeah. LinkedIn the algorithm likes a more much more of a controversy than a whole like, yeah, you know, sorry, made little mistake. Anything. Nah.

Mike Ryan (17:09.26)
Yeah. Like I so that you can't take that back and there y you know, I can never take that back, which is a pity, I guess, but

Anu Adegbola (17:18.143)
Yeah. Now like you know, and yeah, and I I know we talked about this and I know you take this seriously now. Several of us, I I do definitely I take this seriously as to, you know, what I'm gonna say and even like what I'm even gonna say about brands or people I've worked with or you know, how things, you know, may or may not work. And because yeah, people sometimes people are like, LinkedIn it's like, you know, false and all these AI comments and people have a lot of negative things to say about LinkedIn.

But for some of us, it is a bit of a stage. It's as much of a stage as it's going to be, you know, for our career. And, you know, there are people we respect that respect us that we want to, you know, do right by and be like, well, this is the one platform I care about, you know, and you know, and and and and people have have are are are willing to to listen and hear and allow that to take that and then implement stuff into their campaigns and you know, do you know, move things going forward and

It's really important that a lot more people in our industry approach it with a bit of this kind of integrity than I I I perceive from you, Yugu, because some people would just be like, well, I got my engagement, I knew it was wrong. who cares? You know, I'm not gonna name names, but there are some people that come to mind right now. I'm like, you couldn't care less. You couldn't care less that you made a mistake there. God, yeah.

Mike Ryan (18:24.43)
I know.

Mike Ryan (18:32.636)
unfortun More than more than some, unfortunately. But yeah, I mean I I I really feel a sense of responsibility and like I I also like man, I made some statements on a podcast recently about the August seventeenth updates and suddenly I'm getting emails from Google about this stuff. I'm like, what what has my life become? but

Anu Adegbola (18:52.983)
Jesus. Yes. Yes. Yeah. When I get the Google emails of like, you know, I know that thing you wrote about, my brain goes, God, what did I do? Yeah.

Mike Ryan (19:02.83)
Yeah, yeah. But but I I really like, you know, I've spent I've spent year I've been posting on these places for years and you build up a reputation and stuff, but I I I take it seriously and I I think that's you know, I hope that I've earned that reputation in a sense, but I I I don't want to mislead people. It's and it's so easy to be hard on Google, but when you're hard on them for the wrong reasons, that's not okay either.

Anu Adegbola (19:16.439)
Yeah.

Anu Adegbola (19:25.974)
Yeah.

Anu Adegbola (19:29.993)
No, no, no, you're not helping anybody else. And honestly, and you know, I think the oil industry will will agree with me is that it's that yeah, Google made a great decision in the form of hiring someone like Ginny Marvin, who we knew pre her joining Google, she cared. She cared about the work, she cared about you know integrity of the work. So it was this is this is someone that's going into Google who's really trying to be like, guys, we need to, you know, understand advertisers more. And I think we've definitely, you know, seen seen a shift.

Mike Ryan (19:41.389)
Yeah.

Anu Adegbola (19:59.018)
in that way so everything is not going to be Google's fault. Like yeah, maybe some some things will be, but it's not, it's not going to be everything. And you've got to you've got to do a bit of digging yourself into how these things, especially these AI tools, AI tools work. One thing I do want to get get get to ask about though, which which I think you know the the the listeners will be really interested to to hear. Like so these kind of even in both examples we saw about how you know yeah AI Max went after

Competitive keywords, AI Max, you know, went for search partner network. what what how do you prevent that? Because I I much less than you. You say you're not an account. I have, I do not know what a Google Ads, you know, interface it looks like a lot of the times. So how do you avoid, especially in break examples, but like starting with like, you know, not going out for competitive keywords? Is it about just adding neck making sure you're more on top of negatives? Or are there other tools to help with that?

Mike Ryan (20:48.536)
Yeah.

Mike Ryan (20:54.05)
Well y yeah. They yeah, I mean they they they offer negative keywords, brand inclusions, brand exclusions, and and so you you can you can control a situation like this. The tools are in place. So that that's that's not the problem. And like in that case you can yeah, apply a negative at the campaign level or there's you you can solve it different ways. and with something with Search Partner Network, you can

Anu Adegbola (21:09.066)
Ja

Anu Adegbola (21:18.559)
Mm-hmm.

Mike Ryan (21:22.57)
opt out of search partner network so these these are not like insurmountable problems but i you know i encourage people you should go and check your search term report and see what the heck is going on in there you can like like also in a max's defense the reporting is pretty good you can you can look at that ai max match type and so you immediately know what is ai max doing here

Anu Adegbola (21:24.086)
Yeah.

Anu Adegbola (21:37.175)
Yeah.

Mike Ryan (21:50.105)
They also have the match source call column so you can understand is this stuff coming from the broad match type part of AI Max? Is it coming from the DSA type part of AI Max? Like what's going on here? So y like the tools are all there in order to kind of figure out what's going on and to solve many of these problems. there could be ones that are that are thornier or harder to control, but

Anu Adegbola (21:52.321)
Yeah.

Anu Adegbola (22:01.76)
Yeah.

Anu Adegbola (22:07.262)
huh.

Mike Ryan (22:19.818)
ultimately like the these problems they don't need to occur. but it's just that A AI Max or any AI tech for as smart as it is, it's not strategic. It doesn't always understand your business context or, you know, the the decisions that you have made or haven't made or like th things that you've chosen on purpose not to pursue. it won't necessarily have that knowledge or those constraints and it's

Anu Adegbola (22:25.312)
Yeah.

Anu Adegbola (22:40.213)
Mm.

Mike Ryan (22:49.206)
It's kind of agnostic, you know, and it's just gonna explore opportunity wherever it can find it. And that is again where you need to put guardrails in place.

Anu Adegbola (22:58.581)
Yeah, yeah. Are there a lot more guardrails that we need to put in place now than let's say five, ten years ago?

Mike Ryan (23:06.422)
Yeah, yes and no, because like people used to really heavily structurally enforce their their ad account and their strategy. Like you would you would have maybe device campaigns or like y like and you know, using these device modifiers and stuff like that. and the funny thing is that Google has gone back in this direction a bit. Like even with with PMAX you can

Anu Adegbola (23:22.591)
Yeah. Yeah.

Anu Adegbola (23:34.463)
Yeah.

Mike Ryan (23:34.967)
You can basically create these, you can kind of sculpt your way toward a device specific campaign. but the question is, should you? Because Google does consider the device in their bidding and stuff like that. and it now there is for people who have a highly like granular account structure, there are so many cases where that's probably your F up like.

Anu Adegbola (23:39.327)
Yeah. Yeah.

Anu Adegbola (23:44.319)
Okay.

Anu Adegbola (23:48.437)
Yeah. Yeah.

Anu Adegbola (23:59.49)
Yeah.

Mike Ryan (24:04.79)
one of the single biggest leverage things you can do is to consolidate more data for the for the AI to work with. So like if you if we're talking a search campaign, your segmentation needs to be kind of thematic. And if we're if we're talking about like shopping or PMAX, I think you need just every segment that you that you make, it needs to have a reason

Anu Adegbola (24:09.493)
Yeah. Right. Yeah.

Anu Adegbola (24:19.777)
Yeah.

Mike Ryan (24:33.432)
For existing. It needs to kind of defend its own existence. and so I'm not saying that because I would not put everything all into one big dumb PMAX either. That's not gonna be the solution either. but you need to really like if you're gonna make a device split, for example, you have to think that through like, how will this affect my monthly conversions? And do I do it like the is each campaign still gonna have.

Anu Adegbola (24:36.692)
Yeah.

Anu Adegbola (24:45.34)
Yeah, yeah.

Anu Adegbola (24:52.62)
Yeah.

Mike Ryan (25:01.514)
at least thirty or ideally sixty plus conversions per month. There's a lot there's a lot to look at.

Anu Adegbola (25:07.979)
Yeah. Yeah. Absolutely. There's a lot to look at and in terms of like there's a balance, isn't it? There's a balance of making sure you're consolidating to get the amount of the right amount of data. So the AI does its thing, but there's also also a bit of like, yeah, separation theming things well to ensure that you're actually getting your your your your the right customers are seeing the right ads, you know, even like something as simple as that.

Mike Ryan (25:18.125)
Yeah.

Mike Ryan (25:29.688)
That yeah. Yeah, that's right. I think like in the past we would have had these very granular campaign and and account builds that were and and and that was very tactical. And now like that segmentation should be more strategic and rather than tactical and and not s so ex intense. 'Cause I I still come across accounts that have hundreds or thousands of campaigns in them and you know, I I I'll tell you another this was

Anu Adegbola (25:36.469)
Yeah. yeah, I remember that. Yeah.

Anu Adegbola (25:44.812)
Mm.

Mike Ryan (25:58.857)
another F up. It wasn't well mine in this case, but I I I found it for this person. They they had they had like margin splits. and this was actually dictated from their CFO. They had to have these super granular margin buckets like from I don't know zero to five percent, five to ten percent, ten to fifteen and so on. And there was no regard and and they and they were assigned okay you will have this bucket and this will be your ROS target for that bucket.

Anu Adegbola (26:03.551)
All right.

Anu Adegbola (26:11.177)
Okay.

Anu Adegbola (26:27.486)
Right.

Mike Ryan (26:28.716)
And I I think they they were like, our CFO came up with this concept and it's it's advanced and it's but it's so that CFO has no understanding for how these these channels or technology work. And so like, you know, it wasn't even close to accurate to these ROAS targets because there were it was just totally volatile. What was going to be happening this month or next. and it's like, okay, let's find a middle ground here where you can have

Anu Adegbola (26:33.865)
Yeah. Yeah.

Anu Adegbola (26:52.47)
Yeah.

Mike Ryan (26:57.856)
some kind of margin differentiation, but what is the point of having a hyper specific ROS target if the algorithm will never achieve it because you've split your data into all these campaigns needlessly? yeah, so

Anu Adegbola (27:09.877)
Yeah. Yeah. No, absolutely. Absolutely. Okay, like yeah, I we we've gotten into a lot of information there. I feel like half an hour has just gone by in a second there. And I've got a couple more questions to to to go through on this story before we now even get into like our next our next segment of talking about, you know, mistakes that you've perceived other people do. I you've already given one example, but even like more. Yeah, exactly. There's there's so many to talk about. So like

Mike Ryan (27:21.176)
Happens.

Mike Ryan (27:32.952)
Yeah, I get yeah.

Anu Adegbola (27:39.105)
So like going back now, like based on that experience, what do you what's your like biggest advice of people to like make sure that you should have done differently to kind of avoid those mistakes happening?

Mike Ryan (27:47.235)
Well.

I think that the the greatest virtue that there is is curiosity and it also in this age of AI too, like you need to stay curious, you need to think critically. do not take things at face value in the platform or anywhere. it's it's worth having that level of skepticism, critical thinking, and curiosity to just dig further. there's there's almost

no amount of curiosity that in my opinion is too much. and yeah, had I been more curious or less credulous, more incredulous, then I would have caught this on day one because it was it was catchable. Yeah.

Anu Adegbola (28:22.656)
Right. Yeah.

Anu Adegbola (28:32.853)
Yeah, yeah. Yeah. Yeah. Absolutely. Yeah. Yeah. Like yeah, check check things as soon as even even if you have a little bit of time and you know, yeah, no matter how confident you are, still go back in and just let that that confidence be confirmed. There's nothing wrong with that. So that yeah, you check and catch things. And like if someone has just gone through that, you know, seen that that you know, you know, there's been like that big spike in impressions that doesn't make sense and

Mike Ryan (28:43.426)
Yeah.

Mike Ryan (28:49.73)
Yes, yes.

Anu Adegbola (29:01.451)
they have to figure it out that it has gone wrong and now they're about to have that client meeting. What's your advice to them? How do should they approach that that moment?

Mike Ryan (29:11.574)
Yeah, I mean these things it's like kind of li when you're in it it can feel really big. but it's always this two shall pass and also remain remain optimistic. Like ultimately as I said this client was more grateful than anything else. I mean it wasn't so good what happened, but just

Anu Adegbola (29:17.505)
Yeah.

Yes.

Anu Adegbola (29:29.002)
Nice.

Anu Adegbola (29:34.229)
Yeah, clients clients surprise you, don't they? Like if you you could even be like a difficult client and all of a sudden this mistake happens. You get a bit terrified to tell them. Then when you tell them it's like, not as bad as I yeah.

Mike Ryan (29:36.834)
Yes.

Mike Ryan (29:44.439)
Yeah, they were they were more they were more upset with Google's technology than with us in the end. and yeah, but I would say just you know, remember that this is everything sadly everything's in your rear view before you know it, even the nice parts too, but it'll be in your rear view, it'll look and feel smaller once it is. in most cases there's some there's some big F ups out there, of course, but yeah, you you're gonna make it. You'll survive it.

Anu Adegbola (30:05.451)
Yeah. Yeah.

Yeah. Yeah. Yeah. Yeah. Yeah. Absolutely. Yeah. So that's that's how the the final word we're gonna have on that. You'll survive it. It'll get it in your review mirror mirror sooner than you actually you realize or expect 'cause you know, that's how time time works. so yeah, that's great. So we'll leave that story now back into its PTSD box. I feel like for some of my guests I'm like, 'cause some of them I just think, my god, can't believe that happened. But now moving on to when, you know,

your your your company or your you've one new client to yourself, your department has one new clients and you're trying to, yeah, improve that performance, see those quick wins that could be, you know, beneficial for your clients because an f-up was made, a strategy that was, you know, was not the right strategy for those clients. for that client was made by the previous, you know, agency. What are some of those like shocking ones or even like more recurring ones that you're like surprised that that happens so often in terms of

Mike Ryan (31:04.172)
Mm.

Anu Adegbola (31:08.363)
you know, strategies that shouldn't be shouldn't be implemented right now.

Mike Ryan (31:12.266)
I think a a lot of advertisers as I mentioned before don't have a great idea about how to segment. they're and and it it can cut both ways where people are segmenting too much or not enough and I don't wanna be like a gatekeeper and be like only I know the correct amount of segmentation but but but it's really it just

Anu Adegbola (31:21.727)
Mm.

Anu Adegbola (31:33.599)
Yeah.

Mike Ryan (31:40.993)
needs to be approached thoughtfully, strategically, because like probably the the most common thing that I see are these like brand and category splits for like PMAX for example or shopping. And I have my own view on this, which we don't have to get into a lot of detail now, but I I think that you can have a smaller amount of campaigns and still get an

Anu Adegbola (31:45.1)
Yeah.

Anu Adegbola (31:56.117)
Yeah. Yeah.

Mike Ryan (32:10.698)
awful lot out of them, a awful lot of control. I I mean that maybe leads me into the next thing. Like I I like to make use of custom labels to bring business data in and context that isn't in your feed or is like Google isn't going to know about necessarily because these are these can be much more interesting ways of segmenting or steering their algorithm than just like the brand or the category. So like

Anu Adegbola (32:34.667)
Yeah, yeah, yeah.

Mike Ryan (32:40.714)
happens all the time that I see custom labels are empty or custom labels are some completely incoherent weird numbers from five years ago and no one knows what they mean anymore. They were like built once and like yeah and and custom labels are really a way of you injecting your domain and your first party data beyond your audiences, but like like for example

Anu Adegbola (32:45.427)
Yeah.

Anu Adegbola (32:56.639)
Yeah. Forgotten about. Yeah.

Anu Adegbola (33:07.914)
Yeah.

Mike Ryan (33:10.238)
that is your way to get margins in. It's a great way to get margins in is through there. you can put your you can put your sell through rate or your t your return rates. and these are all things that are gonna impact ultimately how profitable your campaigns are and how well they work for your business. That Google has no idea about Google Google knows that the item got sold and they do not know that it got returned. they have

Anu Adegbola (33:15.509)
Yeah.

Anu Adegbola (33:20.15)
Yeah.

Anu Adegbola (33:36.074)
Okay, yeah.

Mike Ryan (33:37.324)
You know, they have no window into this, and it's not Google's fault. The algorithm is great at what it does, but you know your return rates. You can you can bring this context in and help Google. So that's kind of a missed opportunity that I see a lot is that people don't know that this is right there in front of you and you can you can do this.

Anu Adegbola (33:57.857)
Yeah. No, absolutely. So, like with even like the structure thing, I think with all of these kind of situation, like I think SEOs felt that they owned the phrase it depends. But I'm like, guys, in the PPC world, a lot depends as well. And when it comes to structure, some people might go, Well, some of my very granular campaigns, you know, it works because of XYZ. Is there would you say that?

Mike Ryan (34:11.168)
Yes.

Anu Adegbola (34:23.076)
that that may be the case and there'll be some exceptions to, you know, the fact of like, yeah, maybe granular could work.

Mike Ryan (34:31.33)
Yeah. Yeah, I mean, I think within within reason, but I I really like y you know, if you are i you if you can support a large amount of conversions across that, then it it could be okay because there is also this cuts both ways. Like in my research you really need at least thirty and as I said, sixty or more conversions, month monthly conversions for School Smart Bidding to

Anu Adegbola (34:55.391)
Yeah, yeah, yeah.

Mike Ryan (35:01.24)
kind of work consistently work and but that number does plateau at a certain point like if you the whether you have you know I'll just say some numbers here whether you have 200 monthly conversions or a thousand then it probably doesn't matter that much anymore and so when you know that's that's always the thing but for small and medium businesses

Anu Adegbola (35:02.111)
Yeah. Yeah.

Anu Adegbola (35:07.019)
Yeah.

Anu Adegbola (35:15.073)
Yeah.

Mm.

Mike Ryan (35:26.744)
You need to be super careful about that 'cause you probably don't have that total conversion pool available. and it and it's true that yeah, a bigger advertiser who has a larger total conversion pool, they can start having more fun. And that's the way it is, unfortunately. that's an advantage that they have. But

Anu Adegbola (35:31.435)
Yeah. Yeah.

Anu Adegbola (35:40.012)
Yeah. Yeah. Yeah. Yeah. Yeah. You almost feel like, you know, even though it just makes me think about all like the GMLs and all the different updates and you just feel like, Go, I I know you're it's almost feels like they only focus on, you know, the the the fun stuff that the big big companies and some of the S and Bs or the ones doing B T B they're like, Well, can we use any of these things? 'Cause we don't have that

Mike Ryan (36:03.822)
Yeah.

Anu Adegbola (36:05.025)
Ca kind of like, you know, leeway with the data or like the you know, the different structures and where each of the different structures has enough data and has enough pull up data to pull in. So yeah, it can be tough for them.

Mike Ryan (36:15.116)
Yeah. It but it's just for for so many campaigns it's gonna be a huge win just to boost the the number of monthly conversions you have compared to all these very sophisticated and good on paper things that you could do, or that sound nice. Like it it's frustrating and it's dumb as anything, but putting more conversions in is often just such a high leverage action that you can take.

Anu Adegbola (36:25.484)
Yeah.

Anu Adegbola (36:30.784)
Yeah.

Anu Adegbola (36:40.575)
Yeah. Yeah. Yeah. Absolutely. No, that makes sense. That absolutely makes sense. And you know, and even we'll kind of we'll talk coup come back even talking about like in AI Max, because when you were talking or giving like your the mistakes, my brain like, I have ten questions about AI Max, about how the new updates have come and what you think and so yeah, so coming back to you and just like, you know, like mistakes that you've seen people are making in terms of, you know, how they're approaching AI, how they're they're using it.

Mike Ryan (37:06.712)
Yeah.

Anu Adegbola (37:08.679)
Yeah, what what are your thoughts about that? And especially even like with the newer updates. I think the latest one that surprised me. I don't know if you heard about this one, is where I think it's for P Max even that you can now the kind of is like a toggle for each of the channels where you can have more focus or not. And I'm like, all these controls, are they actually beneficial? yeah, what do you think?

Mike Ryan (37:19.021)
No.

Mike Ryan (37:31.875)
Yeah. I so these that feature is n not widely available yet. I think we might have like one account that has access to it, but I I haven't had a huge amount to look at here. And again, I want to be careful because the technology's so early, but the this it's very weird the way that this is structured. And so I think you have to be super careful because it you'll see screenshots of this and it's like, there's just a slider where I say if this has this channel like

Anu Adegbola (37:59.842)
Yeah.

Mike Ryan (38:01.027)
has low importance to me or high importance. But that is it is not going to be that simple because behind behind the scenes what you're doing there is you are changing the tolerance toward your CPA target or ROAS target. And if you are you know adjusting say YouTube saying that's less important to me, it doesn't mean that you're that you're not you can't just turn off your spend to that or something, but

Anu Adegbola (38:03.691)
Yeah.

Anu Adegbola (38:07.86)
No.

Anu Adegbola (38:25.132)
Yeah.

Mike Ryan (38:30.71)
you're telling Google that they need to be really strict on the on the rawest target, for example, for that. And I think the mis like it's very confusing because these each have their own default weight by Google. And then you can change that weight. But until we can see like I need to see the numbers. Like where does this what are the actual bounds or ranges on this slider?

Anu Adegbola (38:38.475)
Yeah.

Anu Adegbola (38:46.955)
Yeah.

Anu Adegbola (38:58.412)
Yeah.

Mike Ryan (38:59.362)
The other thing is like what I wouldn't do, I would I think that you can pretty safely demote channels that you don't feel comfortable being on in PMAX. but it might seem like second nature to then take shopping and say, yeah, that's super important to me. Let me tell it's super important. But you're then telling it that by telling it it's important, you're telling it that they can have a lot of freedom on your ROAS target for that part. And

Anu Adegbola (39:14.581)
Yeah. Yeah. Yeah.

Anu Adegbola (39:24.361)
Yeah. Mm.

Mike Ryan (39:28.076)
that maybe isn't the message you thought you were sending. So I I think that the way some of these features are delivered is is weird and there's probably gonna be plenty of pitfalls there, but it's very early and we'll learn more about that as that gets rolled out more.

Anu Adegbola (39:30.465)
Yeah.

Anu Adegbola (39:43.678)
Absolutely sure. And like but like what have you seen as well? And and I know like yeah, you you answered my question perfectly and going to like, you know, just yeah, just jumping off from that is that how are we approaching these updates? And if you you wouldn't mind, you know, also giving insights about the August seventeenth update of like, you know, Rose targets and, you know, how Google is changing that. How how do you feel people are responding to these updates that you think that's a mistake?

Mike Ryan (40:07.395)
Yeah.

Mike Ryan (40:13.036)
Yeah. well we have okay. I'll give you I'll I'll give you one here. So next week I'm gonna be doing some more follow-up research on August 17th because the data is more attributed at that point. But one thing that I was able to look at immediately was how did people like behaviorally respond to this? did did people open up more budget? Did they change their ROAS targets? Stuff like that. And so I looked at the median ROAS targets.

Anu Adegbola (40:32.279)
Exactly.

Mike Ryan (40:42.562)
I don't have these numbers in front of me, unfortunately, but and I looked at people who were limited by by budget and I saw that like the median ROAS is very stable. It does the median target ROAS, but what people are asking for from the platform. This line changes slowly over time or with big seasonal events like Black Friday, but I saw that it was really jumping up the week before and the week of August 17th and has stayed inflated.

Anu Adegbola (41:03.958)
Yeah.

Anu Adegbola (41:08.208)
Mm. Mm.

Mike Ryan (41:11.874)
and so that tells me that people were saying, like, okay, I'm nervous what's gonna happen here and I want to preserve some of the performance that I've been getting on my limited by budget campaign, if it was over achieving the goal. which is perfectly valid. That's one of the actions that Google said you can do. But the surprising thing that I found was that I I also looked at the campaigns that were not limited by budget.

Anu Adegbola (41:19.67)
Yeah.

Anu Adegbola (41:28.161)
Yeah.

Anu Adegbola (41:41.847)
Okay.

Mike Ryan (41:41.847)
And I all and I also saw that a w basically a lot of advertisers, I think this was Google's fault. They there's so much fear and uncertainty and doubt about this change, and it was going viral on LinkedIn and all this stuff and and on Reddit and YouTube and everywhere. And and so also people who didn't need to touch their ROAS targets, I think people were like, Is the auction gonna break next week? And and so like the

Anu Adegbola (41:54.805)
Yeah, yeah, yeah. Yeah.

Anu Adegbola (42:06.059)
Yeah yeah Jesus

Mike Ryan (42:09.186)
The market at large increased our ROAS targets. And this is actually, from Google's standpoint, a disaster. because because no, they didn't need to do that. They didn't need to do anything. Google didn't want them to do anything, but they were so spooked. And this resulted in less spend. And also for all these advertisements like less revenue. So th this was a big market-wide F up because

Anu Adegbola (42:10.485)
My bird I'll go.

Anu Adegbola (42:16.767)
Yeah, do you need to do that?

Anu Adegbola (42:24.15)
Yeah.

Anu Adegbola (42:31.98)
Yeah.

Anu Adegbola (42:36.877)
wow.

Mike Ryan (42:38.042)
And like actually what you would want to do in a situation like that now is to behave countercyclically because if everyone is needlessly being more conservative than and and leaving revenue and spend on the table, then you should actually maybe lower your Rose target to get some market share. But this was the this was the craziest thing and I haven't really published or shared about that a lot yet. So I don't know when this will air, but maybe you're getting a sneak preview of that. But

Anu Adegbola (42:46.55)
Yeah.

Yeah.

Anu Adegbola (42:53.844)
Exactly.

Anu Adegbola (43:03.991)
Yeah, this is gonna air quite quite quite quickly after us recording. So yeah, this is gonna be very interesting. And and I'm I'm yeah, I'm sure listeners will be like, my god, when is Mike gonna come out with that? So yeah, when we will I'll keep promoting this episode and when you come out with the article, I'll make sure we share, share both. And yeah, really, really excited to see that point of view. I think like, yeah, people there have been discussions we had someone in like

Mike Ryan (43:07.665)
okay.

Mike Ryan (43:24.344)
Sure, sure.

Anu Adegbola (43:31.765)
our PPC Live Slack in terms of like, no, you know, what should we do now? How should we be doing our and it's yeah, it's just interesting to see actually what what people did. I I was just thinking as you were talking about it reminds me of the whole like, you know, Y2K, like you know, fiance people just thinking that, my God, what's gonna happen? What's gonna happen? Maybe I should do this. And then, you know, it came and went and you know, it's like, not much.

Mike Ryan (43:37.506)
Yeah. Well

Mike Ryan (43:45.869)
Yeah.

Mike Ryan (43:52.141)
Yeah, it it was it was it's it was the most Y2K thing ever because like no seriously like first off everyone imagined this whole thing like a cliff like you know my limited by budget campaigns are gonna fall off a cliff and that is not what Google said. They said that your performance will trend toward the actual or toward the toward the target. So that means it will de it will develop over time and it might never even like

Anu Adegbola (44:14.999)
Yeah. Yeah.

Mike Ryan (44:21.954)
hit that that level, but it will go in that direction over time. Not like it's gonna fall off a cliff and and so people just freaked out. you know Google's advice was to wait one to two conversion cycles to evaluate the change. So that's why you know as a time of recording next week we will finally hit the one conversion cycle mark for most advertisers, 30 days. And we'll finally because there were a lot of stuff done LinkedIn like

Anu Adegbola (44:23.916)
Yeah.

Yeah. Yeah.

Anu Adegbola (44:37.429)
Yeah.

Anu Adegbola (44:45.399)
Right.

Mike Ryan (44:50.73)
I'm seeing this already. I'm seeing that already. You you're not seeing anything yet. The I prom you're not seeing anything yet. This is sorry, but like, you know, it's been a whole month which feels like a lifetime in these, you know, twenty four hour news cycle and stuff like that. But only now can we actually start to assess what happened. So,

Anu Adegbola (44:53.365)
You shouldn't. You shouldn't.

Anu Adegbola (45:00.836)
my god, I love that. I love that.

Anu Adegbola (45:12.757)
Yeah, yeah. No, I think like if I another aspect like angle I saw it at is that, you know, people went all of a sudden going, I have something to say on LinkedIn that might go viral because it's August 17th, and I'm like, guys, you can't I mean I took advantage of it, I can't lie. And I was like, Yay, I got like two hundred likes on a post for the first time in a long time. But my my my approach wasn't like, guys, what are you doing? For me it was like

Mike Ryan (45:21.838)
Yeah. I know.

Mike Ryan (45:28.834)
Yeah.

Anu Adegbola (45:37.996)
Guys, these kinds of things have happened before. Like when when P Max first came out, when AI Max first came out, like, you know, you you wait and see. You wait and see whether it's it's gonna affect you or not, and that kind of stuff. Like, why are we all like, yeah, anyways, but so yeah, that's a very interesting point that you know it's actually like from like it's the 30 day mark, it's mid September, then you can really now actually start effectively going, this was the effect of that change for you guys.

Mike Ryan (45:45.933)
Yes.

Yes.

Mike Ryan (46:05.186)
Yes, yes.

Anu Adegbola (46:07.427)
Lloyd. Mike, this has been so great. We've gone into like so much things and I'm like, my God, I was like, I I remember asking, like, do you have a heart stop? And I feel like we've already gone over. So let's let's start let's start rounding things up. Really grateful for that amazing insights you've already given. I think this is one of those episodes where like because in previous episodes I think to myself, god, I'm talking a lot, but this time I was just like, Yay, I think I talked the least of this one and I love it.

Mike Ryan (46:17.684)
It's all right.

Mike Ryan (46:31.591)
I'm happy to r I'm happy to run my mouth. I've no problem running my mouse.

Anu Adegbola (46:36.609)
I love it. Thank you so much, Mike. but yeah, as I said, so last question, a nice not far, not not really a PPC question. If your P P C career were a movie, what would the title be?

Mike Ryan (46:38.733)
Yeah.

Mike Ryan (46:48.334)
I'm gonna go with what is the name of that movie now? Everything, everywhere, all at once. Yeah. That's that's my choice because that feels like it's my job and it's exhausting. And it also feels like it's the way Google Ads works nowadays because it you know there was a period of time where it's like

Anu Adegbola (47:00.541)
Yes! my god, I love that movie, Michelle Yo, my god, love that movie.

Anu Adegbola (47:10.845)
Alt Yeah.

Mike Ryan (47:18.718)
yeah, you've got your broad match and your DSA and your AI max for search and your P Max and and are they all doing the exact same thing as each other? Well yeah, they are, but that's fine. Everything's fine. But yeah. Anyway. Yeah. More more is more. yeah, throw some smart bidding exploration in too. More is more. But yeah, no, everything everywhere all at once. That's what my that's what a day in my life feels like, so

Anu Adegbola (47:32.023)
Everything's fine. Everything's right. You need to be everywhere. You need to just just just figure out what how every

Anu Adegbola (47:43.507)
Everywhere all the time.

Mike Ryan (47:48.12)
Yeah.

Anu Adegbola (47:49.016)
I love it. I love it. And you know what? I don't I I I I was even gonna say that I I feel that you're also everywhere in terms of events, but not necessarily. You do you you're very strategic. I I feel you're not as as and at as many events as I'd like you to be at, you know, but like when when's the the next event that you're gonna be at if people wanna come and see you talk first.

Mike Ryan (47:57.43)
No

Ha ha

Mike Ryan (48:10.126)
was going to be a D Mexico this month, but actually I canceled that. But no, actually honestly I try to because I've got young kids and I try to do as few events as my marketing department will allow me to do. but I I'll I'm I'm sure I'm gonna be at one of these March conferences like Friends of Search or Search Mar S SMX Munich. Both great events. I recommend them both a ton. And it's not I love to be at events. I just

Anu Adegbola (48:23.191)
Sure.

Anu Adegbola (48:32.117)
Right. Yes. Yeah.

Mike Ryan (48:39.682)
I can't right now be everywhere. but you can find me, you know, online. I I try to be interactive as much as I can and try to make myself available. So yeah.

Anu Adegbola (48:41.259)
Yeah, yeah.

Yes.

Absolutely.

Awesome. Well, for everybody we'll share the places that you can find Mike in terms of, you know, if you wanna also like, yeah, pick his brain about August seventeenth and when they should actually these people should you should be seeing what the effects of the update was or just in general, you know, how to be yeah, you know, really careful but also be testing all these like new new tools. It's it's very they're very interesting, but there's there's ways to do it, I feel, and ways not to do it. So you should all just be careful and

Mike Ryan (49:01.923)
Yeah.

Anu Adegbola (49:20.479)
And go in with our with our eyes open. yeah, Mike, this has been absolutely fantastic. Again, once again, thank you so much for joining me on the podcast. And yeah, hopefully we'll see next year. And also, if I yeah, I I'm I'm we're we're Mike and I are gonna work very hard to schedule him coming in person for PPC Live events. We don't know when that's gonna happen yet, but I am gonna work hard on that.

Mike Ryan (49:41.678)
We we we were we were tantalizingly close last year, so two 2027 gonna be the year then.

Anu Adegbola (49:48.107)
I know.

Yes, yes. Hoping for a big thing for twenty twenty seven for sure. All right. Absolute pleasure. Thanks, Mike. So

Mike Ryan (49:55.128)
Thank thank you for having me. Really appreciate it.