Jul 28, 2020
Show Resources:
How Does the LinkedIn Auction Work?
LinkedIn Learning course about LinkedIn Ads by AJ Wilcox: LinkedIn Advertising Course
Contact us at Podcast@B2Linked.com with ideas for what you'd like AJ to cover.
Show Transcript:
LinkedIn Ads are unapologetically expensive, but what makes them that way? Let's break it down.
Welcome to the LinkedIn Ads Show. Here's your host, AJ Wilcox.
0:19
Hey there LinkedIn Ads fanatics. I'm so tired of hearing people say
the same thing about LinkedIn Ads that they've said forever. I've
heard it's really expensive. Why is that? I'm gonna dive into
exactly why that is and what you can do about it. Let's hit it.
0:34
So how does LinkedIn decide how much you pay for a click? Well,
there are two big pieces that weigh in here. The first is what your
bid is. Now you can bid by either cost per click, cost per thousand
impressions, or automated bid, which really is just a cost per
thousand impression bid that you don't have to babysit as much. So
we're going to concentrate mostly on cost per click here because
I'm assuming most of you listening are bidding are really smart
advertisers and are bidding on a max cost per click model. The next
piece that factors in here is your relevancy score. Now, LinkedIn
calls this relevancy score, Google calls it quality score, Facebook
calls it relevance score. And it's all the same concept. It's
essentially, how often do I make the network money when they show
my ads. So each individual ad creative gets a relevancy score. And
it has to come up with this. So let's say you launch a brand new
ad. Well, LinkedIn doesn't know what relevancy score to give it.
And so it's probably going to borrow from other ads in the
campaign. If all the other ads have been performing really well,
then it's probably going to get an assumption of a really high
relevancy score. If the others are performing really poorly, it
will probably assume this one will perform poorly too. So other ads
in the same campaign. But let's say this is a brand new campaign.
LinkedIn doesn't know how it's going to work, then I think they
might either borrow from other campaigns that are running
concurrently, or borrow from the account in the whole. So your
relevancy score is mostly made up of what your click through rate
is. And this makes a lot of sense, because if I have a 1% click
through rate on my ads, and my competitor has a half of a percent
for every single time that LinkedIn shows my ad, or my competitors
ad, there is an opportunity cost in it for them in saying, hey,
even if AJ is, you know, he has a 1% click through rate, he's only
willing to bid $5 per click, his competitor has half that, and the
competitors willing to bid $8, well, LinkedIn looks at it and goes
well, for every hundred times. I show AJ's ad, I'm gonna make $5
and every time I show as competitors ad for 100 times, I'm going to
make $4 so even though I'm bidding less, LinkedIn still says that
I'm making them more money than my competitor. So this is how this
relevancy score works. It was really pioneered by Google back in
the early Google AdWords. Now Google Ads days. And it makes a lot
of sense. So it's based off of how often you make LinkedIn money.
Well, that is pretty much your click through rate, it's for every
time I show the ad, how likely is someone to click it, and
therefore LinkedIn get paid, but they're also paying attention to
things like comments, likes, shares, and even negative member
feedback that they can provide on ads. So you can actually find
what your relevancy score is by running a Campaign Performance
Report. And then look at it's at the time of recording column BI is
called campaign quality score. And so this is a snapshot at any
given time, you can see what your campaign quality score is. But I
would argue this isn't all that helpful, because this is something
that changes quite regularly. And you can't always wait around to
watch if this number has changed up or down. I think you just need
to go by maybe what you're ad performance is looking like today.
And if performance looks bad today, and over the last four days,
you can assume it probably has a poor relevancy score, and you can
try something else.
4:10
Auction
So the auction, LinkedIn calls it a second price auction. And the way it works mathematically is we're going to put you versus a competitor. So you have a bid. And we're going to multiply the bid by the relevant score to get a combined score. And we'll talk about what that means here in a minute. Okay, so you're bidding $8 and you have a relevancy score of 6, because you've got a click through rate that's slightly above benchmark because you created a good ad. So you take your $8 bid times your 6 relevancy score, and that gives you 48. So that's your combined score. And at this point, that doesn't mean anything. Now your competitor comes along and bids $12. So they're bidding $4 more per click than you and I know you're sweating right now you're thinking for sure I'm going to lose this bid. But they have a relevancy score of 3, because their click through rates are well below the average. And they're just pulling teeth to try to get LinkedIn to give them traffic. So a competitor bid of 12 times the relevancy score of 3, they're getting a combined score of 36. Now, this is where the math gets a little bit tricky, but you have a combined score of 48. your competitor has a combined score of 36. So what the auction does is it divides the losers combined score that's 36, in this case, divided by your the winners relevancy score, which is a 6, and that equals $6. So that means that that really is how high that the winner would have to have bid, you know, you bid $8, but you actually only needed to bid $6 to beat the loser. And because it's a second price auction, what you need to do is pay one more cent than that. So in this case, you bid $8, but you're only going to pay $6 plus the one cent to beat them by a cent. So that is how a second price auction works. Again, this was really pioneered by Google. This is how digital advertising has been forever. So does this actually happen in practice? Well, not from my experience. And I'm not saying that LinkedIn is disingenuous here in what type of auction they have. But go and do this experiment for me, go for half a day, bid $50 on a cost per click basis. And tell me if you only paid the $12 and one cent that it would take to beat second place. Chances are, you probably didn't. You bid $50 aand you probably paid something like $48.30 for it. So I'm not sure if it's because it's not really a second price auction. Or maybe it's because there's enough competition around all the audiences they are always going to pay really close that there's always someone nipping at your heels with better performance or bidding higher. So the edge cases here, you may not want to download this report, you know, three times a day to see how your campaigns are doing and realizing that the campaign report will only give you the campaign relevancy score, not at the individual ad level. So it's helpful to understand what it looks like when you're really high or really low. So edge cases here, if you have a relevancy score of 10, you are getting tons of traffic, even if you bid all the way down to the floor. And because you're performing so well, with such a high click through rate, if the floor didn't exist, you should be able to pay less than the floor. But because there is a floor on everything, you can't do that. So a little hint here, when we get into the bidding budgeting episode, we'll talk about how you can use CPM bidding to pay even below the artificial floor. So that's a relevancy score of 10. Things are looking great. You can keep lowering your bid and still spend all of your money very efficiently. What if you get a relevancy score of 1? What that means is you were so uncompetitive in the auction that LinkedIn won't even show your ads. So you could bid you know $30, $40, $50 a click and it's not even until you're paying super huge when you start to overcome the relevancy scores of your competition. So you're most likely going to land between there.
8:12
The floor
So I mentioned that LinkedIn has a floor. And the concept of the bidding floor was really started to keep the riffraff out. I think at this time, LinkedIn probably saw other ad networks that were that they had a lot of crap on them, think like the belly fat ads on Facebook. And so what they did is they said, Hey, we know that competition is going to rise prices normally, but what we're going to do is set a arbitrary floor here that you're not going to be willing to pay if you're just being opportunistic, trying to get cheap traffic. And so they make sure only advertisers are coming to the platform that are serious. And then once you hit the floor, then competition kind of rises normally. And Google did this really well by you know, initially in the early days of AdWords, they set a five cent floor on every single keyword out there. And so at the early days, you could bid five cents and still get traffic. And Facebook followed suit. I don't know what Facebook's minimum was because they've always gone CPM, but it was really low. And then they let everyone come and have success on their platform, and then go and tell their friends who then jumped on and used it. And competition kind of raised naturally for everyone. And I call these really generous platforms. Because, you know, early on early days before there was competition, you really couldn't do any wrong. You'd come to the network and inevitably have success unless you were a crappy marketer. Okay, so the way this goes history was back in 2008, LinkedIn released a $2 floor for all text ads. And that was, you know, basically from 2008 till they launched in 2013, the sponsored content ads, and I was the first person to launch sponsored content ads, which was awesome. I got to use them before there was a significant floor. So back then when sponsored content came out, there was somewhere around a $4.50 cent floor. I saw it down to $4.25 for some really uncompetitive audiences. I saw it, you know, $6.75ish for more competitive audiences. And it stayed that way for quite a while. And then just recently, last year in 2019, when LinkedIn released their objective based advertising platform, prices increased about 35%. So what used to be a $4.50 floor is now $6.08. The way this works is back before objective based advertising, we had a website visits objective, and that was how you would just get traffic to your website. And that is currently called legacy. But back then it was just called website visits. After objective based advertising, they moved to where now website visits has a 35% increase in the floor, so you're going to have to bid higher and it Raise the auction for everyone. So a little trick here, if you're still sending website visits traffic, and let's say, back in 2018, you had launched some of these campaigns, try just editing the campaign to repurpose it so that you can keep it as a legacy campaign. That will be something really cool to help you get, you know, 35% less cost per click at least anywhere near the floor, as your competition is getting. So you might ask, well, if you're not bidding right at the floor, what's the harm in LinkedIn raising the floor price? The challenge here is that when suggested bids and the floor price go up, people begin by bidding higher, either because they're forced to now or because it's suggested and they just don't know better, so they're going to take the platform's recommendation. And now when everyone is bidding slightly higher, competition gets higher, and those who aren't bidding as high start to lose traffic or maybe if their relevancy score isn't doing as well. So now prices have been raised for everyone and It feels quite artificial. And I'm sure it's very obvious to you, I'm not a fan of artificial price hikes, I'm not a fan of even a floor price, I honestly believe that you should have a platform that is really inexpensive to get in, and then you let competition naturally raise the prices up. But since we do have a floor, it's helpful to understand what else may contribute to LinkedIn's higher prices.
12:24
Low inventory from low use
So aside from the artificial floors, which I obviously don't support, one of the things is low inventory from just general low use of the platform. So absolutely nothing against LinkedIn here. I love it as a platform, both paid and organic. But it is also the network that you don't think to spend as much time on because it's professional. So meanwhile, people are going and visiting Facebook 18 times a day, on average they're probably going to LinkedIn three or four times a month. Now there are certainly more active users. But with the vast majority of users checking in, not very often, and especially early on with LinkedIn, it was a platform that people only came back to every like six months to update their resume, or when they were looking for a job. This ended up creating really low inventory.
13:09
Low variety of types of professionals
So the few advertisers who were bidding on that inventory, it cost more for them. Now, it's really exciting that LinkedIn is getting used more and more. So we're seeing a lot more inventory open up, probably keeping prices relatively steady, even though a lot of new advertisers are coming in. Something else that tends to lead to these high prices, is the fact that there's a low variety of the types of professionals and businesses that there are out there. So follow me on this. If you are bidding on keywords like with Google, your competition is built naturally over billions or even trillions of combinations of words in the written language. And that's pretty cool. That means you can keep prices pretty low for a long period of time. Facebook is more audience focused, but they built over consumer based information. And other facets. So you can split any individual person up into lots of different segments, you know, are they a cat person? Do they love coffee? Are they into music? And so you have all of these different ways of categorizing a human. And it means that you can spread that competition out really naturally. On the other hand, LinkedIn has a very few set of facets that you should be able to target someone by. There are only so many job titles out there. There's only so many levels of seniority, there are only so many finite numbers of companies of certain sizes. And so because of that competition is going to naturally group around the more competitive aspects of who someone is. Certainly if you are targeting the highest rank, so like a CEO of large enterprise companies, that's going to be a really expensive click.
14:52
High competition from big brands
You also have high competition from really big brands because they have these deals that are worth millions of dollars to them. The same reason why if you want to try to outbid Salesforce in Google ads, you're going to be spending $80, $90 per click on something like sales CRMs. Really expensive because you have a company like Salesforce, who are such finely tuned marketers, and they have a large lifetime value. They know what they can afford to spend. And so they're willing to spend big and bid high. And so you've got this high competition from brands who the click is probably worth much more to them than it is to you. And so it drives competition up higher for everyone.
15:33
Competition will naturally increase over time
The fourth cause here to higher prices, his competition is naturally going to rise over time. More and more people hear about and consider the platform and then they start having conversations with their bosses about what channel should we get involved in. And inevitably, they're going to think, hey, we should check out LinkedIn, I've heard some good things. And also advertisers are getting more sophisticated at tracking. And as they do this, they're going to find more and more valuable In this really highly targeted, really high quality traffic, and they're going to make the decision to scale and keep spending and maybe even keep bidding higher as they start to get into Salesforce land.
16:11
Timing of day
The fifth cause of high prices here is actually the time of day. And you wouldn't think about this usually. But there are certain times of the day when people tend to be on LinkedIn. And there are certain times when they tend not to be. And what happens is, you know, advertisers will go in with a budget with certain bids. Sometimes during the day, they will outspend their budget, and they will drop out of the auction. So you can imagine some advertiser demand may taper off towards the end of the day. So that may affect things. When people are online more, it creates more added inventory, bringing overall prices down. And when people aren't there, let's say in the middle of the night, and everyone's bidding for that just few insomniac visitors who might come, those cost per click are going to be really high and it's probably not a great time to be bidding where people their not really in their right mind. There's also this issue of time of year where, you know, in December, people are really blowing their budgets up because they're trying to finish the year, the quarter, the month really strong. And so you have these big advertisers who are just you know, draining huge budgets. And in January all that kind of disappears. January's traditionally a very good month for LinkedIn advertisers. No one's in a hurry. It's the beginning of the year where people are back in the office and don't have things on their calendars. They're willing to entertain. They now have higher budgets, or they have new fresh budgets that they can start to think about allocating. So timing of day timing and general seasonality plays in here nicely.
17:40
Bidding each other up
Okay, number six cause for high prices is of course bidding each other up. So one competitor says, hey, I'm getting less traffic than I'd like. I'm gonna go ahead and raise my bids by you know 5% or 10%. And you'll at some point, go Ah, you know what, I'm not getting traffic anymore. I better go on raise my bids. And so you tend to bid each other up over time.
18:03
Bidding yourself up
Seventh cause of high prices is bidding yourself up. So when you have a single account and two campaigns who are targeting the same user, you are not going to bid yourself up, you're not providing yourself competition, which is really helpful to know. But there are companies out there who have multiple accounts, let's say using an example like Microsoft, it doesn't make sense for the Xbox team and the Microsoft Office team to be using the same account. So they might theoretically be targeting the same user, even though it's both coming from Microsoft. And in that case, they would bid each other up, you know, Xbox says I'm willing to bid $6. And Microsoft Office says I'm willing to bid $6.50 and you can see competition is, you know, the same company is now driving their own costs up.
18:52
Platform generosity
I mentioned this concept of platform generosity a little earlier. And I want to throw this out there as a concept I've been thinking about over time, and the way I qualify it is, it's the feeling that the network really supports you. And it inspires confidence to test new things, knowing that it's not going to embarrass you or cause you to make expensive mistakes. So Google and Facebook definitely have a platform generosity. I've tested this with Facebook, where I go in and just say, okay, I'm going to take Facebook's recommendations in creating a new ad set. And I'm going to bid what they say to bid. And I'm going to target how they say to target. And generally I'm going to have pretty good performance unless my ads suck. Google's kind of the same way. It's certainly getting much tougher now that everyone on the planet is advertising on Google. So quite a bit more competition there. I don't feel like LinkedIn has a high amount of platform generosity, because if you go down and create a brand new campaign, you're going to tick the audience expansion box which generally muddies your targeting and decreases your lead quality. You're going to bid automated bid and that's going to oftentimes You know, depending on your click through rate, but if you have an average click through rate, you're going to pay much more for that click, then you would have if you bid CPC. So that's one. The other one is if you do decide to pay by cost per click, which tends to be the lowest cost most of the time, LinkedIn will recommend a really insanely high bid to you, regardless of what your budget is. And they'll have a floor that's quite high in a suggested bid range that's quite high. So I would absolutely love to see LinkedIn start focusing on platform generosity. And in talking to LinkedIn's product people, this is certainly not what they intend. They don't intend it to come across like we're, you know, less focused on the user having success and more on padding the bottom line, but I think an extra focus here and attitude shift could go a long way in helping people understand that if I test something on LinkedIn, I'm not going to be embarrassed to my boss because I overspent the budget or just paid way too much. Okay, here's a quick sponsor break, and then we'll dive into what you can do to pay less on LinkedIn.
21:11
Thank you for listening to the LinkedIn Ads Show. Hungry for more?
AJ Wilcox, take it away.
21:22
Alright, let's jump into this. So now you know the history. And if
you're paying too much for LinkedIn Ads traffic, here's what you
can actually go and do about it. So a few different scenarios here.
Let's say you're not being competitive enough. If you're running
text ads falling from the first position to the second ad position,
it's just a few pixels below. So there's really not a huge problem
to bidding low with text ads, you'll fall down to the next position
down and still be above the fold. Sponsored content is a little bit
different of an animal here though, because the first ad position
starts as the second item in their newsfeed. And if you drop down
to the second position, that's five more slots down. So now you'll
be at slot six for your ad position. And if you're up at the very
top, you're probably going to have a much higher click through rate
than you would if you fell down one or two positions. We've
actually done a test here where we found that if we had a 1.2%
click through rate in the first position, it actually dropped down
to a 0.2%. So it dropped an entire full percent, just going from
slot one down to slot two. It's important to understand here that
your lead quality does not decrease when you drop down positions or
when you bid less. The common thought is that it does, but in a lot
of testing we've done we have not found lead quality, as long as we
set our targeting really tight, we have not found lead quality to
fail us when we're bidding low. If you are bidding by cost per
click, LinkedIn is much more incentivized to show you in higher
positions because that's where the ads get clicked. If you are
bidding by cost per impression, then you will tend to fall down
positions much more quickly. So if you want to pay less, there were
three real mechanisms you can do. The first is you can bid less.
The second is you can launch ads that get a better click through
rate. And the third is you can actually change your bidding
strategy to bid smarter. So I'll put it out there this way. If you
took LinkedIn's recommendations, and you bid really high, let's say
you bid, you know $15 a click and you have a $50 per day budget.
That means you're only going to get just over three clicks a day
out of your budget. And you could probably hit that in the first
half of the day. So imagine you spent all $50 in the first half of
the day and then for the last half of the day, you just got
nothing, your account was pretty much off. If you would have been
half that you could have paid half the price and gotten double the
amount of clicks and it might have actually gone all day, you know,
maybe it still cuts off earlier in the day. So be smarter about the
way you bid, pay attention to did I actually hit my budget during
the mid day? I like to use this little rhyme, if you hit your
budget during the day, you paid too much for clicks along the
way.
24:19
Armchair Economist
So now I get to sit in my in my armchair here and play the role of an armchair economist. Now I know LinkedIn has Stanford and Harvard MBAs who know way more about economics than I do. But I'm going to attempt to postulate here some ideas that I think LinkedIn could do to improve their pricing model. So first off, since the pricing has been set by artificial floors, I would love to see LinkedIn artificially cut its floor prices in half. And what would happen is, if we're already past the level of competition, where people are paying more than the floor, which we are in most cases I'm sure, LinkedIn has no need to worry about the riffraff. The floor will have already kept them out. And anyone who did attempt to come in and bid too low for what they're actually offering, they wouldn't be shown anyway. And of course, the better advertisers, they would be paying close attention to okay, how low do I bid before my traffic cuts off? What this would do is it would cause everyone on the network to come and have success and go tell their friends, just like early Google and Facebook. And I postulate that if the only thing that people knew about LinkedIn Ads, who hadn't tried it was no longer "I've heard it too expensive". We'd probably see LinkedIn ads gain mass adoption, and then pricing would come back up naturally. And I think it would be a lot better received. And I think what mass adoption on LinkedIn Ads would look like is everyone in B2B, LinkedIn would be the de facto platform that they would jump on and learn when they were very first coming out of college or be the first platform they try, rather than trying Google first, then Facebook and then kind of failing over to LinkedIn when Facebook got too expensive for them. And certainly, if this were the case, the Microsoft stock price would take a hit for a quarter or two. But then pricing would come back up naturally rather than artificially. And then everyone in business to business would recognize LinkedIn as a tier one platform and not a tier two or tier three. And this is not without precedent, because we've already seen that LinkedIn knows the effect of artificially lowering prices. When I very first tried sponsored Inmail back before it was in the platform, you had to go directly through a sales rep. It was a $3 per person you sent to and since it came out on the platform as self service, it turned out to be a I think they started the floor somewhere around 25 cents to send it and the floor now has been dropped twice. So they've artificially lowered the price of sponsored in mail twice, maybe even three times now,which is fantastic. We've seen much better performance from our sponsored in mail. And I'm so grateful that LinkedIn did that. And then even just recently, dynamic ads saw the same shift. And this one wasn't quite artificial. What happened is LinkedIn gave up that inventory that they had given to the programmatic exchanges for just general 300x250 banner ads across the site. And what that did is by dropping that inventory, it opened up the inventory to all of us advertisers to use both text ads, and dynamic ads that occupy that same space. So now dynamic ads dropped their prices more than half. It was actually down to about a third to a half of what they were before, which is fantastic. And now I can actually recommend using dynamic ads again, back then I couldn't because they cost more than than sponsored content, and they had really low engagement rates like text ads. So I would absolutely love to get rid of the high floors. I would love to see sponsored content. Sure text ads has always had a $2 floor. I'd love to see sponsored content adopt the $2 floor as well. And of course, if that doesn't get me any traffic, so be it, then I have to raise my floor. And of course the platform could inform you of that. Okay, I've got the episode resources for you coming right up, so stick around.
28:20
Thank you for listening to the LinkedIn Ads show. Hungry for more?
AJ Wilcox, take it away.
28:30
Resources
Okay, I've got a couple of good resources for you. There's a really cool video back from 2013 where LinkedIn explains how their auction works. So if you weren't quite following with like the what do I divide by what and what got multiplied together to decide how much I pay? Go watch this video. It's there in your show notes. I've also linked to a YouTube video that I created where it discusses both bidding and budgeting strategies, and definitely check that one out. Next, I've got a course on LinkedIn learning com. It's called Advertising on LinkedIn. Definitely check that one out. The link is in the show notes here for you. And it's a really inexpensive way to learn from the ground up how to advertise in the most efficient way. As always, I would love it if whatever player you're on you do subscribe and then you make sure to rate and review as soon as you know how much value you're getting out of this. Make sure you leave a great review telling other people that they can find a lot of good stuff out of this podcast that I hope that you're getting. Finally, definitely email us Podcast@B2Linked.com with any ideas you have for future shows, we'd love to take your your recommendations here your advice, would love to cater to you. So anything you're looking to learn, I'd love to hear. I will see you back here next week. And I am cheering you on in your LinkedIn ads initiatives.