Welcome to the LinkedIn Ads Show.

Resources for this week:

LinkedIn Learning Course
http://j.mp/linkedin-ads-course

Contact us at Podcast@B2Linked.com with ideas for what you’d like AJ to cover.

Transcript:

Getting the right account organization and AB testing strategy on LinkedIn ads is crucial. Buckle up, we’re about to get nerdy.

Welcome to the LinkedIn Ads Show. here’s your host, AJ Wilcox.

Hey there LinkedIn Ads fanatics. So as you know from the teaser, today, we’re going to be diving super deep into account organization, and AB testing strategy. We’ll cover things like your optimal account organization, your ad rotation settings, how many ads you should have in each campaign, and more. Let’s hit it. For a count organization. Let’s start at the account level. So on LinkedIn, you can have one page. That’s a company’s organizational page per account. But you can have many different ad accounts for a single page. And that’s for purposes like let’s say if you were Microsoft, you wanted to have a separate team or Separate agency running your Xbox ads and from someone running office or someone running Windows. So you can have multiple ad accounts for each page. But for any single ad account you have, you can’t publish to multiple pages. And at the account level, when you very first created, there are a couple things that are totally final that you can’t change afterwards. You can change things like the name of the account, and all kinds of things, but you can’t ever change your currency. Once you’ve locked in which currency you’re going to be working in that account is done. So we had a client where we launched in US dollar. And as it turns out, we needed to use a Canadian credit card to fund the account. And we had to just throw the entire account away and start a new account that runs just off Canadian currency. As soon as you associate a company page or an organization page to your ads account, that’s done. It’s totally now baked into that account, and you can’t disassociate it later. All right past the record. counts level, we get into an organizational grouping called a campaign group. Now campaign groups, from my opinion, are pieces of garbage and the bane of my existence. But what they’re supposed to be what they were originally designed to be is kind of like with Google ads, we had ad groups and Facebook ads, we have ad sets. But instead of putting them underneath the campaign, they put them up above in the hierarchy. And it didn’t really make a whole lot of sense. So as I understood, they were supposed to be a way where we could put in like multiple credit cards on the same account. So companies like Microsoft could break it up and have, you know, a different credit card within the same account for the office team, the Xbox team, the windows team, etc. And then those totally separate groups could be on separate budgets without having to create a new ads account. Unfortunately, this really never happened. And they weren’t very well engineered, and now they really can’t be easily removed. So we’re kind of stuck with them, and they don’t seem to be doing a whole lot of touch up or development on them. So they’re a little bit of a nuisance when you have campaigns in multiple campaign groups, and you can’t see them all in one view, you have to click up, and then down a different level to, you know, compare different performance. So I don’t love them. But there are some ways that we work with them that tend to work pretty well. So any account that is an online account, meaning that it’s self service, you set it up yourself, LinkedIn didn’t set it up for you. You’re going to have a campaign group in there called default campaign group. And it’s very rigid, you can’t delete it, you can’t rename it. There’s a lot of things you can’t do with it. But what it is, is a campaign group that you cannot turn off or back on, you can’t rename, but it is in every account, and it can hold up to 1000 campaigns. That is its only strength. That’s the only campaign group that can hold a lot of campaigns. So if you have a big account, and you’re likely going to have 200 or more campaigns in it, go ahead and just build everything in the default campaign group. And that really, that’s going to be your best bet. If you go and create a new campaign group, you get a little bit of a perk, you can turn it on or off, meaning you can pause the entire campaign group. You can also name it something descriptive. So it’s not just default campaign group taking up space. The big downside is they can only hold up to 200 campaigns total, which is why they are the bane of my existence. So anytime I have a large account, I can’t rely on a new campaign group to hold everything. I’ve got to go with the default. But one really cool thing you can do here because you can turn these off and on is, let’s say your landing pages go down or, you know, something at the company means you know, we’ve got to pause all ads. It’s really easy if they’re all in a single new campaign group to go in and just hit pause on the campaign group and be done. You’re not sending traffic to a landing page that’s down or something. Also, if you were doing manual day parting or scheduling, it’s nice to have a single entity you can turn off or turn back on. As you’re heading out of the office or coming in. It’s a lot easier than what I used to do, which was go by individual campaigns, one after another, pausing and pausing that kind of thing. So my rule of thumb here is, if you do plan to have fewer than 200 total campaigns, go ahead and build it right within a new campaign group. Name it something descriptive that you care about and if you need a logical separation between campaigns, like for the quarter or something like that, sure, you can do one of these new every quarter, but I sure like to build an account that is evergreen that is everything is exactly as it will be forever in a single campaign group just so things aren’t changing around. I can still do more granular reporting. Having a separate campaign group for a different initiative doesn’t help me very much.

5:57
All right, past campaign groups, you get into to the individual campaign level. Now from Episode 02, you’ll remember we talked all about the targeting the different things that live at the campaign level, are number one, most importantly, your targeting. Also the ad format you choose whether it’s text ads, sponsored content, message ads, dynamic ads, all of that lives at the campaign level and can’t be changed. You also have your bidding and budgeting section, that these things can be adjusted. And you’ll remember these from Episode 06. Everything about how you bid, what your budgets are, everything like that is living there at the campaign level. And then finally, your objective lives at the campaign level to which you also can’t change. That means if you chose one campaign to be a website visits, a focused campaign, you will not be able to choose to change that if you ever want to run a conversion focused one or a lead gen version of that campaign, it has to be a whole other campaign. And your accounts structure is really, really important for general account organization. Just understanding where you’re at and making it easy to read and interpret what’s going on. But also for reporting, as you’re building campaigns, think about how easy this is going to be to break down performance later for your reporting. So I’m going to share with you my personal strategy on how I build all of this out, so that you can do the same if you are so inclined. I found this strategy to make it beautiful for reporting, as well as making it really easy to understand what’s in an account, and then interpreting the results outside of the platform as well. So as you’re naming campaigns, I want you to think evergreen. That means every campaign you build, plan on this being evergreen, something that will run forever. Don’t think that you’ll ever get rid of a campaign or have to create a new version unless you absolutely have to. The reason why this is is in a single campaign that represents an objective, an ad format, and an audience, you are going to be building history. You’re building what LinkedIn calls their relevancy score in that campaign. And relevancy score, we’ll have a whole separate episode on this. But basically, this is Lincoln’s way of pinning to your individual ad or your campaign, how well it’s going to perform. And in a way, it’s level of risk. Because LinkedIn, if they keep showing an ad that never gets clicked on, and you’re paying by cost per click, then that is a risk to them, that’s money totally lost. Versus if they start showing your ad that gets clicked on a lot, then that takes away their risk and makes them more money. So you build this history in a campaign, the more dollars you spend in it. And that means that let’s say you have two ads active in a campaign, you pause both of those ads and launch two new ads. LinkedIn looks at that and goes, ooh, the two new ads, we don’t know how these are going to perform. We don’t know if they’re going to have a good relevancy score or not. We Don’t know if they’re high risk. So what we’ll do, we’ll just look at the relevancy score of the campaign. Historically, how has have all the ads in this campaign performed, and if there’s some good positive history, they’ll give you the benefit of the doubt. They’ll say, Okay, the last relevancy score in this campaign was an eight. Let’s assume the new ones are going to be an eight as well. And so immediately out of the gate, you get better performance, they won’t charge, you as much you’ll win more auctions, have access to more traffic. And it goes like that. If you tried to build a brand new campaign around the same audience, LinkedIn wouldn’t have any sort of history to rely on. And you’d start out with presumed lower relevancy scores, and you’d have to prove your way out, meaning that at least for the first few days to a week, you probably pay more per click and get less traffic than you would otherwise. So try to build your entire account to be evergreen. Build this history so that everything you launch, you’re giving LinkedIn the best signals and the best chance to let your ads perform welll. Then, like I shared earlier, the strategy here won’t be, hey, I ran my 2020 webinar ads, you’re not going to throw that campaign away. Because now that offers dead, you’re just going to pause those ads that were promoting it. And instead launch two new ads promoting, let’s say, your white paper. And so now, if you want to report based off of your, you know, certain objective, you just are going to report based off of the ads that were in that objective, don’t try to go to the campaign level, and be sad that you didn’t name your campaign name like webinar. So it’s easy for reporting, just bubble your ads up that were around the webinar report based on those bubble up the ads based off of your, your new webinar and report separately on those. So it’s still super easy, and you’ll get better cost performance. And on that note, I really hate looking at accounts where people have named the campaigns things like 2018 webinar, and I look at those and go, Ah, that just tells me nothing. What you want to do is name your campaigns, after the things that don’t change. So you don’t want to put like CPC bidding or auto bidding in the campaign name, because you can change that stuff at any time. But do name your campaigns after the three things that don’t change, your ad format, your objective, and your audience targeting. That way this campaign will always be evergreen, as long as you are still targeting that audience, or as long as you’re still using that, that objective this campaign will live on and be a great receptacle for whichever ads you have going to that audience. A good example and I’ll put this here in the show notes. A good example for how I might name a campaign would be something like if this were a sponsored content campaign, I might do sc for sponsored content and then I like the vertical separator bar, but some kind of delineation here. So sc and then the the bar or pipe is We call it and then if this is a website visits campaign, I might do WV for website visits another pipe. And then I might call this campaign CEO job titles, pipe, company sizes 51 and above pipe, banking industry, pipe US. And you’ll see in the show notes exactly how this comes out. You don’t have to follow my recommendations here. You can name it however you want. But what this does is it allows me very quickly to segment all of my campaigns all of my performance, because the ad format itself are the first two letters. When LinkedIn automatically, alphabetically organizes my whole account, I’m going to see all of the different ad types right together, which is great because they tend to perform similarly. I don’t like it when my sponsored content and my text ads data is all mixed together, because they’re so different. It doesn’t mean anything. Also, my objective is there which makes it really easy, especially if I’m pulling this out to Excel or or something else, I can keep all of my objectives together and report on them separately. And then because all of my targeting is there in the URL, as I’m looking through my account, and I see something that’s performing well, I already know what that performance represents. In this example. I know that this represents people with CEO job titles. I know the targeting was around title. And I know these are CEOs. I know they.re companies with at least 51 people. I know that they’re in the banking industry. And I know that this is just the US. This makes account organization beautiful. It makes segmenting things out amazing. It also means that if I ever wanted to run a report and say, hey, how do we tend to perform at companies with more than 51 people, I could easily pull that segment out in this campaign name. And I could compare it against campaigns that let’s say they had one through 50 employees. I’ve consulted for some companies who need the initiative somehow in the campaign name, or maybe the the internal name, they give personas. And that’s fine, I would just encourage you to make your account as evergreen as possible, so you can always take advantage of that building history in your campaigns relevancy score. All right after a quick sponsor break, then we’ll dive into information about ads, and AB testing strategy.

14:18
The LinkedIn ads show is proudly brought to you by B2linked.com, the LinkedIn Ads experts

14:28
B2Linked is an ad agency that LinkedIn ads is all we do. So go ahead and fill out the form on any of the pages of B2linked.com to get in touch, and our team will help you enact all of these strategies and more to help get you the best ad performance. All right, now let’s jump into ads. I get this question all the time. How many ads should I have per campaign? My answer 99% of the time is two. Two current ads in every campaign and here’s why. If you just put a single ad in any campaign, you won’t have anything to compare against. And that ad could be terrible, or it could be amazing. Or it could be average. And you’d never know. It’s not a good AB test, you won’t ever learn anything from running just a single thing. As soon as you launch a second ad, now you have something to compare against. And both of those ads will definitely receive at least some exposure. So if one was truly terrible, LinkedIn would optimize it out of the running and just run the one that tended to perform much better the whole time. But, most of the time, you’ll probably get some kind of split, it might be like 60/40, split between, you know, ad one and ad two in there, but both will get some exposure. As soon as you add three or more ads, and remember, this is the same on every other platform, Google and Facebook, they all have the same issue. One of those ads will get a lot of exposure, one will do okay. And the rest will, unfairly as I call it, get labeled with a poor relevancy score, and get nearly ignored. So you might have one ad that gets 1,000 impressions, one that gets 600. And then ads three, four, and five might get 24. 3 impressions, I’ve seen some get one on Facebook, or literally zero, it’s never ever been seen. And the auction can say, Ah, that’s a bad performer, I know it and just refuse to even give it a chance. That means those ads that you’ve launched, that didn’t get a chance that were labeled with a poor relevancy score from the beginning, there is no reviving them. It really could be a great ad, but you’d never know because you could bid up way high on it, and it would still never get shown. Those ads are effectively dead. So if you’re asking me if there’s a value in going ahead and creating ads 3, 4, 5, 6, 10 in a single campaign, I would say definitely not. The ad test that I launched never really had a chance to gather any data, so it won’t be statistically significant. And I put a whole lot of work into writing, you know, six extra ads for this campaign that likely won’t ever get used. So my recommendation here is always to run two ads at a time. And if you’ve got four ads or four offers that you want to run, just run them two at a time. Run, you know, the first two for a week, pause them, then run the next two. And that’s, I think, the right way to do it. Now, I mentioned I only do this 99% of the time, the other 1% of the time, I recommend launching for ad creative in the single campaign. And the reason why is because LinkedIn has this rule about their max frequency cap in sponsored content campaigns, where by default, each member can only see one ad from a company per day. And that’s how they make sure that people can’t get annoyed or or won’t get annoyed as easily. They won’t be fully saturated by someone. However, and this is as of October of 2019, this is very recent. You can now if you have at least four ad creatives in a single campaign, you are now eligible to receive five impressions from a member every 48 hour period. So that means you used to be able to have to be seen during that 48 hour period twice. If you’re bidding really aggressively. Now you can be shown five times during that time period. Before October, we’ve had this access for a while. Before October, it was you could be shown four times in 48 hours. But now it’s five, which is potentially really good, especially from those types of offers. Let’s say you have a webinar coming up three days from now or an event starting, and you want as much exposure as you can possibly get. This is the way to do it. You run four ad creatives in a single campaign, qualify for the higher frequency cap, and you’re off to the races. Now why don’t I do this normally, Well, normally, what I want is a slow burn across all of my audience. I know the average user on LinkedIn only logs in three or four times a month. And that’s okay with me. So if I’m bidding steadily, and I’m showing that with the same rules steadily, I’m going to evenly saturate that audience over time. And I know it’s usually about every 27 to 33 days before an ad will saturate an audience. But if I’m showing more times more actively, I’m going to burn out my most active audience members in my segment faster than the average member. And there are quite a few people who will log in and use LinkedIn every single day. It’s so hot on the organic side right now. So because of that, I know some people will get way more saturated. And I’m going to start segmenting my audience in a way that I can’t track. Some people will be really upset but they’re seeing you all the time. And some will just, you know, think that you’re normal. And then as soon as you you show that audience a different ad, they’re going to start reacting in a different way. So I’m a big fan of a slow burn for that average user. I want it to show evenly over time. And I really like LinkedIn frequency capping, that’s pretty conservative for that reason. But if I really want to blow an audience out and get the word out about something where there’s a time limit based off of it, I will do this. So you might ask, okay, well, now I have to write two extra ads, because I was running two campaigns or two ads in a campaign. And now I have to write two more. So I can have four. I don’t recommend that. I still want data around my AB test, I don’t want to have just two extra ads that I know potentially will get ignored, like we talked about. So what I’ll do is I will make a duplicate of each of the ads. So let’s say I have ad A and ad B in that campaign, I’m going to make a duplicate of A and a duplicate of B. Now, according to LinkedIn, I have four ad creatives in the same campaign and there are there are four separate ads there. But because two of them are identical to the other two. Now when I go to report and I bubble up all of the performance by ad, I will still only have two ads in the account. And all of the data will aggregate together, which we’ll cover in a lot more detail here in just a little bit. I told you we were going to get nerdy. And one other reason you might want to take advantage of that five impressions every 48 hours thing is if you have a lot of budget to spend, and you’re just trying to spend it as effectively as possible. You could do that by doing things like turning on LinkedIn audience network, so that there’s more inventory to show in. You could also do that by bidding up. You can also do that by bidding CPM because you’re kind of bypassing the auction at that point. You’re taking the risk away from LinkedIn and showing your ads. And then of course, you can show more impressions to the same users by using this strategy as well. Every once in a while, I’ll hear someone advocate using the Rotate ads evenly function. And I came from the world of Google Ads many, many years ago. Because of that, I definitely understand the viewpoint, I understand that if you launch two ads, you want them to show as evenly as possible. So you get a really good AB test and your data makes a lot of sense. The idea of this is really alluring to us data geeks. But on this instance, don’t be fooled. I do not recommend using the Rotate evenly function. The default that LinkedIn puts you into is optimizing your ads for the best click through rate. And obviously, the higher click through rate doesn’t necessarily mean that it’s going to convert at a higher rate. But what it does, a higher click through rate qualifies you for better pricing. And so if you will let LinkedIn do that, you won’t pay as much. So this rotate evenly function, I oftentimes call it the charge me more and show me less button. The way that it works is it doesn’t actually show your ads evenly. So it’s not accomplishing the purpose of what you’re actually after. What it does is it enters both of your ad creatives into the auction evenly the same number of times, but it doesn’t show it evenly. And there’s a big difference here. The difference is that each creative is given its own relevancy score by LinkedIn. And assuming that the campaign has the same bids for both ads, the one ad with the poorer relevancy score will lose more auctions, meaning that you’ll get less traffic every time that ad is shown. And the auctions that it does win, you won’t win at as favorable of a price. So when you’re doing this rotate evenly, assuming that one of your ads doesn’t perform quite as well, you will be shown less, you’ll get fewer impressions while you’re doing it, and your cost per click will rise. So certainly not what I like. This is one of those features that I just can’t figure out a reason why I would ever use it. So my definite rule is don’t use rotate evenly. It’s never a good idea. Certainly, if any of you have a good use case for this, please let me know. I would love to change my tune at some point. And I don’t know if I’m allowed to share this, but I’m going to anyway, I know LinkedIn is working on a real AB testing solution where it won’t be affected by auction dynamics. And you actually will be able to do a split test of ads. So I’m excited to hear more about this when LinkedIn releases it. But as for right now, just use the default, which is optimized towards your click through rate. And we realize what that means is one of your ads isn’t going to get as many impressions it won’t be a perfect AB test. But you’ll get better pricing along the way, which is great.

24:37
So now we get into AB testing strategy. And I apologize, I’m going to get really geeky here, so bear with me. So you now know why we run two ads per campaign. But you might be asking something like, So does this mean I’m going to be writing thousands of unique different ad copy for each of my campaigns? And that answer is no, absolutely not. For all of your campaigns where you’re targeting a single persona, you’re going to put the same two ads in each campaign. Now from LinkedIn perspective, they will be separate ads, each on the back end will have its own creative ID. That means you as an advertiser, you’re going to have ad A in every campaign. Here is my intro. Here’s my title. Here’s my headline. Here’s my landing page. And you’re going to put it in separately, creating a whole separate ad. But, when you’re looking at the data, everything tells you across the whole account that there are only two ads running. All intros of ad A are the same. All headlines for ad A, all destination URLs, aside from tracking parameters for ad A are all the same. And ditto for B. And because you have an individual tracking parameter in every single one of these URLs, if you’re a really sophisticated advertiser, you’re going to be passing all of those separate tracking parameters into your CRM for tracking past down funnel into the sales process. Because of this, when your team closes a deal, they can come back to you and you can see the exact ad in the exact campaign that introduced that closed deal, and everyone wins. The shortcut here is if you do create a single ad that you share among all of your different campaigns, that is a shortcut. It means you’ll be building fewer individual ads. But you know, at my peak, I can get each ad created in less than a minute if it’s, you know, text ad or sponsored content. So I’m not super worried about having to spend an extra half hour or an hour once a month creating ads. I more care about the tracking and performance aspect to it. If you do something like a boosted post, where you create the ad on your company page publicly, and then you go and promote that ad to your audience. The one cool thing there is that you preserve all of the social proof in that campaign. That means the numbers likes, the comments that occurred are all going to carry with it into all of those campaigns. But then what you lose is all of your tracking, you have no way of telling once they convert and go into the sales process, whether they came from an organic company page follower, which is going to have a higher affinity for your brand, a higher likelihood to convert versus your ads, which might be going to cold traffic, which are far less likely to convert. And you’re actually spending dollars for those. So because of that I like to create individual unique ads for every campaign, but just keep the same AB test running in each of the campaigns for data analysis purposes. If you listen to Episode 02, you’ll know how I micro segment my audiences by targeting type and other facets that you might like to break down. I really like to break down my audiences by seniority, that’s kind of my default, but you can also break them down by company size to see if enterprise responds different than mid market. So now you’ve placed the same two ads, same to you, but according to LinkedIn, they’re all different, in all of these campaigns. And in some of the campaigns because you’re optimizing towards click through rate, and sometimes LinkedIn makes the wrong decision, and wrongly awards, the losing ad the higher relevancy score. Sometimes the, let’s say ad B will win out when it should have been ad A that won. But because you’ve placed these in so many different campaigns, multiple campaigns, the exact same test, the majority of the time, LinkedIn will choose right. And the winning ad by click through rate will win out across the whole account. And then when you get all of your data into Excel, a single pivot table will aggregate and tell you whether ad A or ad B how it performed across the entire account. Realizing each of these individual campaigns, let’s say they only spent $20, $40, $100. Each one individually wouldn’t tell you much. You would look at very small sample sizes. But because you aggregate all of the performance of ad A and B across the entire account for that time period, now you’re getting very, very quickly, very relevant and statistically significant data about your click through rates. Telling you very quickly, what messaging your audience likes, dislikes, what they resonate with. And, like I said to significance, this is really fast to do. If you’re wondering how I do this in Excel, what I do is aggregate each individual ads performance. And I do this because I know that let’s say if I’m running a messaging test in my intros, for instance, so everything else is the same between two sponsored content. The image is the same, the headline is the same. But I’m doing a test just on my messaging in the intro, and this is a test I like to do very often. What I’ll do is, I understand that only the intros are different. So I’m going to in my pivot table, break down all performance by intro, and then I’ll break out and understand, you know, intro A, intro B, all of their performance will be there. And I can very quickly and easily tell which of those two ads wins. You can make this really easy to by putting something in the ad name that maybe includes your tracking parameters or includes a designation of whether this is ad A or ad B. So one test that I might recommend doing is in my intro, maybe I’m using a motivation that’s aspirational, like I’m gonna make you feel like the hero if you take the action I want you to. And I might test that against an intro that is more fear based, like you’ll get fired or you’ll look bad if you don’t take the action that I’m I’m asking you to do. Every one of my aspirational ads, I might put some kind of designation in the ad name, which is optional on sponsored content. And I might put the day’s date the ad was launched. And the letter A and the fear based one might be today’s date, and the letter B. We’re actually a little bit nervous. Hear on tracking. So we have a unique tracking code for every single ad we launch that we based off of the day’s date, and the source and the medium. So we would do that. We’d have our unique tracking code, and then the letter A, or unique tracking code, and B. So now when we build a pivot table, or even in our dashboard in Google Data Studio, we can aggregate all of our performance metrics by the ad was launched and whether it was an A or a B. And then it becomes really easy to test and tune, because you can go and say, Hey, it looks like you know, after four days, we found the ad B is way outperforming or converts much higher. So we can go and say let’s go pause every ad that has B in it, and we’ll know very quickly, what’s performing and what’s not. Alright, coming right up is our resources section. So stick around.

31:56
Thank you for listening to the LinkedIn ad show. Hungry for more? AJ Wilcox, take it away.

32:06
All right, I’ve got a couple of great resources for you. If you’re looking to get into LinkedIn ads, there is a course that will stick here. It’s actually the course on LinkedInlearning.com. It was by me. And I walk you through essentially the first hour and a half of what I would teach if you hired me directly to train your team or your company. And because it’s on LinkedIn learning, it’s only $25. Or if any of you on your team have LinkedIn premium, it’s free. So definitely check that out. The link is in the show notes. Also, whatever podcast player you are listening on, please hit that subscribe button. I want to have all of these episodes right in your earholes. So you can be a super sophisticated advertiser on LinkedIn. So definitely subscribe, and then a plea for you. Please, please, please do rate and review this podcast on whatever player or network you’re listening to it on. This is still new, and so the more ratings and reviews we have, the more people will get to enjoy this podcast and the more friends you’ll get to geek out with over the watercooler. If you’ve got ideas for the show, any sort of episode that you want to hear about, please reach out to us at podcast@b2linked com. We’d love to hear from you. And of course, as we’re building the show family, it’s so great to hear from you, what you’re looking for, and what we can deliver. Alright, we’ll see you back here next week. We are cheering you on in your LinkedIn Ads initiatives.

‍