The Best Way to Set Shopping Campaigns and Optimize Them – All Things PPC Podcast Ep3

We would always recommend only have a product active in one campaign at a time, which is a pain.

Okay. So I am here today with Rory, who is one of our senior account managers. I’m very experienced for our e commerce clients. How are you doing Rory? Yeah, good. Thanks. How are you doing? I’m good. Thank you. So I wanted to bring you on today to do. A semi deep dive on getting into some of the specifics of what to set up with shopping campaigns and optimize them in the best way, because I think, I don’t know what you think about this, but I think too many people think you just sort of whack all your products in a shopping campaign and then just leave it to it, especially with performance max.

But it’s definitely more complicated than that, and there’s definitely a lot of things we can get into. So, when it comes to setting up a shopping campaign, where do you normally start? Is it a case of putting everything in one campaign, or what’s the first question you’re asking of a client? If you’re starting from fresh You need to gather some data to get some idea of what works and what doesn’t you probably already have some idea of what works and what doesn’t just from having your website set up.

What about existing accounts that we take over that are underperforming, for example? So you’ve got to look at what, what is the reason for it not forming. Is it because too much money is going on products that don’t work? Is it because it’s going into avenues that don’t work? Generally find when we run audits for people that they’re spending a lot of money on products that aren’t even getting sales through.

So it’s about splitting that off from most people start with. A catch all here’s everything in one and just let it run exactly like you said, and it’s about, okay, so what doesn’t work? Let’s you either exclude that if you just don’t want to advertise it anymore, but I know people have stock. So split that off into its own campaign, manage it separately so that it’s not wasting all of your money.

Yeah, but it might. Work those low value products or low performance products might get something if they’re just left on like a low budget, um, to kind of run separately. How would you start that process? Because something that’s always eye opening for me when we, we take on a new client or we do an audit of an account is that e commerce managers or digital marketing managers, they just don’t understand which of their products are even selling.

So what are you specifically doing to start analyzing? You know, at a top level where the performance gains and where the wastage is. I mean, there’s a bit of nuance with it. So if you’re a seasonal type of industry, you can’t just look at like the last 30 or 90 days data and get an idea of what works.

You got to look at what was working last year and get some idea of that. If you’ve got. You know, we, we’ve got some accounts with tens of thousands of products where that just isn’t feasible to look at everything in my new details. So instead of looking at individual products, you might be looking at product categories, which categories don’t work.

Yeah. What do you think? Well, I mean, I think one of the first things I like to look at is. I think you’re right about the seasonality on products. And then we always, we don’t always know about stock levels and things like that. But just looking at which products have even had impressions in Google Ads and then which conversions, because there’s often two things.

There’ll be a good portion of the products when you sort by impressions. That just haven’t had any, and then you’ve got a good portion of, well, a massive portion of the products that haven’t had any conversions. So I think there’s two questions in there. One, let’s tackle one at a time. So if you’ve got a set of, say, let’s keep it simple, a thousand products and say 40 percent of them aren’t even showing, what are some of the things you need to start looking for there to understand why those products might not be appearing at a high level?

So sometimes it’s because of how the products are set up in, in the feeds, the kind of basic product data. If Google can’t understand what your product is enough to sell it, then it’s not going to show it very often. But also, if you’re throwing a thousand products into a campaign and expecting, uh, that to all, then to all get even distribution of the budget, that’s not how it works.

Um, Google is going to focus on, well, we’ll test out everything, but then after a while it’ll go, okay, these things work. So I’ll prioritize putting the budget there. And then sometimes products just don’t get seen for quite a long time. Um, we see that quite a lot with some of our accounts and generally it’s a good, uh, kind of sanity check to do when you’re managing an account and have been running it for quite a while to do let’s look at what’s not.

Not even had any impressions and can we do like a product push on those? So kind of cut it from the main campaign put it out into a separate campaign Will those products get any traffic and if they’re still not getting any traffic in there Then there is some kind of problem with your product data Yeah, so I think that’s a good that’s a good first point of if if you were to start with one campaign That’s where we start to have some Some structure, okay, you’ve got a subset that aren’t getting any impressions.

That’s where you’d have a different structure in your account, wouldn’t you? You’d split those out into different areas. Yeah. What other classic kind of structural setups will you go for in a typical account? Feel free to use an example of one of the industries we work with. So, for example, we work with a flooring company that does different types of flooring, don’t they?

So, how do we segment those out? There’s different factors there. So particularly nowadays, there’s a kind of balance that you have to uphold with, I want fewer campaigns for the reason of you’re putting more of your products together. So you’re collecting data better. So that the algorithms can work faster so that your campaigns can work better.

But then also, can you make those, um, more specific to the product? If you’ve got, um, products with different profit margins, then that’s a classic example of. You know, if, if your main products have a 10 percent margin and these have a 20 percent margin, okay, let’s, let’s put those out into a separate one and manage them different differently.

Yeah. So we were, we would have maybe different KPIs for those different campaigns, wouldn’t they? Yeah. Yeah. Something that comes across quite a lot is splitting out by product value and that can work, but generally I find. Um, with, with lots of our accounts in the furniture industry, for example, where you’ve got quite a big range of, um, you know, some products that are thousands of pounds and some that are dozens.

That if you split off your high value products into a different campaign, there’s not enough data that’s going to run through that campaign for it to get any traction on your high value product. You split off the low value products and generally that works quite well. But what you want to look at is what’s the performance like for the different pricing that you’ve got, what, where are the most sales coming through?

Because that’s the bulk of your, your good data. And go, right, I’ll, I’ll have those campaign, uh, those products in, in this main campaign and the stuff that’s higher value than that. So you put in all of your, your high value products that you want to sell with all the good data so that the, the bidding algorithm is going to, um, be able to sell more of the high value products.

Yeah, it’s something we’ve talked about on other podcasts as well that, um. A sort of perverse point when you’re running a PPC account is the more keywords you have or the more products you have with the same budget, the thinner you’re getting spread. So it takes longer to get any traction and data of statistical significance on what’s working and what’s not.

So when you’ve got a client that says, we’ve got a pallet of this in the warehouse and we’ve, you know, we need to get it out, how can we push? Push those kind of products more forward that aren’t getting the traction. How can we give them? Do we supplement them with search campaigns? Do we do discovery campaigns?

What options are there? Uh, you could do search campaigns. I generally don’t find it work as well. Particularly. I mean, if you’re using performance max, you kind of already getting a good holistic, your search, your video, your display. If they work, um, they’ll put more money into it. But if you’ve got like a small set of products that you just want to get out the door.

Um, and you’re not going to have time to build up data and go through all those processes. Um, then, yeah, it’s probably best to just stick them either in with that kind of the main campaign and they might, um, sell over time or just as I was saying before, like a product push type campaign where it’s on its own.

The budget is only going to go through those products. A lot of people are using stuff like Target ROAS, um, as a bidding strategy on campaigns that are spending maybe three, four, five pounds a day. And that is not giving the algorithm, it’s not giving the bidding strategy enough data or enough throughput to adequately try and target that target ROAS.

So instead you’re better off using something like maximized conversion value and just keeping an eye on the budget levels. The target is. More like handrails for a higher budget campaign, right? Yeah, it’s something definitely we’ve seen taking over accounts when people think they just set the target ROAS as if it’s some magic number.

I guess I think that is part of the problem with how Google’s labeled it. It’s kind of people hear the word target and think that’s an actual don’t they? But you have to use common sense, especially for low volume products that how to get the bidding algorithm to work in your favor. Yeah, the common number that I’ve seen kicked around is to use target.

You want at least 15 conversions a month. Right. For that one campaign, that’s not across your entire account. It’s not going to, the learning from one campaign is not going to go into the other campaign. Yeah, yeah, I understand. So yeah, and that’s about gathering enough data. So what, what about when it’s the flip side?

So we’ve talked about if people just have everything in one campaign, and then there’s a few scenarios where you might want to split them out, undersellers, bestsellers, margin difference, you know, there’s. There’s a lot of nuance there before we talk about analyzing data. What happens if you’ve got multiple campaigns with the same product?

Cause I feel like we’re seeing this more often. I have a theory why, but I’d like to hear what you think and the impacts of it. So number one, if you’ve got a product in more than one campaign, you’re segmenting the data of how Google learns to advertise that product. Hmm. You’re, you’re splitting up what search terms are good, what, um, uses it once or all of the performance data, which is going to make it harder for it to learn how to sell that product and what type of user is going to be interested in that product, which is going to make the performance of that product worse.

So, we would always recommend. only have a product active in one campaign at a time, which is a pain, especially if you’ve got thousands of products. Um, when you’re trying to segment out different products individually, but it is worth doing for the performance benefit. It’s okay. So if I play devil’s advocate of someone who doesn’t know as much about shopping campaign setups as you, you might say, well, if I’ve got the Google’s only going to show one.

one at a time. I feel like Google will just show the most expensive version. Is it just for the data or is there some, is there a risk it might not show at all or how does Google get confused? Well, the other potential risk is that it could be registering. Both products, uh, in the same ad auction, which means you might be raising your own cost per click by advertising the product in two different campaigns.

Yeah, and then it’s supply and demand and Google will just say, thanks very much. I’ll have the highest cost per click, which in turn limits performance further down. I’ve got a theory that the reason we’re seeing this more in the accounts we audit is because Performance max has been a thing for a few years now, and it feels like people are just attached to feed to their performance max campaign, left it for once and then come back and set up a sub campaign because they’ve read some tips somewhere and they forget that.

They’re then duplicating products. So that’s, that’s my theory of why it’s happening. And we’re actually seeing it more. It used to be more on the search side. You’d have, and we still get that problem where you have multiple keywords, the same keywords in multiple campaigns. But that, that’s my theory of it’s kind of almost too easy to do on the shopping.

Yeah. That does seem to be a, an idea that performance max is this, uh, magic wallet. Um, I mean, it. It is very good. It is very useful, but it also needs its own management. And I think because part of the, uh, the intricacies of managing performance max is that Google don’t show you very much data to kind of feedback what type of users, uh, are advertising or what they’re searching for.

They’re a little bit more, it’s kind of opening up a little bit. Um, so performance max took over smart shopping. In Smart Shopping, you could not see what your search terms were. Now in Performance Max, you can see what your search terms are, sort of. You can see a big, here’s the categories of searches.

They’re still holding some of that back because there’s personal information there and privacy information. And that’s, that’s a whole different, there’s a whole different podcast. So when you’re, I guess, what, what, aside from not saying everything you might as a, as a Google ads marketer want to say, what other frustrations do you have with shopping campaigns?

Is that even stuff within Google ads? Is it more the feeds and stuff? What are the classic things that hold? You, you back to get the best results. The main thing is dealing with, I think I talk more now than I ever have about demand. It used to be more, uh, simplified in, in how you manage things where you could just try and push things harder with bidding on individual products.

Whereas now it is more about where’s the demand for the product and that’s going to dictate. How often it’s seen, how much performance you’re going to get out of it, what kind of sales you’re going to get out of it, and then that kind of loops in and yeah, it is frustrating, especially, um, I mean, it happens sometimes where you feel like you feel like you want to do more, um, to try and benefit the performance and you’re kind of stuck against this barrier of, You know, the, the performance was worse on, on this weekend.

So, um, I guess that’s mostly from demand dropping down for the weekend. Yeah. You’ve, you’ve always got to overlay the demand that you see in BBC with all the channels as well, I think. Often people will say, if there’s a dip in PPC, they’ll say, well, why isn’t it spending? But when you start looking at organic over time, or direct, there might be other factors at play there.

Yeah. And it’s not just one element. On the flip side, if your budgets aren’t managed properly, or you could, you know, we’re taking over an account that’s not been tracking things or double tracking. It does take time to, to get things back on an even keel, doesn’t it? Yeah, that’s, that’s definitely a frustration and we’ve seen that happen before with multiple accounts.

That’s a kind of, if you take over, uh, a new account and there’s something like double tracking of conversions, then with, with all, uh, Um, the best will in the world that we can to, to manage the account as best we can. The algorithm is still, still dealing with historical data that’s inaccurate and trying to, it’s looking back at that time and going, what was working so well, then why isn’t it now?

Yeah, and it’s unfortunately, people don’t like to hear it, but it’s like ripping the plaster off. You’ve just got to take the pain of getting the tracking right, getting it set up right, structuring the campaigns. Correctly. Um, outside of the Google Ads account, but still within the Google ecosystem then is Google Merchant Center.

What’s the main issues you see in Google Merchant Center when you’re checking client accounts that people trip over? The main one is probably GTINs. Um, whether you have access to them or not, um, it’s just, it just seems to be a constant recurring thing of products are missing GTINs or you flagged that there is no GTIN for this product, but Google says there is, um, which is just a recurring, everyone has that same issue.

There are, there are some ways around it, but it doesn’t always work, doesn’t it? You can set the flag to false and things like that, but like everything that we do, the better the quality, the data we get in. The better the quality, the results that come out. And that’s about as much as having all your products in there, it’s having as much data within those products.

Listed in the correct way, good quality images, products actually worth selling, sold at the right price against competitors, and things like that. The more complete that is, the better you’re going to get the performance of your ads account. And then when you’re analyzing it, you can, you can see a true point of where you can portion and get better results.

Yeah, I do think, um, probably more than it ever has been. We do more. Analysis of why things are better, why things are going wrong. And ever before, um, I think the more automation helps, the more we have, the more time we have to spend, um, analyzing how the automation works and why it’s done things that it’s done on picking it.

Yeah, absolutely. Okay. So what, what are some of the, if you were training one of our more junior staff on, on taking over an e commerce account, what kind of, what kind of two or three key things do you think they should be looking at? On a weekly and daily and monthly basis specific to shopping campaigns and feeds.

It’s difficult to say about the, the daily and the weekly. Because particularly with accounts that have lots of products, you need to build up data and it’s, it’s difficult to be difficult to manage accounts when it’s, uh, like reactionary to the small amounts of data. You kind of got to look at things in the bigger picture.

The main things that I would say to go look at are, um, your product performance over time, uh, your, the strength of your assets, if you’re, if you’re in a performance max campaign, whether you’re, whether you can improve your, your images or include videos in the, um, I see a lot of accounts tripping up on, we’ve set up a performance max campaign and we’ve given it images and.

Text assets, but no video because we don’t have that. Um, so then Google auto generates a video for you with your image and text assets so that it can put you out on video. Um, some of those are terrible to put it mildly. Well, generally, it’s a good rule of thumb that any auto applied stuff you want to check before it’s auto applied anyway.

Yep. Um, but one, uh, quite nice snippet is that if you do get, cause you can, uh, view which, uh, videos Google’s automated. Automatically create for you and um, if any of them are any good, you could submit those as your video and particularly if you do something like submitting a landscape video that would stop these automated square and portrait videos from being used, which would take you off something like YouTube shorts.

Which can, in some accounts, rack up a lot of traffic and spend without actually converting because most people watching YouTube shorts are just scrolling through the ads anyway. Yeah, the other thing that I’ve found really useful over the last, uh, kind of six months, particularly now that more people are getting Really deep dive into it is looking at, um, like scripts, automation and new tools that we can put together to to get ourselves some better data on what performance Max is actually doing.

There’s a really good, uh, script that we use that shows. What percentages of your budget are going into search, shopping, video, display? Honestly, quite a lot of them, the video and the display were not getting shown at all. It’s clearly gone. We’ll try it. It doesn’t work. Right. No more. Other thing that’s really useful and something that I think we do with most of our accounts is Um, brand exclusions, a lot of people are finicky about where and when their brand traffic shows.

So while, while you can’t add a negative keywords or it’s very, very difficult to add negative keywords to a performance max campaign, specifically, you can add a. Brand exclusion list, which will then cut all of your brand traffic out of performance max. And we see so many accounts come through when we audit them where that’s not been selected.

And a lot of the performance that they have been getting in those campaigns has been brand terms. Yeah, absolutely. That’s a, that’s a real thing. We see many digital marketing and e commerce managers and business owners just aren’t aware of. They’re just looking at the top level stat. We’ve got several clients like that who we’ve taken on.

One of our clients in the trades industry, shall we say, they, I remember when, when we started working with them, 90% Of their conversions came from brand it’s like you, yes, you need to advertise a brand, but you’re going to get those anyway. And then when we work with the clients, we’ve got in the food sector where it’s a consumable product.

So, ideally, you know, people come back and buy it again and again, because they’re drinking the protein shake, eating the meal, you know, down in the chocolate bar, whatever. Brand is important, but you have to split it off, don’t you? And I feel like too many in house teams and. And agencies just hide behind brand.

So, yeah, it’s a really good point to call that out. So, what, um, what, I guess to wrap this up with, is there, is there anything else that’s kind of, I feel like we’ve talked a lot about how we get around the pitfalls of some of the shopping things, you know, not seeing things in Performance Max, or people just putting things all in one place while it’s time to get data.

But what are some of the nice things about, Shopping campaigns that are useful. I mean, it’s a great route to market for e commerce stores. For example, we can run promotions and things that, what are some of the things that, you know, when client says, yes, we can run a promotion or whatever, it’s like a, a nice, easy, good win thing to have with shopping campaign.

I think more people now are using, uh, Google Shopping as in the search, not the, the ads. Um, to, to just generally scroll through, I think more people are getting confident with. I can look at that platform and see various different products. So it’s a huge boost of visibility now. And it’s one that actually works and converts.

You’ll find that generally the cost per click is a lot lower than if you’re advertising through search. We’ve got, um, 1 of our clients who, um, they’re in the trade industry and the majority of their, uh, account has been historically. Uh, search, even though they have tens of thousands of products and, um, works great has has worked great for many years.

But now they’re looking for that next level step up of how we can, um, get in front of more people and get more sales. And I think they’ve been hesitant for a while about, uh, investing in Google Shopping. But since we tried that, we’ve had just amazing results in it. Shopping has been consistently one of their best performing campaigns for probably the last 18 months.

But that’s because I think in that example, they’ve also got a website that can convert because, yeah, you can get cheap traffic from shopping, but, you know, if your products aren’t right, or your website doesn’t convert, it’s all goes together, doesn’t it? And the whole chain of things. Yeah, everything’s all gonna feed back into.

The basics of, is your pricing right? Is your website right? Is it easy to set up an account or buy something? Is there a good incentive to come back? Do you even make enough profit? I mean, we work with a, a retailer of tech products, and they, they sell, they sell, try to work with us, they sell headphones.

You know, the Bose headphones, and they were, they were selling like hotcakes because they had some legacy deal with Bose. Um, so it looked on the front end, because they were only looking at conversions, that they were making loads of profit there. But the, uh, sorry, loads of revenue there. But the difference is they sell headphones at 15 quid, uh, 500 quid and 1500 quid.

If you’re always targeting the same thing there through a shopping ad. Then if you’ve got the same ROAS targets there, you’re going to be making drastically different. And when we actually went into the profit, they were losing something like 15 on every sale, whereas on the higher margin items, there’s just so much scope to do things, is what I’m saying with shopping.

There’s just, there’s so much to analyze that’s outside of the account and it just needs constant, constant maintenance. And so I think. What we’ve gone through today is great because it all ties back to, you need a strategy, you need to measure everything and you need to focus on the right things, which is kind of our core message of what we always say here.

So thanks Rory. That was, that was great. We’ve got some good insights into e commerce and shopping campaigns and thanks for your time. We’ll see you on another podcast. All right. Great. Thanks Chris.