12 The Challenges and Opportunities of a Google Ads Program - Phillip Mejia Jan 13 2022

Tim Curtis: [00:00:00] Welcome to the Digital Velocity Podcast. I'm Tim Curtis from CohereOne,

Erik Martinez: and I'm Erik Martinez from Blue Tangerine.

Tim Curtis: Today on the show we have Philip Mejia from Blue Tangerine to talk about the challenges and the opportunities of running a modern day Google Ads program. So, Phillip, welcome.

Phillip Mejia: Thank you for having me.

Tim Curtis: Lots been happening obviously in the space related to Google. Before we get started on that, we thought we might get a little bit of a background from you. Why don't you tell us a little bit about yourself and your particular journey?

Phillip Mejia: Yeah. So, my journey to working with Google Ads [00:01:00] and just overall online marketing in general started last century, so back in the 21st century, 20th century.

Tim Curtis: Back in the 1900s.

Phillip Mejia: Yes. Back in the 1900s.

Erik Martinez: Back in the 1900s.

Phillip Mejia: I actually started working in call centers. This was kind of around the same time the internet started to come out. That special thing. Working in the call centers, I got a lot of introduction into customer service, customer behavior, reporting, upselling, all those types of things. Getting to learn that part of a company's business and understanding how they work with customers really helped set the tone for what would later be my journey into online marketing. From the call centers, one of the bigger things that started happening in the late nineties, early two thousands is, as we're working in the call centers, a lot of our bosses at the time would start to say, hey, how do we start to move these customer calls onto the web?

We really want to get them [00:02:00] shifted over there. We don't want them to call in anymore. We've got to shift our money over to that side. This is the wave of the future, and so luckily I was in a position where I got to be part of that journey in understanding how we were going to migrate a lot of that call into kind of a self-service model on the web. Working with some high-tech companies back in the nineties and early two thousands, I got to see that migration over to the websites.

From there, little by little, I started to pick up and understand the buying signals, buying behavior and that translated more into the digital side. So, no longer was I working with people on the phone, but how they worked with the websites, what they found interesting, and then slowly, how do we market to them. The big thing was we want to kind of pull back on our spend with the marketing that we're doing outside of digital, how do we transfer that back over, and so we started to see that migration to websites as well.

Erik Martinez: So, Phillip that's [00:03:00] the work side. Tell us something that nobody outside of our listening audience would know about you.

Phillip Mejia: Gosh. So, I have three big things. I don't know who's going to know me outside of our little group here. I worked for Apple for a number of years. I was part of the candy colored Mac launch, the iMac launch, that they did in the late nineties. Another personal thing that's happened to me, I've made 10 skydiving jumps, two of them solo and survived. The other thing that's probably not as well known, but some people do know, is I played trumpet for about eight years.

Erik Martinez: You know, even I didn't know that and I've known you a long time, so...

Phillip Mejia: I know.

Tim Curtis: I'm okay with everything, but the skydiving.

Erik Martinez: I think the skydiving is cool, and it says something about you.

Tim Curtis: It does, but that's a hard pass here, so...

Erik Martinez: He's also forgetting the time that he decided that he was going to do a triathlon.

Phillip Mejia: I did one and that happened just before the pandemic and I have [00:04:00] not done another one since, but I do plan on testing myself out over the next couple of years.

Tim Curtis: Wow. You're quite the Renaissance man, Phillip.

Erik Martinez: So, Phillip, in episode eight of the Digital Velocity Podcast, we had a nice conversation with Pat Grady from Amazing Ads and it is his opinion that Google is working really hard to "hide" information, automate the whole process so that the decision-making for how a brand runs their ads in the Google Ads platform is really removed down to an algorithm. What's your opinion on that, and if that is the case and what are you seeing to kind of lead you to that conclusion, one? Two, what's the objective here? What are they trying to accomplish?

Phillip Mejia: So, let's start with what they're trying to accomplish. They want world domination. No, I'm kidding. They don't want world domination, not quite yet, but they do want to be able to automate a lot of your [00:05:00] more simple tasks on an ongoing basis. They're doing a number of things to push you in that direction. One, if you have Google Analytics, you've probably seen in your Google Analytics account about Google Analytics 4. So, that's a new analytics package that's going to take over the current version and what that one is going to help do is to help close some of the measurement gaps and more information about behavioral data, the visitors that are coming to your website. So, they're kind of adjusting where you're going to be able to find some of that data.

The other thing that they've done is earlier this year, they changed their keyword match type. Previously, they had this broad plus match that did a great job of bringing in all types of traffic, great conversion information. They've tightened that up a bit. Again, they're wanting to get you used to less information that's being provided to you, but allowing you to see your performance [00:06:00] still, without seeing a lot of the background data. The other piece they've done and that they're introducing is smart bidding. So, now they're taking that whole bidding piece off of your hands as well, and they've introduced a number of different variations.

The idea behind a lot of the stuff that they're introducing is to help automate some of those simple tasks that you would go in on a regular basis and do yourself. The whole idea of automation is to help you with some of your daily tasks, give you more time to work on other things, but you don't want to set it and forget it. That's the piece that people need to really understand is, all of this automation is not to just let you forget about what you've got going on, but to really understand what pieces that you can automate. So, yeah, I do agree with what Pat Grady said in that they are hiding a lot of that information. Again, because they want you to get used to not seeing as much information as before, and to have you start to trust and use the new [00:07:00] systems that they're coming out with.

Erik Martinez: Doesn't it feel a little self-serving? I'm an advertiser on Google. I'm spending a few million dollars a year in the Google Ads platform. You're now trying to push me to automate some of the simpler tasks. In order to accomplish that, you're hiding the information that I would use for decision-making on the performance of those ads, and you've also moved the goalposts in terms of the phrase match types and all those things, then now are allowing more phrases in that we would have maybe excluded to drive up click volumes. What's the issue here? I mean, they're a publicly traded company. They're trying to make money. This is one man's opinion. They're not trying to help me run my business better, or are they?

Phillip Mejia: In one respect, they are. In another respect, they are also trying to protect the user's data. So, what they're saying is, you know, Hey, we've got all of this wonderful [00:08:00] information about our people, but privacy is a big issue. We really don't want to have to supply you with everything that's going on. We'll supply you with the bulk of what we see happening, but depending on the types of searches, the volume of those searches, we want to protect some of that user data. That's the information that they're providing at this time, but again, they're also providing additional tools on top of that. So, that while you may not be able to see all the data available, they do still provide additional tools to help you improve your performance without seeing the additional data. Are you still losing out? Absolutely. I do believe you are still losing out on some of that. There are work arounds to find that missing data, but they're making you dig for it now, and just be a little bit smarter about how you're going about that optimization process.

Erik Martinez: So, what you're telling me, under the guise of privacy from a company that gives you Google Analytics and collects all this data [00:09:00] about individual users, links that up to their Google accounts across platforms, across devices, under the guise of privacy, they're taking information away from us for decision making.

Hey, you know what, if we're in healthcare and there's such a granular search that I can individually identify somebody from that information, totally get that. That makes sense to me, but if I'm selling a statue online, really? Has it come to that? That we've got to protect somebody's information about buying a statue, which when it's aggregated information anyways, I don't really know anything about that individual user, right? Unless I've got all the individual tracking tools, trying to identify those people in the background and you know the story. Tim, I know you're a big privacy guy, so weigh in on that cause I'm confused.

Tim Curtis: So, first off, let me say that I think when you're talking about Google or even Facebook in this [00:10:00] manner and you're kind of focusing on the question at hand, are they hiding the information?

Well, first and foremost, I think it's a problem because both Google and Facebook, perhaps now more than ever, are suffering a crisis of faith. They have a very poor reputation in general. I believe they're significantly underwater in terms of people's belief that they're acting on their best interest. They're being called out daily by increasing numbers of consumer rights advocates, privacy advocates, and just in general people across the aisle. So, both Republicans and Democrats are calling them bad actors. You're starting from a position of a crisis of faith, and so any move that these companies make is going to be more heavily scrutinized.

Do I believe that they're making some of the changes in the interest of other people? I think they're making the modifications in their interest, but I think they're doing so because the market is indicating that they need to batten down some of that information. The number of legislative pieces that continue [00:11:00] to iterate, so you take California, California Consumer Privacy Act, CCPA. Now, we've moved into the California Privacy Rights Act, which goes even further than CCPA in terms of what's highly sensitive personal information versus just personal information.

So, things are getting more granular and ad tech is going to have to respond to that by closing off some of those areas. But you think about Google and the amount of first-party data that Google has access through really from Google Analytics. You know, it's deployed on almost every website, right?

So, they have a lot of information, so for them to act as if they're doing things from a third-party perspective, in the best interest of privacy, it sounds good, but it doesn't really align with what's happening internally. So, I think, in general, when you're talking about a company sitting here trying to make decisions. Yes. It's going to make it much, much more difficult, but that's sort of the fruit that comes from privacy regulation.[00:12:00]

What you had available at your fingertips, you're simply not going to have any more. That's problematic, particularly for the ad tech space, because that is where it has been the Wild West. You have not had the regulation. People have been able to do sort of what they pleased. We've had a lot of data breaches that's really impacted every American time and time again. So, yeah, it's a problem. We're setting in a space where as we move that direction, we're going to have to realize that we're going to have less tools at our disposal, but companies such as Google and Facebook are going to use that cover to make changes that will further cement their business and their profitability. Those are a couple of things that you can for sure count on.

Phillip Mejia: I agree with that a hundred percent. They're trying to appease both sides. You know, they want to appease the consumer. They want to appease the ad tech. So, yes, they are hiding some of that data from us, but, again, they do provide other tools that give you more nuances about what is happening [00:13:00] in your Google Ads world.

Erik Martinez: Pat and I have talked about this topic multiple times and Pat's belief, and, you know, he may listen to this podcast to go, I didn't say that Erik, but Pat and I have talked about this a number of times and the one thing that he keeps saying is look, Google is in business to make money for Google, and by limiting your choices, we're going to drive ad costs up. We're actually going to create more waste for businesses. I hear you, Tim. I understand that the privacy legislation is providing cover for this, but at the same time, what do we say to our clients when we can't manage the campaigns the way we want?

Pat believes that Google's goal is to do away with agencies, and that the individual marketers can go in and click a couple buttons, set up their ad campaigns, and trust [00:14:00] Google. We're going to take care of your ad spend for you. By the way, this is like version 4.0 of this because back in the early 2000s, Google would assign,. Once you reach a certain level of spend, Phillip, you definitely remember this, when they would assign a team to you and a team would come up with recommendations. Hey, do this, and every time we tested it, we'd sit there and go, we're just wasting money, right? We didn't really get much value out of those relationships and over time they have systematized those recommendations in the form of tools and knobs and leavers inside the platform to get basically the same result, but now they have absolute control whether you choose to do it or not.

Correct me if I'm wrong, I have heard you say over the past year or two, that they have turned on features without letting us know. So, they roll something out, it turns on and all of the sudden your ad spend goes shooting up and you're like, what the heck happened until you dig into it and go, oh, they [00:15:00] released this feature and turned it off. Is that still happening in your opinion? Or is it happening less? Are they getting better about telling us what's going on?

Phillip Mejia: You're absolutely right. In fact, there's a whole section within Google now called auto applied, where they have a list of all of the recommendations on updating ads, updating site links, updating tracking information, and they rolled this out and had everything pretty well automated. Everything by default was checked. So, it would automatically get rid of redundant keywords, would get rid of duplicate keywords, would get rid of matches that they didn't feel were proper for the account, and they automated that process and so when they announced it, it had already been in place.

We started looking at the change history and we could see some recommendations that were automatically applied to the account, found out, oh yeah, we rolled this out. This is a new feature. We were able to go in [00:16:00] and make adjustments on what we want them to automatically update and not update, but that does continue to happen.

In fact, just a couple of days ago or last week, that auto apply recommendation was updated with new auto apply recommendation features that had been added and old ones had been removed. So, they're constantly updating a lot of this stuff, and this is part of that whole AI and automation and wanting them to be trusted by you in that, you know, a lot of these automation pieces are, Hey, let us update your ads for you. Let us update your keywords for you. At some point in time, they're going to come back with, let us update your product feed for you. So, yeah, this is part of their process. This is part of their, we want you to automate as much as possible. We want to take care of it for you.

Tim Curtis: It does seem as if the end goal is more control over that process, which means effectively marginalizing the agency [00:17:00] component that's been a very, very significant part of the ad tech ecosystem and by setting up pure automation, it really puts a brand in the position that they have to put all of their trust in Google.

Currently, if you have an ad partner who's managing those ads for you, obviously, it's more than just Google, right? But in this instance for a brand, I think it's very, very important, almost mission critica,l that they understand, they go into this eyes wide open that the black box that was Google is simply going to get darker and darker. You're going to have less insight into what happens. More of that information is going to be hard to get, where it's going to take an inordinate amount of effort to get the same amount of information until that loophole is closed. Everybody has to be clear, that's the direction it's headed.

Erik Martinez: And here's where it's dangerous. I had a comment from a client who we were working on taking over [00:18:00] an existing ad account. Their current agency, one man show was retiring, he said, hey, I recommend Blue Tangerine takes over the account. So, we sent over the statement of work, the marketing manager hands it off to one of their contracts people and the comment comes back from the contract person of, well, these ad agencies just charge way too much for this stuff. I mean, I can watch some YouTube videos and learn how to do my advertising.

What I'm hearing you say, Phillip, is we are moving into a world where Google is saying, yeah, hey, advertiser, come on in, do your best. Just give us your budget, we'll take care of it for you. You don't need to have any experience or expertise, but what I'm really hearing you say, it's even more important to have deep expertise and understanding of what's going on if one of those automations gets [00:19:00] turned on and it is not beneficial to your brand, that there's an expert looking at that and saying, no, that's not something that we want to turn on. Let's turn it off. Let's manage it and tweak it and advocate for the brands that we're working with. Is that a fair statement?

Phillip Mejia: That absolutely is a fair statement. There's so many examples where Google has been really trying to take over and automate and take care of the tasks for you. One piece that I haven't mentioned is on the advertising side, originally there was text ads, then it changed into enhanced text ads, and now it's become a responsive ad, and even when you're going in to create that responsive ad Google is pre-filling with things that it wants you to accept. Again, because it's trying to learn what it's presenting to you and whether or not you want to click on it.

It's taking a look at all the keywords that you're bidding on. It's taking a look at your website. It's trying to take into consideration a [00:20:00] lot of different data points and it wants your information, it wants your behavior as you're putting that in there cause now it's learning. Oh, okay, so, this company, or this account is looking for this type of stuff, so we're going to continue to present that until we get it right and they say, hey, they got it. Let's go ahead and accept. They want to try and get as much acceptance as possible. That's part of this whole AI. It's just learning everything you do. Every click, you take every reason where you say no, I'm not going to accept that recommendation. It learns what is getting you to say no and what is getting you to say yes. So, that is a big piece of it.

Erik Martinez: And what I really hear you saying is on December 14, 2021 Skynet went live.

Phillip Mejia: Yes, exactly. Exactly. Exactly.

Erik Martinez: In mountain view, California. Okay. So, if you look at the changes that Google has made over the past 12 to 18 months, what are the most [00:21:00] significant changes that they've made and why should we care about them?

Phillip Mejia: I think I covered a number of them earlier, but the whole changes to the keyword match type is a big one. That was one where they got rid of the broad plus matches. They tightened up more of the phrase match type and then made some additional changes to the broad match, but they put a lot more emphasis on exact match because the more exact matches you have in your account, again, it's understanding exactly what it is you're looking for and it's been giving more weight to the exact match over some of the other match types. They're leading you down the path because they want to know exactly what it is that you're looking for across that. So, that's a big change. Again, they're taking away your ability to see what queries you're matching. They want to know what you're thinking that it should be matched, what it is that you want them to match against. So, that is a learning piece for them.[00:22:00]

The other piece of that is smart bidding. They want to know, are you looking to hit a certain value, conversion value that you need to hit? Do you want to hit a very specific return on ad spend? Is there a cost per acquisition that you're looking to hit? They're doing that because they don't want you to pay attention to the CPC cost. So, a lot of the smart bidding is, Hey, we think this person's in the buying mood right now. So, normally where you would spend maybe 75 cents a click, because we think the buying signals are there, we're going to charge you $5 for this click because we think that's going to convert.

Interestingly enough, there are times where they have a higher cost per click for a search term and it does convert and they're thinking, yeah, this is great. This is what we want to see. We want to be able to charge you higher cost per click for these conversions and their system is learning what keyword at what time may convert, so that they can charge [00:23:00] you a premium on that click. So, that's one thing that they're trying to take out of your control.

Erik Martinez: So, I'm going to play devil's advocate on that one. They're also trying to create value...

Phillip Mejia: Yes.

Erik Martinez: ...in that same process, right? Because who here wouldn't pay a premium if they could get a guaranteed conversion? We'd all pay something, right?

Phillip Mejia: Yes.

Erik Martinez: The threshold, what that something is, is subjective to each brand and their values and their financial metrics and all the details, but that is in a certain way, if you look at it, a way of trying to create value and create value for them, right, cause it's more efficient in that scenario.

So, I remember having a conversation with your team a couple of weeks ago, and we were talking about isolation strategies and how the smart bidding is now replacing these isolation strategies. Can you delve a little bit [00:24:00] into that? You probably need to define what an isolation strategy is and give an example and why it's important to understand why smart bidding might be a better option than using isolation strategies.

Phillip Mejia: So, with the isolation strategies within shopping campaigns, what ends up happening is shopping campaigns offer three levels of shopping. So, you can have a low type of shopping campaign, a medium shopping campaign, and a high shopping campaign. Within those campaigns, you have the ability to kind of direct where that traffic and how that traffic lands on your shopping campaigns.

So, if you wanted to cut out a certain group of searchers, you could say, you know, if someone's just looking for a greeting card versus someone looking for a vet greeting card, you could say, hey, I don't want to find anybody in my searches that comes through with [00:25:00] just greeting card, and so you're saying anybody who does those searches, they aren't the ones that were interested in. We really want somebody who's got vet greeting card in there. And so you add these negatives to the different campaigns to try and funnel them down so that they can end up purchasing for you. You cut out quite a bit of traffic. You really tighten up who you're going after, and you can see some very good results, especially when you're in a very niche market that may bring in a lot of unqualified traffic. That's the ISO strategy that you can create with some of your campaigns?

The problem currently is, as we mentioned earlier, in the podcast, that Google is hiding more and more of those terms. So, you don't have as much control over it. Now enter smart bidding and now Google's looking at everything in your account. It's looking at your search campaigns. It's looking at your dynamic campaigns. It's looking at your account, [00:26:00] negatives, your campaign negatives. It's taking all of that data into account in your smart bidding campaigns.

So, now it's saying, hey, we know based off of what you've done historically on this account, what this campaign should be and so it goes out and it tries to mimic your best performing campaign, knowing what your negatives are and what other parameters that you've set in the other campaigns. In some cases, I've seen smart bidding do exceptionally well with certain campaigns. In others, not quite so good.

Erik Martinez: When you see a smart bidding campaign go off the rails, are there specific characteristics of when it goes off the rails or is it just simply there's just not enough information? That's what I got from what you just said. There's not enough information for the smart bidding campaign to properly optimize.

Phillip Mejia: That is correct. I mean, you're going to start to see a number of different things because the smart bidding is [00:27:00] going to start suddenly looking at, oh my gosh, how do I optimize? It will almost act like a human. It's going to let's raise the bar. Let's see if we can get more traffic. You'll start to see more traffic come in. You'll start to see your click-through rate go down because it's looking for the right type of traffic. Unless you have a good budget set, it's going to continue to spend and max out that budget every single time it's running on a daily basis, but it starts to get erratic in terms of its behavior because it hasn't found enough information to get what it needs, and so it's going to go out and find that information for itself whether it's good or not. Over time, you'll start to see that return on ad spend just tick downward because it is looking for anything it can latch onto, and it could occur over a month, it could happen in a week. You just start to see that behavior and that performance just tick downward [00:28:00] on a regular basis.

Tim Curtis: Sort of tangential to that, I want to ask a question about AI. When you're talking to clients, oftentimes, and they're discussing, their ad spend strategies or where they're headed, there's sort of this belief, and this is really more marketing than anything, that if you turn it over to AI, it will automatically improve. So, a number of agencies have started leveraging some AI in their bidding process, similar to smart bidding. What do you see oftentimes with that AI technology?

You know, those of us who have worked, and I've got over 20 something years in experience in dealing with these type of algorithms, these type of models, what are you seeing when clients begin to deploy that because what I've seen is it's not always beneficial?

Phillip Mejia: Exactly. So, there've been some clients where they just pushed everything through smart bidding and initially seeing great results, but then slowly coming down over time. [00:29:00] One of the things that we try to do here is we try to step into it. So, we don't launch everything under smart bidding. We take a look at what's a great performing subcategory and what's an underperforming sub category. Let's see what we can do to help improve both of them. So, we may take a smaller segment of a shopping campaign and just focus on that and see can the AI improve the overall performance? How long does it take? What's the cost associated with it? Then finally, what are the end results? Did we see a much better improvement over what we've done manually? In some cases, yeah, we've actually seen some good results, but the goal behind those results is what's the long-term effect? Can it sustain itself without additional intervention from a human? Right now, myself personally, I have not seen a sustainability there.

Seasonality might come into play. That's one of the things about AI. [00:30:00] It takes it a lot longer to learn your seasonalities. If you start AI during the holiday season, all the rest of the year it's going to try and replicate that success, and I can guarantee you for very fourth quarter business, it won't get there until the following year because that's the time it learned, that's the time it's trying to understand that part of the business. Conversely, something that starts Earlier in the season, as it's learning throughout the season, you may see better performance from it because it's starting from a place where it's really learning your business from that seasonal standpoint and that beginning standpoint.

So, again, it really depends on when you start to use it and what you do with it when you start to use it. Again, like I said, I'm not an all in kind of guy. I like to tip toe into the stuff that I do. I haven't gotten to a point yet with AI where I feel comfortable enough like, yeah, let's just go ahead and launch all of these campaigns and let it go, especially on the Ecommerce side. With Ecommerce, it's more about what you spend and what you [00:31:00] get on your return versus kind of a lead client. That one, we tend to see a little bit more success. Again, there's not as many moving parts or levers.

Say for instance, if you work with a client that has over 5,000 products versus a client that is looking for 10 to 15 leads a month, there's not as much to work with. So you tend to see a little bit better performance on that lead side.

Tim Curtis: Many clients don't realize when a client does a lot of testing and they work with a partner, whether it's Blue Tangerine or CohereOne to analyze what's happening in their fourth quarter. A lot of businesses are fourth quarter businesses, so let's just make the assumption that we're talking about a fourth quarter business here. What's interesting about fourth quarter businesses, is that when you're really focusing on your acquisition during that fourth quarter, what tends to be the rule of thumb is that customers acquired during the fourth quarter, generally have a lower lifetime value than those who are acquired during the off season.[00:32:00] Even marginal customers at best, are motivated during fourth quarter to make a purchase, that purchase does not translate into additional purchases throughout the year.

So, that strategy, if you're deploying AI, you know, one consideration that you might have there is that you may be shaping an AI, because it really is going to be input output, if you are shaping that AI to look for customers that include that marginal customer, those attributes, you're going to replicate that in the off season and not even as effectively then. So, you really are lowering the watermark considerably throughout the rest of the year. That being said, obviously, the first and foremost, just like with any channel you test. You test, test, and retest. Just like you talked about Phillip with, you know, putting your toes in the water there to really see are we getting the return that we want? Look and consider testing for the long game like I mentioned, the lifetime value. You should understand the lifetime value of those [00:33:00] channels before you begin tweaking a lot of these campaigns, and I would dare say that the vast majority of clients do not understand the long-term lifetime value implications. So, that being said, if someone says we're going to test into AI, where would you tell them to begin those tests?

Phillip Mejia: If they are an Ecommerce client, I would say, start with your shopping campaigns because within your shopping campaigns, there's a couple of things that happen, whether it's a smart shopping campaign or a manual shopping campaign, you're going to see a great deal of new visitors to your site. Shopping casts a larger net, so you tend to get a lot more visitors because of it, but that would be the place I would start. If you've got a shopping campaign, hopefully, within your feed, you've got some good breakouts of your data. You've got some good understanding of your categories, your subcategories, [00:34:00] so that you can really take a look at what to test. For me, that would be the quickest and easiest way to understand, do I have some automation opportunities here? That to me is the place to start.

Tim Curtis: Would you say that then of all the things that would be important for a client to begin, would that be really focusing on a shopping, for example, as a critical component for someone to begin in terms of a campaign.

Phillip Mejia: Yes. Yes. With each one of those products, you already have built-in search queries with each of the product titles, so you don't have to worry about creating any campaigns. You don't have to worry about creating those ads. A lot of that is part of that automation and so it'll kick off pretty quickly. You'll get data very, very quickly in terms of what's working and what's not working. Again, I would do a head to head. Do something that you know has got great performance and something that you know that doesn't, [00:35:00] and let them run together to see what the AI does for both of them.

Tim Curtis: When you're dealing with shopping, couple things that I think would be, I guess I could say important to understand. Number one is shopping does have critical mass. You are able to really reach out and grab, you know, larger audiences from the shopping side. That being said, of course, you go back and you do understand that shopping does not always have the highest lifetime value. So with that, there's a couple of strategies you really need to think about deploying. Number one, consider ranking products based upon their appeal and based upon their margin, so that you have an understanding of what's an A plus type product that will probably do well on shopping. If you have a product, for example, that has moderate appeal, but man, it's a razor thin margin, shopping may not be the best channel for you to utilize. But oftentimes the entire kitchen sink is loaded into Google shopping and I think you really have to [00:36:00] think about your strategies in terms of tiers of products that you can deploy. Is that something that you would recommend or do you see that often?

Phillip Mejia: So, you hit on a very, very good point, which is understanding your product's margins. You know, a lot of times people don't necessarily take into consideration or they just give an overarching kind of margin number, but if you really want to be effective in your shopping campaigns, then you really want to tier everything out based off of great performers, not so great performers, products with exceptionally good margins, and things with razor thin margins because in some cases you may find that those products with razor thin margins, big traffic drivers for you. In some cases, you may make as much as you spend, but the volume of orders and the number of eyes that get to your website because of that product [00:37:00] may have some additional benefits outside of that return on ad spend for it.

Tim Curtis: Especially if you're deploying the second strategy I would say is key, is if you're deploying a second purchase nuture campaign of some kind, you're really trying to get that one-time purchaser that's coming from Google shopping, if you're then trying to get them on a track to really boost conversion for second time purchase. We all know that the trajectory of a customer who makes a second purchase over those that make just that initial purchase is radically different, so understanding that and building strategies and tactics around that too, to boost those conversions is obviously very important as well.

Phillip Mejia: Absolutely.

Erik Martinez: I think one of the key things that I'm hearing and you guys are saying it, but not necessarily, specifically, this way, segment your shopping campaigns. Segment your ad campaigns. Do it granularly, so that you can have deep insights into what's [00:38:00] working and what's not, and why it's working and why it may not be working, so you can continually optimize. So, even with this world of AI and automation, the one thing that's come clear to me is... I attended a conference over the summer, and I had an opportunity to listen to Paul Roetzer speak, and Paul is the head of the Marketing AI Institute and he has five levels of AI, and for the vast majority of AI that's deployed today, we're still at level one.

And level one is basically the idea of automating simple, simple tasks. Everyone talks about machine learning, but the machine is learning in level one is how to automate that task and make it better,t make it more efficient. Right? We haven't really moved past this level even in our marketing ad tech today. It's coming, [00:39:00] it's coming. There are better algorithms that are now able to associate those small little intricacies and the data that if we were trying to do it, we'd have to deploy whole bands of teams and lots and lots of computing power to figure it out, right? The technology is getting to that point. Big companies like Google and Facebook are deploying these types of technologies and making their algorithms better at identifying those small connections.

Phillip you've mentioned today, several examples where smart bidding technology, which is moving that direction, hasn't really achieved that promise yet, so let's not throw all our dollars at the AI shiny object because we're just not there yet. It's coming. It's coming and we need to continuously test using our ad dollars to figure out when you get that level of efficiency and gain and performance that, you know, [00:40:00] when that happens, so you can take advantage of it when it does. So, Phillip, you've been doing this a really long time. Everybody who I've ever talked to who works on Google Ads are Bing Ads, or has been in this space, has their thing that they love about the platform. The thing that it's my go-to secret of how I obtain performance for my client, for myself, whatever it is. What's yours? What's your go-to method or process that really sets what you do apart from guys just like you who have been doing this, or gals, who have been doing this for a really long time?

Phillip Mejia: So, for me, my big thing, and the one thing that I notice a lot of people aren't really tracking as much is the whole free shopping piece within Google Merchant Center. Google introduced, I want to say a year and a half ago, maybe two years ago, the concept of free shopping. It's something they had a long time ago under the name of Frugal. I don't know if you remember that in the early two thousands, but they kind [00:41:00] of reintroduced this as the pandemic started to increase and that's one area that I think is underutilized by a lot of people. If you have products that are constantly getting tons of clicks, always showing up towards the top, but we don't seem to convert on them, don't be afraid of moving that over to the free shopping side. If it's already getting a lot of attention through paid, most likely, it'll still continue to get that attention under the free shopping but move it into the free shopping and start to watch that performance. Can you improve on that? We've got a number of clients where we continue to see that free shopping grow month over month. That gives us more time to allocate some of those Google dollars into the campaigns that we know are doing exceptionally well. Keep growing that return on ad spend and kind of moving away from some of that noise that might be impacting some of those shopping campaigns, but free shopping is something that I think that is [00:42:00] probably the most underutilized right now.

I'm not going to get into how I migrate stuff. Yeah.

Erik Martinez: And nobody wants to sit through that discussion. Trust me. He can talk for hours and hours.

Tim Curtis: Well, yeah, I think, definitely, we can save the hours and hours for drinks sometime. So, I want to ask sort of a last question here in the wrap-up, kind of leaning into that experience you have. You've been there done that. You're continuing to manage these campaigns at a very intimate and intricate level. What is your one piece of advice to listeners who are either managing those campaigns themselves, or they're an executive who is overseeing the staff or the agency managing that? What's your one piece of advice for them?

Phillip Mejia: Wow. I've got a number of pieces of advice, but probably the one that's kind of reared up more recently than not is whenever a client is giving you a goal. Let's say, for instance, they say, hey, as long as [00:43:00] our return on ad spend, is this number, you know, we're going to be profitable.

One of the things I think that a lot of agencies and people who work on accounts and they've been given goals, they need to go back to that client and say, how did you get to this number? What are the metrics that you're bringing in to say this is the number you need to hit? Because over time, we're starting to realize that people just put together kind of a high-level number, and as we dig into the detail of the data and we really look at what products have been selling and what has been working for us, that return on ad spend goal should be higher than what they're thinking it should be. Take a look at return on ad spend goals and really understand if you are making as much profit as you think.

Erik Martinez: That's a fantastic piece of advice cause I've always mentally done that calculation of taking some kind of contribution formula. Folks, if you don't have a contribution formula in your marketing department, you [00:44:00] need one, and that contribution formula has to take into account margin, advertising, fulfillment costs. It has to take in all of those things, so that you can know whether you're breaking even, or making money on your advertising, and it's okay if you are deficit spending, as long as you understand your lifetime value. So, I think that's a fantastic piece of advice. I know Tim and I talk about lifetime value all the time.

Tim Curtis: All the time.

Erik Martinez: Because that's the trigger that lets you figure out how much can I actually invest to grow my business over a certain period of time, and that's what we're all trying to do. So, Phillip, thank you so much. We want to be respectful of your time. If somebody wanted to reach out and talk to you, what's the best way to reach you?

Phillip Mejia: They can call, no, I'm just joking. They can reach me via email, phillip.mejia@ bluetangerine.com. It's Phillip with two [00:45:00] L's.

Erik Martinez: How do you spell Mejia?

Phillip Mejia: Mejia is M as in Mary, E as in Edward, J as in John, I, and then the letter A.

Erik Martinez: Phillip, thank you so much for sharing your experience and some time with our audience today. We really appreciate it. Thank you so much for listening to this episode of the Digital Velocity Podcast. I'm Erik Martinez from Blue Tangerine,

Tim Curtis: and I'm Tim Curtis from CohereOne.