Campaign Segmentation in Search Ads

A few weeks ago, we released a video where I talked about the most common for search campaign structures that we see in accounts that we take over. If you’re interested in learning about those, you can check out this video right here. But in that video, I mentioned that my approach has always been the consolidated or thematic approach.

I have always used that structure. It’s what I manage best. And I got a number of follow-up questions asking how I end up segmenting campaigns based on that consolidated thematic approach. So rather than answering those in the comments or via email, I wanted to put together a video and walk you through the main considerations and some of the approaches I take for breaking out my search campaign structures.

Just like other videos on the Paid Media Pros channel, I’m going to start this one off with a caveat, which I know is what you all love hearing from me. As I was doing the prep for this video, I was trying to create as complete of a picture as I could of my search campaign strategy.

But I kept having questions like, what about this scenario? What about this other scenario? What about this random occurrence that happens within 10% of accounts? And it became clear to me that I was not going to be able to put together a video that would answer every single question that was out there and help everybody create exactly the campaign structure that they want and need.

If you’re here hoping that this video will give you exactly the answers that you need and you’ll have no other questions, I’m probably going to let you down. If you really want your custom search campaign built exactly the way that I would do it, we do offer that as a service at Paid Media Pros, but you’re going to have to pay me for it.

It’s not going to be in this free YouTube video because every business, every account, and every account manager is different, and it’s just simply impossible to put together a video that will cover everything that everybody has a question about. So with that out of the way, let’s jump into my consideration process and some potentially imperfect but hopefully illustrative examples of how I would put together a search campaign using the consolidated or thematic approach mentioned in the other video.

The first thing I do when I start planning for an account structure build is to think about what my biggest segmentations are, what are the biggest delineators in the account that are going to impact my structure. And the first things that might come to mind for you are going to be based on keywords and products and all this sort of thing. But actually, the list lives with the different campaign settings because those are the things that you could only do at the campaign level that you really need to have incorporated into your structure.

One of those is budget. You can really only set the budget at the campaign level. For search campaigns, you can utilize shared budgets, which in a lot of ways, can be imperfect. So I always try and think about how I can structure my campaigns for any budget considerations that I have.

Maybe I’m supposed to focus 80% on our bottom-funnel intent keywords, but I can utilize the other 20% for top of funnel research based keywords. In this scenario, it makes sense to break those two out because we have different budget considerations for them. Geography also plays a role in this if you’re trying to target users in the United States as well as the United Kingdom and Australia, and New Zealand.

If nothing else, those are on very different time zones. So having those all incorporated into one campaign could make it very difficult to understand different day-parting types of performance trends. But additionally, most companies that I work with that advertise internationally also have budget needs that are broken out by different countries.

They also have reporting needs which are broken out by different countries, and those are easiest to address at the campaign level. So think about how you need to report different performance metrics and budget considerations to your client or to your bosses because that could also impact how your campaigns are structured.

Conversion goals can also be set up at the account or campaign level in a Google Ads account. So if you have a set of keywords that are going to be focused on a very bottom-funnel goal, and another set of keywords that are focused on a more mid or top-funnel goal, those should be broken out into different campaigns. So that not only can you have them separated and controlled by budget, you can also have different conversion goals assigned to each of those campaigns so that they’re tracking the goals that matter most to them.

And all of the machine learning can help you optimize to those specific goals and not the account-level goals. And if your account has been around for a while, historical performance can also help you break out different campaigns. In some scenarios, I have campaigns where clients really want to continue advertising on keywords that logically make sense but just do not perform well.

So we segment them into their own campaign so we can control the budget, we can report on them separately, we can include other conversion goals but still, keep them apart from most of the rest of the campaigns. So here are some quick examples of what these segmentations could look like. Maybe you have two different product groups that have their own budget allocated.

Some large companies have teams dedicated to advertising an entire product, and they have their own budget line items. So having those separated by campaign with the subsequent ad groups and keywords reflecting those products is the best way to go.

Same thing for performance. Again, maybe we have the high intent bottom funnel keywords in one campaign, and then we have the lower intent mid to high funnel keywords in another campaign separated out so that we know the performance is different, campaign goals are different, all that good stuff.

Then the segmentation that might look a little bit different if you were paying attention on the previous slides is broken down by geography. If you’re trying to advertise the exact same products, you probably have the same ad groups and keywords across the United States and the United Kingdom.

For the performance example, you’ll notice here that the low intent is ad groups three and four and keywords five through eight because these are going to be different ad groups and different keywords than what’s in the high intent campaign.

But with the geography campaigns, it’s going to still be ad group one and ad group two, keywords one through four, because it’s basically just a duplicate of the United States campaign broken out and targeting someplace else because we’re still trying to advertise the same product or service, but we needed to break things down by geography instead. But not every account is going to need all of these different segmentations.

You could have an account that is relatively simple where you have one budget, one country. They’re all optimizing toward the same goal. There are no additional campaign consideration needs. And with that, you basically get to skip that whole step one of trying to figure out those high-level segments.

So whether you needed to have any segmentation or not, once you have your campaigns determined, now we need to work on utilizing keyword relationships to develop the ad group structures. We’re trying to focus on what keywords are going to live in which campaigns to make sure that we have logical campaign structures and subsequent ad group and keyword groupings.

For this, I’m going to do a couple of examples to try and cover e-commerce and lead generation considerations. So first, let’s take a major retailer like Macy’s. They have tons and tons of products and categorizations on their website. So putting together a set of campaign structures is going to feel really difficult.

So the first thing I try and do is think of all of the different ways that I can organize these different campaigns in an account. So some of the potential segmentations that come to mind, it could be based on the department think something along the lines of clothing verse shoes versus the home decor department, all those things.

Next could be brand, could be clothing type, maybe it’s the customer, so men versus women versus kids. Or it could be margin. There were a handful of different segmentations in that menu that talked about the products that are on sale. But even on top of that, some products have a bigger margin than others. There are a lot of different ways that I could potentially try and structure this account. And at this stage, this is what you’re looking for.

You probably have a good idea, at least a common sense, of what the keywords in the account are going to be for this product or service that you’re trying to advertise. And at this stage, we’re really just looking for any of the potential segmentations of how you can slice and dice those different keyword groups to make it make sense for all the different campaign settings, reporting needs, conversion goals, and also how your mind works.

As the campaign manager, don’t undersell any of the potential segmentations just because it’s the way that your brain works. I’m a big advocate for setting up a campaign logically for the person who’s going to manage it. Because if they’re the person pulling the strings and this is what makes sense to them, that’s probably going to give you the best chance at good performance. So, with that in mind, here’s an example campaign set of how this would work.

And again, this reflects how my mind works. The first piece I chose was Department because it makes the most sense for me to advertise shoes separately from clothes, separately from home decor, all this stuff. So here are a handful of how these campaigns could be segmented out.

And you’ll probably notice that although I started with the Department as the first identifier, I used the other potential segmentations to further narrow it. So in the shoes department alone, I have four campaigns. Two of them are Nike, two of them are non-brand.

So this is where I start leveraging some of those brand segmentations that I could have used. But rather than grouping Nike altogether, it makes sense for me to have a shoes campaign that focuses on Nike. I then also have different segmentations based on customer, so men’s verse women’s.

And then, I have an additional segmentation of running versus casual shoes, which wasn’t really one of the potential segmentations at the beginning. But as I got further into a potential structure here, it made sense to have the purpose of the shoes broken out separately. The same types of trends hold for the women’s and men’s examples that I have here too.

As I said, this is just one way that you could put together these different campaign groups for a major retailer like Macy’s. And you might not always land on the perfect segmentation the first time. So don’t hesitate to mix and match.

Try different organizations and see what makes the most sense for your mind, but also your campaign optimization and controls. I mentioned we’re going to do a lead generation example. We’re going to use B2B SaaS offerings for this one potential segmentations; here could be the service line.

Maybe you have a handful of different specific services that you offer, could be based on the buyer cycle. So kind of their intent based on what keywords they’re searching for, the same sort of thing. Maybe based on the keywords that they’re searching for, you can determine that person’s level of authority to make that decision.

Or you could be able to lean into the company industry if somebody searches specifically for retail verse insurance versus education, something along those lines. So again, think of all the different potential segmentations. Highest level, common sense, logical ones usually come up first. But this will all be different based on whatever keyword group you have in front of you.

For this example, I’m going to use Modern Device Management or effectively when you have an employer who manages your software externally, so you have a work phone, you have a work computer, work takes care of all the software updates, all that good stuff. So here I might have different breakdowns of Modern Device Management or MDM, and it matters to me that it’s an employee phone, an employee computer that are separated from a remote device.

And then there might be instances where people are typing in specific brands like Apple devices or Android devices, but then as a separate service, maybe we offer just IT solutions. So we have employee support and software implementation to get started. So the employee support is obviously helping the end user.

Software implementation might be helping the organization roll out new software to all the employees. So we’re targeting different users within that relationship, and the campaigns need to reflect that. So again, think of all the different segmentations you have.

Mix and Match, come up with different options, and land on the ones that make the most sense for you. Once we have our campaigns broken out and we have a good high-level view of what our campaign segments are going to be, we next need to work on keyword segmentation which will influence what our ad groups are going to be.

And I have a pretty simple rule of thumb when it comes to this. I get this question a lot how do I know when to break out a different ad group for a set of keywords? And here it is if there’s a difference in intent behind the keyword or a meaningful difference in the language that somebody is using in that keyword that I want to reflect in the ad copy, the keyword gets its own ad group.

If all of the keywords have about the same level of intent and they might be using different words but not in such a meaningful way that it would matter to have ad copy reflecting those specific words back to them on the search engine results page, then I’m not going to break it out in its own ad group.

At that point, you’re just bulking up your campaigns for the sake of doing it. But anytime I think somebody has a different intent or I want to write different ads, that’s when I start to break things out. So let’s walk through a couple of examples using the same businesses that we’ve already been talking about.

Here are the potential campaign names that I used earlier for the other examples, and we’re going to focus on just two of them shoes Nike women’s and MDM Apple devices. So let’s assume I have some keywords that I want to get into these campaigns, and I need to think about how I’m going to structure those ad groups and group these keywords together.

This certainly isn’t all of the keywords that I would put into any of these campaigns, but again, these examples are going to be imperfect, but hopefully, you’ll get the gist of it when we’re done. You can see in the shoes campaign I have a number of different options of people searching for women’s Nike shoes, whether they are red track shoes, they’re running shoes, they’re basketball shoes, they’re for CrossFit, somebody’s using ladies instead of women’s, or they’re asking a question about do they even make women’s basketball shoes.

We’ll come back to the legion example here in just a second. But I want to start talking about how these shoe keywords can be broken out into different ad groups. For my mind, it makes sense to focus these on the different type of sport that somebody is using because each of those had a different sport or activity included in the keywords. So I would have four different ad groups here based on those keywords. One for basketball, CrossFit, running, and track. The CrossFit and track are pretty simple.

There was only one version for those, and both of those make sense to me. Depending on the search volume available for red, Nike women’s track shoes versus blue or purple, or no color included, would determine whether or not I broke out separate ad groups for the color included. But for right now, it’s going to be grouped all together because I’m probably only going to be able to advertise a handful of products with the track shoes anyway.

So I might as well leave them all together and let somebody filter for red on the website as opposed to forcing them to it right away. For the running shoes example, this is where the difference in words isn’t big enough for me to make a change in the ad copy. I don’t really care that somebody used women versus ladies. That doesn’t seem like a fundamentally different intent behind that change in words.

If somebody typed in ladies’ and women’s popped up in the ad copy, they would probably be fine. So those are going to live together in the same ad group, the same type of example, but just with intent is for basketball. One, somebody is actively looking for shoes. We don’t know if they’re trying to buy them or not, but they know that Nike makes women’s basketball shoes.

The second keyword. Those people don’t actually know if they make women’s basketball shoes, but I don’t think that writing ad copy saying yes makes sense. It still makes sense for me to just send them to the women’s basketball shoes page on the Nike website and have that kind of inherently answer the question.

Now let’s look at the second example, the MDM Apple devices. So we have remote iPhone management, manage employee iPhone remotely, employee MacBook, remote control, remote access work MacBook, modern device management iPhone and remote manage employee iPhone. Now let’s look at how I would probably segment these if I were building out campaigns for them.

The biggest segmentation I have here is based on iPhone versus MacBook because those are two very different devices. But then, after that, I have it broken out by employee versus general search terms. So the employee iPhone is the first ad group. This to me, says that an employer or a boss is trying to manage somebody’s iPhone remotely.

So I have these keywords grouped together. The second group around iPhone might have the same level of intent, but I don’t quite know that yet. So I don’t want to focus my ad copy as much about being a company managing an employee’s iPhone. I would rather just talk about managing iPhones remotely and let them find their way into an employer or a regular user funnel.

The same is true for MacBook. The first option I have here is remote access work MacBook, and then the second option is employee MacBook remote control. I still feel pretty good about the employee MacBook remote control being an employer trying to manage an employee’s MacBook. But Remote Access Work MacBook to me could be an employer, or it could be an employee who got themselves locked out of their MacBook or left it at the office, and they’re trying to access it remotely.

So here I’ll break those out separately because I do think there’s potentially a different level of intent, and I want to call that out in my ad copy and potentially even on my landing pages. So now that we have our campaign segmentations done, we have all of our ad groups and the keywords associated with those ad groups segmented the way that we want it.

The last thing I want to talk about is match types. We’ve already put together a video that talks about how the different match types function on Google Ads. If you’re interested, you can check out the video at the top of the screen right now. But my approach is pretty simple, and I did put an asterisk here because this is as of January 2023 because things do change. But here’s my strategy at this moment, all keywords in my account start with both an exact match and a phrase match variant.

Both of those match types live in the same ad group in the same campaign. This is kind of the heart of that consolidated or thematic approach. There’s no segmentation for match types at the ad group level or campaign level. They all live together.

And then lastly, Broad is to be tested only if the other match types are already doing well and we need more scale. I have a love-hate relationship with Broad match keywords, and that relationship continues to ebb and flow as the changes happen for Broad Match. If you want to know how Broad Match currently works, we do have a video that talks about Google Ads Broad Match by itself and some potential strategies.

You can check out the video at the top of the screen right now if you’re interested in that. But my personal strategy is still to always start off with phrase and exact, see how the performance goes, and then expand to broad if needed. And that’s pretty much it for my campaign structure strategy.

As a quick overview and revisiting what we just talked about, first consider all the segmentations that rely on campaign-level settings. That’s where you need to start because that’s the highest level, and you need to make sure that you’re taking that into consideration. Then create logical campaign groups for account segmentation for the major retailer.

This could have been based on department or brand, or customer. For the lead generation example, that could have been based on the service line or the authority of the user. Whatever makes the most sense for you and leans into the right campaign settings.

All your reporting needs and what you think you’re going to manage the best is what you should run with next. You need to group the keywords into those campaign segments and then match them at the ad group level, trying to get as close as you can for user intent and the words included in those keywords.

If there’s no difference of intent and you’re not going to utilize a different ad copy, probably just leave them all in the same ad group. But if you want to adjust your ad copy message because of the words or the intent behind them, start to break things out so you can address them appropriately.

And then lastly, I always include Phrase and exact Match variants to start and then look to expand only after I’ve seen good performance from those. So just like with any other video on the Paid Media Pros channel, please feel free to leave any questions you have in the comments below, and we’ll work on getting an answer to you.

Written by Michelle Morgan