Lacing Google Display Network Targeting
Hello fellow advertisers! Today I will be going in-depth on a concept that I’ve talked about in a couple of my previous video blogs about the Display Network: lacing. Lacing is a form of targeting that allows you to tie multiple pieces of targeting criteria together within the same ad group. Since the Display Network is renowned for being too broad, this is a great technique to help narrow your audience sizes down and get your ads in front of a more focused list of users. Enjoy!
Hello everybody, and welcome to another episode of my video blog series. I am your host, Nick Rennard. I’m the head of SEM here at Digital Reach Agency and today we’ll be doing a follow-up to a couple blogs and video blogs that I’ve done in the past regarding the Display Network. I’ve talked about a strategy, I call it lacing. I don’t think it’s an official term, it’s just what I’ve been using that I’ve gotten some questions on and have been explaining it to clients and co-workers. We’ve had quite a bit of success with it so far, and I wanted to do a episode that revolved strictly around this lacing method and kind of explaining tips and tricks and set up and yeah, give you guys an idea of how I’ve be utilizing the Display Network. Let’s go ahead and get started.
All right, so we’ll start off with the table of contents here. I’m going to do a quick recap on the Google Display Network in general, I’ll go over my lacing strategy that I use on the Google Display Network. I’ll talk about the set up, including the settings that you need to make sure you double check, either in the Google AdWords editor or in the browser interface, whichever one you prefer to use for campaign creation. I’ll discuss a bit about the Ad Group set up as well, I do have a section with recommendations on how to bid on these campaigns, which we’ll go over some of the pros and cons of the Display Network, generically speaking, and how to adjust your bids accordingly. We’ll talk a bit about targeting and what my recommendations are in terms of how you should target using this lacing method. Then I’ll do a quick recap at the end just so you have a final list of check points that you’d need to make sure you cover when setting this up.
All right, so let’s start with the Google Display Network recap here. On the Google Display Network, and this differs from the Search Network … On the Search Network, we can only use text ads but on the Display Network, with what I’ll be showing you today, we can use either text ads or image ads. You can take your text ads from the Search Network and run them on the Display Network just the same. One of the advantages of the Display Network is that we do get to use image ads.
There’s a couple of ways of going about creating image ads. I think the most common way that most people do image ad creation is through AdWords. AdWords I do have hyperlink here to their AdWord, if you just … I’ll pull it up here, but if you just Google AdWords image ad creation, I think is what I looked up, you can find this support page here and it will talk you through how to create image ads. They have their own way of doing it through the AdWords interface, which is actually quite good overall. I mean, it’s not as good as having your own graphic designer do it custom tailored to how you want it to get done, but it’s a pretty good placeholder if you want to run image ads and don’t have a graphic designer or don’t want to pay one.
Second here, you can also create and upload your own custom ads. This is the way that most of our clients do … I know that here at Digital Reach we offer image ads to our clients for free. We do have an internal graphic designer that we can pay a small fee to get image ads created for any of our clients. It’s really not that hard, even if you don’t have someone internally, it’s not hard to outsource it to a company like us or anybody that can create image ads. Yeah, if you don’t like the text ads, if you don’t like the way they reformat text ads in AdWords, then you can use image ads. Anyways, that’s enough on that.
The types of targeting that we can use on the Display Network … I’m actually going to be going into this pretty in depth, but the first type of targeting that I’m going to be covering is called content targeting. There’s two forms of content targeting. One is key words and the other is topics. Keyword targeting on the Display Network is very different from keyword targeting on the Search Network, and I’ll go over that once we get to those slides. Then topic targeting, I’ll review that as well on that slide.
The next type of targeting that we can use on the Display Network is called interest targeting. There’s three different forms of interest targeting. Interest targeting it targets users based on their search history. The three different kinds are affinity audiences, which are audiences that Google sets up; custom affinity audiences, which are audiences that you can set up; and then, in-market audiences is another one that they have internally at Google, again we’ll review exactly how those work once I get to that slide.
Next type of targeting we can use is one that you probably all heard of, remarketing. I’ll got over it briefly, but I think most people, at least most people that I work with, are already very familiar with remarketing so we don’t need to go too in depth on it. Remarketing doesn’t really have a lot … I don’t use remarketing a lot in lacing, and I’ll cover why once we get there.
Then lastly, we have placement target. We’ll review that as well.
All right, so, let’s talk about lacing because that’s what the majority of this video blog is going to be about. Lacing is a strategy that I talked about in my previous video blog. It you go to our website, digitalreachagency.com, you go to our blog section, there’s one that I wrote called “Google AdWords Display Network,” where I talk a bit about lacing. That particular blog, obviously, doesn’t go too in depth on it but this blog here, or vlog, video blog, is my follow-up to that blog. If you are interested in learning more on this, I suggest you look into that.
Let’s go ahead an get into what is lacing. Lacing is a strategy that we use to help target a more focused list of users. What do we do that? We do that because on the Display Network the pool of people that we can target is quite large. The analogy that I always draw is that if you think of the Search Network as planet earth, the Display Network would be the rest of the universe. Pretty big difference in size there, obviously. There’s some pros and cons to that. The biggest con to the Display Network being as big as it is is that oftentimes when you set up a display campaign you end up spraying your ads out into the universe, essentially, and it doesn’t really get in front of the right users. It’s just kind of everywhere.
That’s what this method helps you mitigate. It helps you bring your ads in front of a more focused list of users because most people who try advertising on the Display Network through anything that isn’t remarketing, usually ends up leaving a bad taste in their mouth about the Google Display Network in general. Which is why I want to teach you guys this method because it’s actually quite possible to get your ads in front of relevant users on the Display Network, but you just have to do it correctly.
The cool thing about lacing is that it forces the users that you’re putting your ads in front of to meet multiple pieces of criteria. Rather than just using topic targeting or just using one of those targetings I covered on the previous slide, it laces multiple forms of targeting together. Typically, I only lace two of them together. I’ve had success … I’ll go over how I lace that in in a bit, but you could lace more that two if you wanted. It’s really up to you how you want to test, and we’ll go over a strategy for that in the following slides.
Why do I use lacing strategy? Again, Google Display Network is huge so you oftentimes end up wasting a lot of your ad spends in showing in front of users that aren’t likely to convert. Lacing allows us to narrow that list of users down to a more relevant pool of users that are more likely to convert.
That’s kind of a 101 of what lacing is and why we use it. Let’s go ahead and go into a bit of the set up. I’m not going to go over … You should be able to figure out if you go into AdWords in the AdWords editor, and you’re relatively familiar with it, you should know how to create a campaign and the basics. I’m going to go over some of the things that you need to make sure you double check, because if you don’t, again, they’re just little things that you might miss when you’re doing your campaign set up that you need to check if you’re using this lacing strategy.
The first thing here is in your setting, when you run multiple audiences on the AdWords Display Network you need to make sure that you check your flexible reach settings. By default, Google will often set it to … You can see down here on the screen shot how it says “Target and bid”. Google oftentimes sets it to “Bid only”. The difference between these two is “Target and bid” means that it has to meet all pieces of targeting criteria that you have under any of these given sections. So under topics, it would say … If you have it set to “Bid only” then it means that it can bid on anything but if it gets to topics it applies that bid modifier, whatever bid modifier you choose. However, if you use “Target and bid” it forces it to meet all pieces of criteria including topics or placements or demographics, whatever you’re using for targeting, in order for the ad to show up.
For the lacing strategy, you want to make sure that whatever you’re using, let’s say you’re lacing topics and placements together in this case, you need to make sure that both of those are set to “Target and bid”. Otherwise, this strategy won’t work. Again, pretty simple just make sure you go into your AdWords editor, select your Ad Group or whatever level you’re bidding at, click on this little “Flexible reach” tab right here and make sure that all of your settings are set to “Target and bid”. Anyways, I hope that’s a reasonable explanation for why we need to set it like that. If you don’t understand it, just know that you need to go into your settings and set your “Flexible reach” to “Target and bid”.
Now, there are other applications for using “Bid only”. The most common one that’s used is RLSAs. I guess we don’t need to go too much into this, but I guess what I wanted to show you is that while we do using “Target and bid” for the lacing strategy to get our ads in front of more focused users, the application for “Bid only” in RLSAs allows us to bid on other things outside of the targeting that we’re using. In the example of RLSAs, we can bid on the Search Network but if we want to bid more aggressively for users that are populated to our remarketing lists, whatever RLSA overlay we have on our campaign, we can use the “Bid only” setting and that way it’ll bid more aggressively for people that are on that remarketing list and less aggressively for people that aren’t.
Going more into the set up. We went over the “Flexible reach” stuff. The next thing we’re going to go over is you can test lacing multiple different kinds of audiences together. Any Ad Group that you set up, it’s really up to you if you want to lace topics or keywords or demographics or in-market audiences, whatever you want to lace together. It doesn’t really matter. What I am going to advocate is that you keep all of them in their own separate Ad Groups. Each Ad Group that you set up in a particular campaign should have its own form of targeting. One might be topics laced with in-market audiences, the other one might be keywords mixed with affinity audiences. Keep them separate and label them appropriately.
A lot of clients that I work with, once I explain all the different ways that we could lace things together, because really there’s, not infinite combinations, but there’s a lot of combinations that you could do on the Display Network. It’s better to keep it to one or two or three strategies at first. The reason is is that if you set up 19 different Ad Groups, all with different forms of targeting, you’re not really going to gather statistically significant data on any of that. In fact, usually what happens is that if you set up all those Ad Groups it all ends up getting funneled into whatever audience is the biggest audience and you’re just going to have 17 or 18 of those Ad Groups are going to be … Or the majority of those Ad Groups are going to be not getting any clicks at all because all that traffic is getting soaked up by whatever your broadest form of targeting is.
I recommend starting out small, keep in mind that the more different types of targeting that you try at the same time, the longer it’s going to take for you to gather statistically significant data. Just keep in mind that more is not better, in this situation. You should just test one thing at a time, make sure that your set up … It’s better to do less and make sure that your set up is correct. Also make sure that you’re spending time analyzing your data accordingly. If you’re just trying to show off and set up a bunch of stuff to show your boss, or whatever, how much you’re doing in the account but you’re not doing a good job analyzing that data and what’s working or what’s not working and making adjustments accordingly, it’s not really a good longterm solution. Sure, it might be cute to run that campaign for a few months but if you’re not analyzing the data or gathering statistically significant data, then all that money is pretty much wasted.
Another recommendation here is to only run one set of images … I already talked about running on audience per Ad Group, but only run one set of images per Ad Group. Sometimes users respond differently to different sets of images. Again, it’s just the standard rules of a science experiment not to change too many variables at once. I would recommend that if you do want to use different images, I would recommend setting up a separate campaign or at least a second Ad Group that had the other set of images but the same targeting. That way you could see if users have a higher or lower click-through rate on a particular set of images. Again, if you mix and match too much it’s just harder to determine whether the ads had … Which one of those variables actually affected performance overall.
Just a couple of recommendations. Again, try to keep it small at first. Just make sure that you only run maybe one or two or three different types of targeting. Also, I would recommend only starting with one set of ads, maybe two. Long story short … I know I’m rambling a lot here, but long story short, just make sure you don’t spread yourself too thin on your Ad Groups.
All right. Let’s jump into recommendations here, and we’ll start with bidding recommendations. When you go to the Ad Group level while you’re doing the set up, you want to make sure that you’re using default Ad Group level bids. There’s really no need to bid more granularly because after you’ve run the campaigns for awhile you can look at performance more granularly and then make adjustments from there. When you first start the campaign off, I always recommend just use the default level bids. Set your bid to 50 cents or a dollar or whatever you want to set it too, and then analyze it after you’ve gathered some data and make adjustments after that.
If you try to bid more for one audience and less for another audience before you have any data to tell you whether or not that’s the right call, you’re probably just wasting your time because you don’t have any data to tell you that one audience is necessarily better or worse than another. I recommend just using the Ad Group level bid, gather some data, and then you can look more granularly at which topic is performing better or maybe which ad is performing better or Ad Group is performing better. You can go from there once you have some data but just keep it simple and straightforward at first.
Second recommendation here with bidding is to start with small bids. Small is obviously a relative term. Typically, when I create a Display campaign, I’ll start my bids off anywhere between 25 cents and a dollar. Clicks on the Display Network shouldn’t really cost you more than a dollar. There are times where … I set up Display campaigns that have cost per clicks that are above a dollar on the Display Network and I’ve had profitable campaigns but typically speaking, clicks on the Display Network should be less than a dollar. If you’re paying more than that you’re probably paying too much. Start small. Start with 50 cents, 75 cents. See what kind of traffic that gets you, then analyze from there.
In terms of how to analyze your data, the most important metric that you’re going to want to look at after you’ve gathered … Let’s say you’ve let it run for a week or two, the first metric that you’re going to want to look at is impression share. There’s three different forms of impression share. There’s total impression share, there’s impression share loss due to rank, and impression share loss due to budget. I’ll do a quick recap. Impression share is the percentage … Let’s say you have a million users that your ads could show in front of, but your ads only show to 250,000 out of that million. Your total impression share in that case is going to be 25%, because your ads were showing in front of 25% of the total pool of users. That’s what impression share is.
There’s two reasons that you might not show in front of that other 75% of users. The first one is called impression share loss due to rank and the other one is called impression share loss due to budget. Impression share loss due to rank happens when somebody else who’s bidding on the same banner slot either has a higher bid than you … Maybe you’re bidding 50 cents or 75 cents and they’re bidding $5 dollars, they’re going to outrank you. Not necessarily a bad thing. Maybe they’re paying way too much for their clicks and their campaigns are just not profitable at all.
The other reason you can lose out on impression share loss due to rank is if you have a low quality score. Let’s say your image has low resolution or crappy click-through rate, or maybe your landing page sucks, or maybe all of the above, and Google gives you a quality score of two, even if you have the bid of $5 dollars and the other person has a bid of 75 cents, if your quality score’s a two, and their quality score’s a nine, there’s a possibility that they’ll still outrank you even with a much lower bid. Again, make sure you’re executing all the best practices in terms of quality scores. Not going to go into quality scores today because we have a bunch of other blogs already written about that, so you can check out our blog section if you want to learn more on that.
Impression share loss due to budget is the other reason. The way that impression share loss due to budget happens is, let’s say you set your budget limit to $30 dollars and let’s say your cost per click is a dollar. And in the first 12 hours of the day, you get enough clicks to maximize your budget. Well, you can’t get any more clicks after that point in the day because your budget’s already capped. Which means that your campaigns are just going to shut off. Ad Words will automatically shut them down so that way you don’t go over your allotted budget. Any click that you could’ve gotten after that point in which your campaigns shut themselves off, any click that you could’ve gotten if your budget were more open is called impression share loss due to budget. Again, if you want to mitigate that just open up your budget and let it spend more.
After the first few days, make sure you check your impression share. If your impression share loss due to budget is too high, you want to reign your bids back in so you can get a cheaper cost per click and get your ads to show longer throughout the day. If your total impression share is low, you can try increasing your budget, you can try narrowing your targeting, there’s a lot of methods that again, I have a lot of blogs on impression share so you can look into that.
All right, so, audience lacing recommendations. Recommendations for targeting. First, you want to pair one piece of content targeting with another piece of interest targeting. Again, you can lace any two audiences you want together but this is my recommendation. Pick keywords or topics and pair it with one of the three affinity audiences, custom affinity audiences and in-market audiences. The reason that I do this because content targeting and interest targeting are really the only two different forms of targeting. Content targeting is based on the content of the pages, whereas interest targeting is based on the interests of the users. I find the best results when I pair one of those two together, however I have had success mixing two types of content targeting or two types of interest targeting, but again, my recommendation is to do one of each.
You can slice remarketing and placements into the mix, however, I recommend keeping remarketing separate because I think those lists are already narrow enough that you don’t need to mix it with anything else, it’s just good on its own. Placement targeting, I’ve talked about placement targeting and how I’m not a huge fan of it in Ad Words unless you already have a good list of placements that you want to show up on. Typically, you need a pretty extensive list of websites that you’d want to show up on, and honestly, if you’re going to be using placement … Placement targeting is just a form of content targeting and I think that keywords and topics just do it better. Anyways, pair a piece of content targeting with a piece of interest targeting, that’s my recommendation.
Wrapping up here, just a recap on what we went over. First and foremost, pick two audiences that you want to put together, one interest and one content targeting audience, put those two together. Make sure your “Flexible reach” is set to “Target and bid”, you do not want it to be set to “Bid only”, otherwise this method won’t work. Then, third, when bidding on non-remarketing Display campaigns, you want to start out with small bids. I would start anywhere from 50 cents to 75 cents, probably a good start. Sometimes you do need to … You started out at a number like that and it runs for a day or two and it doesn’t spend anything. If that happens, just increase your bid. It’s the nature of this game is you never really know where to start, you just got to do a good job analyzing your data and adjust accordingly.
Fourth, upload your image and/or text ads. You can try both. I don’t necessarily think that one is better than the other. I typically will try both, and I will look for which one has a better click-through rate, or a better cost per click, or a better performance in terms of conversion rates, and then I’ll adjust according to whatever KPIs are important to you. In most cases, probably lead gen. Click-through rate is also a big one in terms of maybe split testing something like that. Then lastly, you want to make sure that you let your campaigns run until you have statistical significance before you make adjustments. Once you have statistical significance, then you can start making adjustments based on your impression share.
Impression share, again, is what you’ll be analyzing first and foremost. There are other metrics, plenty of metrics, within AdWords that we can use as KPIs in terms of where we want to increase bids, decrease bids, maybe it’s an awareness campaign, you just want to generate a ton of traffic. Yeah, there a lot of reasons you might want to change that. Just make sure you have statistical significance first. Don’t spread yourself too thin and start making adjustments before you actually have data that proves it one way or another.
Anyways, that’s all I have for today on the lacing strategy. Hope you guys enjoyed this video blog. I will catch you guys in my next video. Thanks for watching.