Episode Transcript
[00:00:00] Speaker A: Some of my ads are ranked poor. I'm always getting these flashy things, oh, your ads, poor. But they convert the strongest. It doesn't matter what that score is. There's no specific correlation. But I think our minds, psychologically, we just want to achieve those gold numbers. We want that gold star. And unfortunately I see a lot of people chasing and accepting, you know, opta score recommendations, for example, like, yes, yes, yes, I've got a hundred percent. And then they're like, I've had no conversions. I've spent five times what I thought I was going to do. Because we're not always looking at the why or what those changes would mean. We're just chasing that number or that green tick.
[00:00:32] Speaker B: Welcome to AI Ads and beyond. I'm Lisa Ressler. This is the podcast where we talk about how AI is impacting pay per click, advertising performance and strategy. Today we have Sophie Logan. She's a PPC specialist with over 11 years of marketing experience.
She's ranked 12th in the 50 influential PPC experts of 2026 in her role as community and editorial manager at Rough Agenda. She leads the speaker lineup content communications for Brighton SEO and Herocom. Sophie specializes in B2B SaaS and lead gen campaigns and she is not shy about saying something is not working in PPC using AI features.
[00:01:20] Speaker A: Thank you very much. Yes, I'm a very honest person.
[00:01:24] Speaker B: Well, that's great because that's one of my visions for this podcast is to get honest opinions about where we're going with AI and ads. So tell me before we get started, how did you get into ppc?
[00:01:35] Speaker A: Actually was a generalist back when I finished university, so I did a bit of everything. I even did a bit of SEO, a bit of pr, a bit of event management.
But I had some PPC responsibilities, including Google AdWords back in the day and I just enjoyed that part so much and that's what I continued to pursue. So it could have gone anywhere. I could have done pr, I could have done social, but ended up in ppc.
[00:01:56] Speaker B: I think all of those different types of marketing make us better at this job, honestly. I think you have to have a well rounded background in training.
[00:02:04] Speaker A: Yeah. And appreciation for the work that other people are doing as well. I think it's really helped me to see how much work goes into everything that everybody does in marketing.
[00:02:12] Speaker B: I have some questions for you and then at the end we're going to do a rapid fire question round.
[00:02:18] Speaker A: Okay, sounds great.
[00:02:19] Speaker B: B2B lead gen has always been harder to optimize than say e commerce or other types of businesses.
How has AI changed B2B PPC? Is this, is it making it easier or more complicated?
[00:02:36] Speaker A: If you're an account with big spend and big data running through and lots of conversion data, it's probably going to be a lot easier for you. Nowadays if you're very fortunate to work on that type of account, you'll see great success with a lot of the AI tools, features and everything that's going on. I think the problem is it's when you're looking at smaller spend accounts or accounts of lower volume of conversions coming through. I think it's harder for those types of accounts because this AI world is not necessarily built for those types of accounts. So it's confusing for a lot of accounts like that because they can see things working for others and not really understand why it's not working for them.
[00:03:07] Speaker B: One of the reasons why, I suspect, is that in B2B we have a lot less data and less signals.
What happens to your campaigns when you're in that situation?
[00:03:21] Speaker A: Yeah, I'm being very fortunate. I have a few accounts with huge amount of volume and it's great. And like I said, AI is like absolutely two thumbs up or these cool features, all these core campaign types. The problem is when you're trying to put those kind of strategies into a smaller spend account, smaller data, it's just not built for them, it's just not effective. It will run. Obviously, if anybody's got a smallest vendor account, it's always telling you to change the target roas or you know, target cpa. The problem is we don't have that conversion data. We don't have 30 conversions in 30 days to be able to tell it what to do. So they're being told to do it and it's, you know, it's the best practice and it's the gold standard, but it's just not built for those types of accounts. So it's just frustration and then lack of results. Then that's just the whole cycle. The more frustrating, the worst results you get, the less time you want to spend on it. It just goes round and round until sometimes you just give up on PPC altogether.
[00:04:09] Speaker B: So do you think the, the 30 conversions in 30 days, do you think that, do you think that actually works?
Because I've heard conflicting information about that in the industry where they're saying, no, it doesn't work at all. It can automatically optimize with the AI signals. But then in the case of B2B this, we don't have as many Signals.
[00:04:31] Speaker A: Yeah, I think as well for a lot of B2B accounts is that they're very niche products and services. Right. Some of the ones that I use are so niche. So the general B2B signals, you know, in general may not be relevant for the accounts that I'm working on, but from my personal experience it does need conversion data and the more you have, the better. 30 and 30 is kind of like, kind of like a great level that I see. But you know, some accounts I have will have like 10 conversions a month. They're never really giving it enough information. So it's that kind of disbalance between what we're being told and then like What a niche B2B lead gen account actually sees happening.
[00:05:05] Speaker B: Right.
That leads me to the AI optimization.
A lot of the AI optimization is really geared towards that E commerce model, that straight sort of sales funnel.
You know, how do you work with that in B2B? And do you think there's a workaround? It's a longer sales cycle, it's a lot more complicated.
[00:05:28] Speaker A: Yeah, I think it's a bit of a nightmare really in the platforms themselves, you know, working with like a 90 day conversion window, for example, if it takes for one of my clients was 183 days on average for a lead to turn into a sale and become actual value to the business. So even though you had your CRM set up and you know you have these amazing campaigns and everything was optimized and perfect, after 180 days you've lost that attribution. I've started working with one client using Dream data as a tool which has like a longer period of like a look back window. It's still very early days for me to kind of understand that and how it works and the results that it's having. But that's the main problem that I see is just the disconnect again as you say, between someone buying a pair of trainers for like £20 and then someone signing up for a piece of software which is, you know, 250, 50,000 pounds per year, for example. That sales cycle is so different, it could be so much longer that the platforms can't necessarily tackle with that.
[00:06:23] Speaker B: I've seen a B2B hosting company that their average sale was 180,000. So they basically needed like one conversion per year to be profitable with the Google Ads campaign. So that's something where I don't know if the AI optimization tools are going to help as much there unless you have like secondary conversions that you're using or loading customer data.
[00:06:48] Speaker A: And like I said, if you're doing reporting as well. This has happened to me so many times, you know, reporting from financial to year to financial year. I'll have loads of leads like in the last financial year and I'm trying to prove the value of PPC for the end of this year. And I'm like, yeah, I still have leads that are open. It's still going to take like six months for them to complete. But trust me, there's going to be money, money coming in. It can be so difficult to prove the value when you're working from financial to year or month on month. It's the long game. And like I said, those platforms aren't all able to kind of deal with that and make it work for us.
[00:07:20] Speaker B: You wrote an article for Search Engine Land and it was about ad strength. Not able to predict success with the AI tools that we have right now, obviously they're extremely predictive and I think there's this seems like there's a big gap between what the platforms are recommending and what is realistic with these tools. And if those add strength or settings are really helping. What, how has your opinion changed or has it since you first wrote that article?
[00:07:55] Speaker A: Even in life? Right? The gamification of getting 100% on anything or, you know, your Fitbit, you know, telling you that you've done enough steps for the day. Our brains just want to make us get those numbers. We want to achieve what we're being told to do. I think the problem with them, as I covered in the article, is you can get a hundred percent, but that doesn't mean, you know, 100% Opti score or, you know, excellent ad strength. That doesn't mean that you're going to get more conversions. You know, some of my ads are ranked poor. You know, the more he's getting these flashy things, oh, your ads, poor. But they convert the strongest. It doesn't matter what that score is. There's no, you know, specific correlation. But I think our minds, psychologically, we just want to achieve those gold numbers. We want that gold star. And unfortunately I see a lot of people chasing and accepting, you know, optiscore recommendations, for example, like, yes, yes, yes, I've got 100%. And then they're like, I've had no conversions. I've spent five times what I thought I was going to do. Because we're not always looking at the why or what those changes would mean. We're just chasing that number or that green tick.
[00:08:48] Speaker B: So when you're Trying to increase the score. A lot of times you'll see it will say to increase the score you need to add more keywords or you'll see these very small things or things that just are not appropriate for your ad.
And I think that is a really big disservice to advertisers right now.
[00:09:09] Speaker A: I think, like I said with the keywords, you know, I'll make sure you're putting your keywords into your headlines, for example. That's great. But if that keyword or phrase that you're using doesn't translate well into a headline, then you're just sticking it in to get the score and you've got a headline or an ad which just has so many keywords stuffed into it that it makes no sense or just sounds, you know, robotic or it sounds like spam. So you end up doing yourself a disservice just to kind of take it off. But yeah, the whole keywords and making sure they're in the ad copy, you know, and having a poor score, again, it's just my personal experience is that my worst scoring ads are my strongest performance.
[00:09:42] Speaker B: Right. And of course you have to have some keywords in the headlines, but there are 15.
So how many keywords do you need? I think there's a lot more differentiation when we do write it ourselves. When it's human, human written.
[00:09:56] Speaker A: I love writing ads. I really enjoy it.
[00:09:59] Speaker B: Shifting gears a little bit. I really want to know your opinion about this.
So one of the strategies, the newer strategies in PPC since you've been doing this for 12 years, is the broad match and AI targeting. Right. So this is, this is broadening how we reach possible converters.
How do you see this in B2B campaigns when there's a huge potential for wasted spend?
[00:10:25] Speaker A: Anybody who's seen any of my content knows I'm quite skeptical of changes. And I'm very risk adverse. Like I'm completely risk adverse. But I have tested things like AI Max research and you know, tried the broad match with the automated bid strategies and as you said, the wasted spend is just too huge. Like the ambiguous nature of a lot of B2B keywords. You know, you're getting top of funnel stuff. Yes. Which can be great, but you can stuff which is a completely, completely irrelevant. They're never going to be right for your, you know, product or service. Every time I have done it, I've done it with a limited budget so I can control it and it not take over. And then I do love a search term report. I know people have scripts and like tools that, that go through them and pick them out. But again, I love a manual sitting down and going through a search term report and controlling whatever I can and building a thick for a negative keyword list. I love that side of PPC still. And doing that with Broad Match is the only way I've been able to get enough control that I'm happy to leave them running. But in 99% of the time I end up calling the Broad Match keywords. Every now and then I'll leave one or two on, but not very often.
[00:11:29] Speaker B: I think when a lot of PPC managers are looking at the search query report, they're taking it as a sort of a one to one relationship like this exact keyword. What I like to do is also go through it manually or downloading it and asking AI to find the patterns and trends in it. So instead of going in detail through every keyword that shouldn't be coming through, looking for those broader trends because you can knock out so many by doing that. And I think that does take, it does kind of take a human to look at that because the computer is not going to understand the context of some of these that are coming through otherwise it wouldn't have triggered it in the first place.
[00:12:15] Speaker A: It's that ambiguous nature, right? Of like we kind of can put that human touch and be like maybe they meant this or you know, this actually means that over here in the UK compared to like you know, Ireland for example. But yeah, like I said, if they were that relevant we would have probably bidded on them anyway and we would have had them show in. But yes, I'm not complete. Again, I'm not completely opposed to it. Most of my campaigns that I am running now do have one or two broad match keywords in there. But in terms of kind of like setting everything broad and just letting it run, I've not. That's not been successful for me.
[00:12:46] Speaker B: What have you seen that is working for lead gen in B2B accounts.
[00:12:51] Speaker A: So something that I've struggled with forever with B2B lead gen accounts is get in form of any image assets. Now if you work E commerce, B2C, you've got cool images of, you know, your clothing or you know, your food that you're selling. I've always had a big problem with working with clients who don't have any imagery to give me either for like social content, social ads or you know, even image extensions. But I think AI in that sense has given them like no excuse humanly possible not to create Something that I can use, even like video. Again, like a lot of the times clients say, oh, we don't, we don't, we don't know how to make a video or we haven't got the time to make it fancy and stuff all the tools. Now we've all got really good cameras on our phones. Get your client to go out, do some recordings, stick them in a tool, make them beautiful, make them fancy, and they've got no excuse. Now, that's the best way that I've seen it.
[00:13:40] Speaker B: Right, so images again, you know, if
[00:13:43] Speaker A: you're Ecom B2C, you're probably like, oh, that's boring. But you know, for B2B lead gen, if you go on a lot of their, you know, these websites, they're not necessarily very visual. You know, they don't necessarily have a way of presenting what they're offering. So, you know, yeah, these tools, I'm finding that clients are getting a little bit more creative, which then may eventually lead for them to book a photo shoot, for example, or book a videographer. But they kind of need a bit of proof that it's worth it to start off with. And the tools, the AI creation tools are great to kind of give them that starting point.
[00:14:12] Speaker B: Are you using the image suggestions in Google Ads native in a platform? Are you using external tools or both?
[00:14:20] Speaker A: Some clients do have like AI tools that they use in any way and they, you know, they've got enterprise versions, so they'd prefer to use those I've tried to use in the platform, like obviously in Google Ads. I wouldn't say necessarily that they've been the best, but again, I feel like B2B Legion, you know, in a few years time will be able to make it work for us. I've seen what some E commerce accounts are able to do in the platform and it's absolutely incredible. I'm just waiting for it to get there for us first.
[00:14:45] Speaker B: And some of those images are really challenging. You have a manufacturer, even if you have software. What are you showing? Screenshots. People.
There's always the people.
[00:14:55] Speaker A: Yeah, screenshot was just like.
[00:15:02] Speaker B: So, you know, it's, it's hard to be creative, I think for AI or for us. You know, how you can visually communicate and differentiate some of those B2B clients.
[00:15:14] Speaker A: Yeah, software especially a lot of my clients are software. And it's like if you take a screen grab of what their tool looks like, it's just, it looks too busy. It's just tiny. You know, there's no, there's no value gained from that. So you either have to get it turned into a graphic or just highlight something in particular. So yeah, I think even for them just, you know, getting faces of people is better than nothing. But yeah, you've gotta love the cheesy pointing pictures or everybody shaking hands or in a board meeting, but it's something is better than nothing.
[00:15:44] Speaker B: Speaking of people shaking hands, in your role as shaping the content and communications for Hero Computer and brighten so the agendas. When you're looking at the topics and you're looking at the speakers, I noticed that there's been a lot of changes in the past year on the AI topics. Do you think that generally the industry is getting a little bit of burnout on the AI topics?
[00:16:11] Speaker A: Yeah, I think, you know, personally I would say yes. Coming from a PPC perspective is we've kind of had AI, you know, and automation and stuff forever, right? We've been using it and exposed to it for a longer time now we're in that real acceptance phase of AI. We just get on with it. Whereas in contrast different industries, for example, like SEO or content or pr, it's still quite new to them, it's still quite scary, it's still quite intimidating. They're being told they won't have a job in five years, which, you know, PPC people have been told five years before that. I feel like PPC side, we're kind of used to it. I'd say for some of the other disciplines they're still overly exposed to it and getting to that point of burnout eventually I see as maybe 20, 27, 28, we're all just going to get to that stage where it is just part of everyday account management, you know, client management is going to just be part of the norm. And rather than saying, oh, this talk is on AI, it will be like this talk is on client reporting and it will include AI. AI won't be the focus. It will just slowly start to integrate into everything that we do and everything that we talk about. I think PPC again, probably a little bit ahead in that because we've been using it for a while. It's pretty much in everything that we have to do. And I think SEO is really starting to get there as well. Also from a non personal perspective, we run priority polls to see what our audience are interested in hearing and we are starting to notice that last year AI was at the absolute top. That's what everybody wanted to hear. But we're slowly seeing people now just wanting, okay, we've, we've Done enough of that, looking for more technical talks, more nitty gritty stuff, more soft skills. I think it's starting to balance out and yeah, hopefully like next year and the years ahead, it will just become part and parcel of life.
[00:17:53] Speaker B: We have always had machine learning. So it was a couple of years ago, it was at the Google Marketing Live where they announced that now it was called AI and so I noticed it was, I was confused because I, because it was the same exact features, they just called it something else. They just repackaged the naming which they do all the time. They change the names of things all of the time.
And so we have already had the, you know, we've already had the machine learning which is part, which is part of AI. They just simplified it, the name, but it didn't really simplify how we optimize accounts because they have, they would position it as something where we can just click the button to initiate AI, like AI initiated and it solves all of our problems. But I think that's why, you know, last year, more so in the year before there was this like a lot of confusion as to, well, these are the same features that you're calling something different and then you're adding more of them but they're not working.
[00:19:01] Speaker A: I remember I was in an agency when automated bidding really started to become a thing and at the time we were using scripts and I remember we had like a two hour meeting where we were all arguing whether we should continue to use scripts or we should start to use the automated bidding in the platform. Like that was so scary at the time. And I remember thinking, oh, I won't be changing bids every morning, what will I do with my time? Obviously I've other things we can be doing. But you're right, it's just being repackaged in a different way because we're exposed to it in life in general, on social media and when you go into the shops, I think it does feel a little bit scarier. Hopefully we can show to everybody that you will absolutely still have a job. There's plenty you can do and it's not set and forget it still needs that human interaction, human control to make it work.
[00:19:45] Speaker B: And we haven't received a lot of information on how to manage the AI.
[00:19:50] Speaker A: No, not from the platforms. Right, because the platforms, as you said, they're saying just do it, it's fine, we've got this, turn it out. I think that's, that's where the community comes in, right? That's where, you know, people speaking at events writing blogs, doing LinkedIn posts, doing, you know, everything, you know, podcasts. It's what everybody is doing about how their actual experience is it, what challenges they found, what controls they've found, you know, little tips and advice. That's where you get the benefit from it because the platforms are obviously going to.
[00:20:18] Speaker B: You turn it on and if it doesn't work now, what do you turn it off? Do you try a different one? Do you just put it on rapidly change things? Then it doesn't, then it still doesn't work, then everybody's stressed out.
So yeah, it's not really, it's not a cure all. I feel like they would really get farther ahead faster if they could acknowledge that and give more advice on how to manage the machines. Because we're still in this transitional period, which I think is why probably for the content where some people want to know, how do I manage these AI features?
[00:20:58] Speaker A: Old school, Is that what we own?
[00:21:02] Speaker B: I don't know if I, I'm using old school to refer to manual.
Maybe they're turning back to that. If the AI isn't working, you know, if we're still not hitting our, we're not hitting our goals, the clients are not going to be happy. It doesn't matter how you package up the AI or what buttons you push. So we need to move to a solution to get them there.
And that's where I think the, the stress and the skepticism of these features comes from.
[00:21:32] Speaker A: Yeah, we're not really seeing under the hood, we're being told it works, do it. But how does it work? You know what the intricacies of it again, like, how would you do it old school, how would you do it manually?
A lot of people like me, I just want to know what they're accepting, right? If I'm accepting a recommendation or a tool, like I want to be able to understand it. I don't want to just say, okay, I trust you, especially for B2B lead gen accounts because you know, nine times out of 10 it doesn't work exactly how it does for everybody else. But yeah, I'm starting to use manual bidding again on some campaigns for my clients, which feels really, really old school and really, really old fashioned. But it's about doing what's right for the account rather than just saying that you've, you know, you've got Google's gold star.
[00:22:14] Speaker B: What's the one thing a B2B advertiser should never fully hand over to automation?
[00:22:20] Speaker A: I would say data interpretation tools. Get that data, get the stats, get the Graphs make it look all fancy, make it look great. But that kind of contextual summary and breakdown and plan of what to do next, I think that needs to come from a human.
What's going on with the business, with the account, with the world. If something happens world news wise, your performance could absolutely tank. The tools don't understand that they're just going to be like, oh, she didn't get that many clicks this month. So I think kind of like having that human touch is still really important. And it's also good for you to actually look at the numbers yourself and think about the next month or next quarter ahead rather than just click and send and washing your hands of it.
[00:23:00] Speaker B: I think having the experience as a manager, PPC manager, is invaluable.
I had some results that I was comparing yesterday on a campaign that was Microsoft ads versus Google Ads. The same exact budget, the same exact campaign.
And I had the AI compare them and while Microsoft had more impressions and overall clicks, it wasn't as relevant of traffic, so it was more irrelevant traffic. So the AI determined that the Microsoft ads were working better.
[00:23:38] Speaker A: That's it. Like, if you just accept recommendations and stuff from these tools and platforms, you could just shove more of your budget on Microsoft. You know, a few weeks, few months later, when the client like, we've not had many conversions or we've got spam leads or we're not making the same money, then you've got a lot to answer for because you just trusted something that you were talking about using your brain, get yourself from really messy situations, situations. But it's the knowledge that you had in the context and digging deeper that kind of safeguarded you from, you know, just all of a sudden putting all your money on Microsoft.
[00:24:09] Speaker B: Right. I would have a lot to answer for, wouldn't I?
[00:24:13] Speaker A: We have clients to answer for. We do. The reporting is our invoices which get paid. You know, the responsibility is on us. You know, you can't build these tools and complain that, you know, they gave you bad advice, for example. It doesn't work like that. We are the safeguard.
[00:24:27] Speaker B: Yes. And you can't say, Google told me to do it.
Not an acceptable response.
[00:24:37] Speaker A: Absolutely not. You're like, google told me to do it then. You sound like a child. Right, yeah. We're being paid for our specialism. So like, you know, flex that skills, flex that experience that you've spent a long time, you know, building.
[00:24:47] Speaker B: The other guy did it, I didn't do it. And then what are you there for? Okay, are you ready for the Rapid fire round.
[00:24:53] Speaker A: I think so I'm ready.
[00:24:54] Speaker B: So one would be don't touch it and then 10 means I'll trust. I love it.
Oh, okay. And then give me maybe a sentence if you want about why.
[00:25:05] Speaker A: Okay, sounds good.
[00:25:07] Speaker B: All right, the first one. Smart bidding with target CPA6 if you have sufficient data. Broad match paired with smart bidding Again,
[00:25:16] Speaker A: it's not there for B to B Legion in my opinion.
[00:25:19] Speaker B: Automatically applied recommendations a zero.
[00:25:22] Speaker A: I'm not even going to give it a one. I'm giving it a zero.
[00:25:25] Speaker B: Yeah, I agree. I'd give it a negative.
[00:25:29] Speaker A: I'm breaking the rules.
[00:25:30] Speaker B: Generated headlines and responsive search ads. S3 Add Strength Score as an optimization guide.
[00:25:39] Speaker A: As long as it's guide, I'm going to give it a 2 because sometimes it does have some really good examples in there.
[00:25:43] Speaker B: I like to have the suggestion there might be some golden nuggets, golden wisdom in there that you didn't think of. Google's audience expansion suggestions again for B2B
[00:25:55] Speaker A: lead gen, I'm going to give it a one. Yeah. Control over everything for that demand gen
[00:25:59] Speaker B: campaigns for B2B pipeline.
[00:26:02] Speaker A: Ooh, I'm going to say a 3.5. There's been a lot of new announcements and developments which are, you know, really piquing my interest. Yeah. 3.5. Not quite four yet.
[00:26:11] Speaker B: Okay. AI max for search or from personal experience.
[00:26:15] Speaker A: A three. Oh three is my favorite number here. Can you tell I'm risk adverse?
[00:26:20] Speaker B: I agree. They're not customized and tailored to B2B. That's the biggest issue, isn't it?
[00:26:26] Speaker A: Yeah. I'm so jealous of people that can use them and it works. It's not like I don't like them for a reason. It's I'm just jealous that they're not right for what I need them to do.
[00:26:33] Speaker B: Okay. Google's campaign objective recommendations, I would say a 2.
[00:26:38] Speaker A: Every now and then there's some good suggestions in there but also there's a lot of suggestions which come up which kind of go against what you're trying to do.
[00:26:45] Speaker B: Thank you for going through that rapid fire round with me. I think what I'm getting out of it is a lot of these features are not tailored to B2B in any way. So I think it really just surfaces that there needs to be more work done on that. And if someone is looking to reach out to you, where would they find you?
[00:27:16] Speaker A: Nowadays you're best finding me on LinkedIn. I'm always talking about B2B lead gen, PBC and my dog. So find me on LinkedIn. It's Sophie Logan. L O G A N like Wolverine is how I remind people.
[00:27:29] Speaker B: Excellent. And you know you have a special
[00:27:31] Speaker A: offer for Hero Comp and Brighton SEO San Diego, which takes place in September. We have a 20% discount for tickets for listeners, so we'll provide you with a link. Super easy. You don't even need to remember a code. You just click on it and the discount will be instantly applied.
[00:27:47] Speaker B: We'll make sure that those are in the show notes and the descriptions for the podcast and in the post. 20% is an excellent discount. Thank you so much for offering that. And thank you for being here. And thanks everyone for listening or watching to AI ads and beyond. Be sure to subscribe wherever you get your podcasts and on YouTube to learn more about how AI is impacting performance and strategy for PPC and digital advertising.
[00:28:19] Speaker A: Thank you everyone.
