About Susan Wenograd
Susan Wenograd is responsible for the management and optimisation of client’s media buys across multiple platforms in order to deliver the best ROI possible, with a speciality in Facebook Ads. Previous to her current role, she spent over 10 years both client-side and agency-side, in roles that spanned media management and team management for brands such as General Motors, CircuitCity.com and many others.
In her Learn Inbound talk, Susan will review how to skillfully craft a sales funnel that will get you more purchases and more leads. Leveraging the multiple campaign options in Facebook Ads can help you drive a string of objectives that will work together towards your goal, without selling your soul to one Campaign type that keeps struggling. By understanding Facebook’s optimisation and objectives, you can use your target market expertise and creative to get more in control of your outcomes.
- How to create a sales funnel that will help you drive more purchases and leads
- Learn how to leverage the multiple campaign options in Facebook Ads to drive your objectives
- How to better understand Facebook’s optimisation and objectives to get more in control of your outcomes
I heard that there's not a lot of talk about Facebook advertising here, so we are going to change that this evening. Before I start, and you're wondering who this fast-talking person dressed in green, green fingernails, I am so American, for St. Patrick's Day. Me and my lovely husband, who's here in the crowd tonight, give him a cheer. My daughter, we live in a magical place called Richmond, Virginia, they have wonderful beer there. If you ever come to Virginia give me a holler, I'd be happy to take you out.
I do all kinds of geeky stuff, speak at all kinds of places, right? Talk, do all that kind of stuff. We're going to cover some stuff tonight. We're going to talk a little bit about how advertisers typically do Facebook. I covered this, because it's usually the wrong way. We're going to talk about funnels, and people are like "What are those? Are we making something?" We're going to talk about creative options and the ways they drive me insane. We're going to talk a little bit about going beyond selling, and what Facebook has planned for the future. When I'm all perky about Facebook ads, and I'm like "They can work great." This is typically the look that I get from potential clients, they're like "Those things suck, they don't work, we wasted money." I'm going to talk a little bit about how they typically use it, and why they see these results.
Usually, what they do is they open up a Facebook account, and they choose a conversion focused campaign, and that is what they run. That doesn't work well. We're going to talk about why that doesn't work well. Take a step back, the reality of the Facebook auction is that your targeted population that it tells you is more than you're actually going to reach. It's going to say something like this "You are going to reach 1.4 million people" it's not really true. It's like that when you pick the actual targeting, but in reality, here's that 1.4 million people, and in that pool everybody is likely to do something different. They're not all likely to actually convert. Some are likely to watch video. Some are likely to engage with the post. Then, finally there is that little piece that's like to convert.
When you pick that conversion campaign, you are picking actually a very small subset of that 1.4 million, and, by the way, conversion campaigns have the highest CPMs, they're the most expensive campaign type you can run on Facebook. Essentially, what you're doing is you're picking the most expensive type of campaign, then you're trying to optimize towards a very high value action, which by the way, if you run a conversion campaign, Facebook wants you to have at least 50 conversion events firing per ad set per week. That is a lot. Especially if you are not a high volume e-commerce person.
Most advertisers do not meet that threshold, that's another problem. Now you've got expensive media with not enough data for Facebook, and the campaigns do shitty, and people are like "Facebook sucks." And it's like "Well, no, it's just they don't really explain how to do it well." Here's why you might be doing it wrong. Again, conversion only campaigns are expensive, they limit your audience, and usually people are going to need more than one experience to buy or convert. A lot of people that I run Facebook ads for, they're used to paid search. They're like "Paid search, it converts at 8% and whatever." Yeah, but those are people that are telling you they want what you have.
Facebook is not the same thing, these are people that have no idea who you are, no idea why they should buy your stuff, so you're having to convince them that's a much longer sale cycle. When I talk about funnels, people are like "Funnel cake?" No, I wish. I don't know if funnel cake is a thing here, deep-fried dough, it's amazing. But, no, this is a less fun funnel.
When we talk about funnels in Facebook, essentially what we're doing is we're talking about that long sale cycle. We're figuring out the awareness piece, and this all goes back to marketing 101, which we've kind of abandoned, because we all got used to paid search, and kind of that quick hit of people that come to the site, buy within a couple days and you're done. But, when you talk about funnels, you're talking about kind of the large, cheaper target of awareness. Where the breakdown happens is that most users are in this awareness stage on Facebook, but advertisers are using conversion campaigns. These are people that don't know who you are, but you're using the most expensive media possible to make them aware of you. That's usually where the largest disconnect happens. After awareness, you go to the consideration stage, which is people that may be aware of your brand, they might know something about it, but they might not be quite there yet as far as purchasing or converting. Finally, you have conversion, which usually comes from the progressive messaging retargeting.
When I talk about this, people are like "That's marketing 101," and it's like "Yeah," but, if you ever look at, when you set up a Facebook campaign, Facebook wants you to do this. They have awareness campaigns, they have consideration campaign types, and they have conversion campaign types. This is what they actually want you to do, they just do a shit job of explaining it. We're going to explain it.
Funnels are your friends. Let's talk about awareness, then kind of the warm audience, and then the hot audience, looking at that funnel. When you're looking at the awareness part, you want to do that for the cheapest amount possible. Yes, there are times where people will convert in that awareness stage, it does happen, it's not usually the reason that you run it, however. They do have campaign types that are a lower cost option, kind of lower stakes. Things like video views, reach campaigns, website clicks, post engagement. These are things that are a little more passive in nature. You're not really asking people to leave the site to do anything, the site being Facebook. It's more of a passive kind of thing. They might do website click, but there are things that really kind of keep them on Facebook, and start to make them aware of you.
Then, you have your warm audience. Usually for these, these are going to be things like you could probably jump to a conversion campaign, depending on what you're selling at that point, website clicks, sometimes reach. But, usually what you're doing is you're basing this warm audience off of people that did something in the awareness stage, and I'm going to show you how to do this. Then, finally, obviously, retargeting. That's kind of, I think, the first step that people do when they start with Facebook, is they do retargeting first, and they're kind of like they don't know what to do after that. Retargeting is usually the part that marketers understand best. A lot of times when I bring this up to clients, they're like "Wait, I got to run like a video views campaign? Then I got to run retargeting campaign? That's stupid, that's like two different campaign types, when I could just run a conversion campaign. That sounds great in theory, but I just don't think that that's going to be cheaper." Let's talk about facts, shall we?
Here's an example of a client that had a conversion campaign, they were running a 1% to 2% lookalike audience, another conversion campaign, 3% to 5% audience. What I said is "Okay, let's do this. Let's run a video view campaign," super cheap, right? "Then, we're going to re-target to people that watched 10 seconds of that video, using a conversion campaign, and let's just see what it does." When you look at the results, naturally, the 1% to 2% has the highest conversion, which is great, but it's not really scalable. When you're dealing with that small of a lookalike audience, it tends to kind of fall off over time, because it's like we said before, it's a really small audience. Then, you get into the 3% to 5%, which was 2.6 return on spend. Video views, we actually had a 326% return on spend. Like I said, you will sometimes see people buy in that super cheap awareness stage, which is kind of just a really nice perk, but then when you re-target to them, you also got a 262% ROAS. The return on spend was really healthy, it actually outperformed the 3% to 5% lookalike in the conversion campaign setup. Overall, your effective CPM comes in a lot less than just running that expensive conversion tactic alone.
The thing about funnels is they can get complex. Especially if you have a multistage thing where people sign up for registrations first, then you re-market to then to become a sign on page subscriber, there can be a lot of steps with a lot of different audiences that you're dealing with. I'm going to take you through how I do this with clients, because it actually breaks it down into super manageable pieces.
First thing I do is I do a creative audit. The biggest thing I see, people get super excited about all the targeting in Facebook, and they realize they really don't have anything to offer people. I usually start with "Let's see what we actually have before we start going off and making shining objects." "This is amazing, we can target fit moms, who don't smoke, and who like pot." Whatever. "You can do that, but like I don't know that you want to, unless you have something you want to offer them."
First, we start with what do you actually have? I usually make a list of things that could be used as downloadable content, or something that gets them into your funnel. It might be something like a PDF checklist, how to video, some type of quiz you have, if you have a webinar maybe if it's short, payment calculators, anything like that. You might be surprised how much content you have that you can repurpose. Because, when you have to start going through, you've probably created content for SEO purposes, or other reasons, that can be easily repackaged into something that could work well in a Facebook ad.
Usually, we start with this, and then we figure out what are the user commitment levels required with this? Low commitment means little to no effort, it's passive. They're watching a video, they're just kind of passively doing something, it's not requiring a lot of them. The medium is kind of low time commitment, but the user is in control of the consumption. Usually, it's trading value for their email address, it might be something gated, something that requires a little bit more of them, but they're not necessarily buying anything, or making a huge time commitment. Then, finally, high, so time and effort for commitment by the user. Things like webinars, 45-minute webinar, people aren't going to start with that when they see you, but if they've downloaded your free PDF, that's super awesome, and you're like "Hey, by the way, we have this webinar." It's a much more natural progression, because they know that you offer value.
Now that we kind of have the "What do we want to offer as bait?" worked out, let's talk a little bit about campaign types. We discussed already the expense involved in the different kinds. But, I also encourage people to look at those assets that we just listed, and figure out what might actually be a good, cheap way to run them. Instead of just kind of knee-jerking to the conversion campaign type, which it makes sense, because Facebook is like "Hey, what do you want to do?" "Well, I want a conversion." Of course, you pick that.
But, start thinking about your assets, and how they align with the campaign types they offer. If you have video, you may want to start with video views campaigns. Super cheap, way cheaper than conversion, a good way to build some warm audiences, which we'll talk about. Downloadable content, if you have not run lead ad campaigns, they can do fantastic. If you have healthy conversion data, feel free to test conversion campaigns. But, again, whatever you know is that conversion, you'd have to get 50 of them a week per ad set. It's a pretty high threshold, so just keep that in mind if you decide to go for those.
Aligning with audiences. Email lists and website traffic are usually the two main ways that people create custom audiences. You can upload your email list, you can specify all of your website traffic, you can specify from certain URLs, you can say if they visited this, but didn't go here, there's all kinds of combinations you can create, just like you would in any type of remarketing.
In the last year and a half or so, I guess it was, Facebook releases Engagement Audiences, every time I show this to my clients, they get super excited. Because, basically what it's doing is Facebook realized users don't like to leave Facebook, they don't want to click off to another site, they want to sit there and look at their cat memes, and gerbil videos, and whatever it is they're watching. They don't want to leave. What they started doing was they started creating ways to re-target these people based on what they did on Facebook, so they don't even have to go to your site anymore to become a retargeting audience.
They started doing things like how long you watched a video. If you go to create a custom audiences, and it gives you all these options, the one that you want is engagement down at the bottom. You've got how long they watched a video for? Did they fill out your lead form? Did they interact with your Canvas ad? Have they interacted with your page? Did they comment on a post? Have they responded to an event? There's all these things that you can based on their Facebook behavior. Every time I show this to clients, they're like "Oh, hell, yes." It's a whole treasure trove of data that has nothing to do with whether they ever clicked off on your site. Those are extremely powerful, especially for retargeting.
One of the things that happens with this is that people will create custom audiences off that and then they get a little overwhelmed by the lookalike section. When you create a custom audience, you can tell Facebook "Hey, I want you to go find more users like this." And they call them lookalikes. When you do that, Facebook says "Well, how much of the population do you want us to match it to?" Because you know they have like 80 bazillion users. For them, they're like "Do you want to match to 1% of the population? 2% of the population?" You can go all the way up to 10%.
1% is going to be the people that look most like the audience you have specified, but it's also the smallest pool. It works really well if you just have a smaller budget, and you're just trying to concentrate it, and get some stuff out there. If you have a really large budget, you'll burn through that 1% lookalike very, very quickly. I usually tell people "Scale your lookalikes to match up with your budget." Large budgets run through them, especially with conversion campaigns, like we said, remember, it gets to be a really small pool. But, if you have a nice big budget, 5% lookalikes can work great, 10% lookalikes can work great. Because, the data is just so large, Facebook has a lot to work with to figure out who it is they should be targeting.
Knowing all those things, it starts to make it easier to assemble this stuff, because then you already know "Okay, here are the lookalikes we want to test. We want to exclude whoever," we want to exclude people that have been to our site, to make sure that they're cold. Do we want to have different placements if we want to do the feed versus Instagram? What type of campaigns do we want to run? This is the big one. What type of campaign, don't do conversion. Then, what are we optimizing for?
If you have more than one thing, if we're going back to that whole you want to have 50 of whatever that thing is each week, when you set it up, and there's a whole field for what do you want us to optimize this ad set for? You want to pick something that has enough of them to justify running a conversions campaign. That's kind of how you start to assemble these things. Every time I show this, people are like "Oh, my God, what the hell do I start with? If I'm just starting what do I want to do?" Okay, here's my advice, if you're just starting, and you're like "Where the hell do I start?" If you want to do a conversion campaign, I would not do it, unless your audience is 1.5 million people or higher. I know it sounds really high, but it used to be that Facebook would do great with like a 300,000 to 500,000, it does better with more data now. If you can make it around 1.5, that would be great, the percentage of what that lookalike is, is going to vary greatly. It might be 1%, depending on the audience you give it, or it might be 5%, but, if you can keep it there or higher, even better.
If you're looking at it from an acquisition funnel standpoint, my favorites are if you have video, video views campaigns, and then re-target to people that watched 10 seconds or more. It works great. I almost always get conversions off of video view campaigns, and they are so damn cheap, it's fantastic. Then, obviously, remarketing. Please, do not just let people interact or go to your site, and then never speak to them again. It breaks my marketing heart when I see someone spend all this money to interact with people, and then they never speak to them again.
Another question I get a lot is what do I do about bidding? They're like "It feels creepy to just let Facebook do it, aren't they gouging me?" Here's the thing, there are some instances where manual bidding will do better, it is so rare, you guys. Believe me, I come from a paid search background, I want to control every frigging penny that I bid on something. It never works as well. I've run a bunch of tests, in general, auto-bidding does a better job. Mostly because there is a black box factor in Facebook, where they have information that we do not have. In something like paid search, it's a very one-to-one transaction. We know that it's this cost per click, they convert at this percentage, therefore this is what we should bid, it's very black and white.
Facebook is different. They know users intent, and they don't share it with us. Facebook can bid a little more smartly than we can, because we don't have all that data. They're assembling millions of data points about who is likely to buy, what they've done in the past, what they're likely to do in the future. We don't know any of that. When you start capping bids, you're just telling Facebook "I don't really care their intent, I'm just only willing to pay this," and you're going to see your results almost always go down.
New feature alert, just hot off the press, they're rolling out for e-commerce people the option to optimize to value, which is basically like optimizing to return on spend. I have been testing it with a large e-commerce client, it is magical. I love it, it does so well. Again, so weird of me to say, because I'm like "It's big brother, I want to control my bids." I'm a total fan girl over some of their auto-bidding like this. It works so, so well.
The client that I'm testing on, it's not out in every account yet, the client that I'm testing on right now, they spend about $300,000 a month in Facebook ads, they have a nice big data pool. For the ad sets that I've turned it on in March 1st through 9th, they were about 313% return on spend just in the first four or five days, they're up to 335%, and they're getting more purchases, which is unusual. Usually, what we see in some things, especially in the Google side, is when you start optimizing for value, it'll hit the value you want, but I'll be really conservative, so the number of purchases will go down. I have not seen that to be the case so far with Facebook's. I'm just waiting for them to completely screw it up, because that's what they do, but right now, it's working fabulously, so I highly recommend it.
Some people ask about creative, this is the other one, people get so hung up about creative on Facebook, because it's so exciting, there's video units, and there's canvas units, and we can be creative, we can do all this amazing shit. I don't like to make generalizations about creative, so here are some, because this is what we have time for.
People ask, they're just like "What should I run? What should I not run?" Here are some observations, your mileage will vary, however, this is what I generally see. Video, everyone is like "Video, video, should I make video? I have to make video. We've got to make video." Okay, video works really well, it has very high engagement, however, I do not find that it does best for conversion. It's really good about sucking people in, isn't necessarily the one that's going to get them to always convert. I tell people a lot "Don't rely on it totally for conversion campaigns." I'll tell you the ones you'll want to test for that.
Amazingly, static images still do well. It kind of blows your mind, because they have all this super fancy ad units, static images can still carry the day, it's kind of funny to see. They're cheap, which probably helps.
Slideshows, if you've seen these, this is where you can pick the images you want, it'll automatically animate through them. I've never gotten these suckers to work. I was like so excited when they came out, because I'm like "Yeah, for my clients that can't do video, this will work great." They suck. I've never gotten them to work. Please, feel free to test, however, I have never had luck.
Carousels, those are the ones I actually have great results with these. These work really well on remarketing. These are the ones where you can have a whole bunch of images, and users can manually swipe through. My personal feeling, and just gut check on it, is that it's probably because users have control over how they consume the information versus a video, which just kind of races through it, the user doesn't feel like they have a lot of control over it. With carousels they can kind of pick and choose what they're interested in, and click through on the card they're interested in, so it gives them a little bit more control over how they interact with you, and what information they see.
Finally, Canvas ads, again, I'm a marketer, I was so stoked when these came out, never gotten them to work. Super expensive, high CPMs, have not gotten them to work. I'm testing them with some lead gen now, because I've never gotten them to work on e-commerce, so I'm hoping, maybe, there will be some life there. But, I really want them to work, just have never had them do so.
One of the things beyond just getting things to perform well with creative is why does it matter from a cost perspective? Because, at the end of the day, I'm a numbers person. Things can look super cool, but if the results suck, I don't really care. Pardon me. Facebook has this thing they call relevance score, it's a lot like quality score in AdWords. It basically measures how interesting are you, how relevant are you to the people that you've chosen to show it to.
The higher your relevance score, they reward you with a lower media cost. Relevance score is basically a one to 10 metric, you can see it if you want to add it as a column in your reporting, it will tell you. But, it basically behooves you to have it be higher, because it'll start dropping your media spend. The way that they calculate relevance score, they don't give you the exact calculations, but based on all the testing that I and a bunch of other people have done, we pretty much determined that social interaction is the number one thing, shares, comments, likes, loves, all that kind of stuff, virality. Then, also, click through rate. Those seem to be the two things that really, really affect that number.
This also brings up the "How come I see ad sometimes and they have like 900 comments, and all of the social interaction? How they do that? I don't get it." We're going to talk a little bit, this is kind of like a pro tip, so if you're running multiple campaigns. You'll notice that these two ads are exactly the same. One was literally created and copied into another ad set, everything about them is the same. But, notice that the social proof is different. This one has 18 shares, two comments, a bunch of likes and loves, this one just has nine little shares, and 16 likes, but everything about them is the same.
Here's why. When you create a creative unit in Facebook, it assigns it a post ID, it treats it like it's something that you posted on your page. When you create this, and put an ad set A, it gets an ID. When you duplicate it and put an ad set B, it gives it a different ID. Same exact ad, different ID. There's a way around this, because what happens is, when they have two different IDs, they accumulate social proof by ID. While this guy gets all these shares and stuff, it has no bearing on this other one, it only accumulates to the one, which sucks, because you look it's the same ad, I want it to get all the stuff. I want all the things on the ad.
There is a way to do it, they don't talk about it, takes an extra step, so worth it. We're talking about stacking social proof. The idea is, you want to get everything rolled up under that one ID. You want to have one ad that runs in all these ad sets, but all that love, all the shares, all the comments come into that one ad ID, makes the relevance score way higher, your CPMs go down in each of those ad sets. It's important, it actually works. Beyond being eye-catching, because I do this stuff for a living, if there's something that has 405 comments or something, you can bet I'm clicking into that, because I want to see what the hell people are talking about. I get suckered into that, and I do this.
Here's how you do this. When you create an ad that you know you want to do this with. There's that little preview arrow up there. If you click that, and you view the Facebook post with comments, it's going to open up on your tab, and it shows you your page, and it shows you the ad. It looks like a post on your page. But, when you do that, if you look up in the top, there's a URL, and there's this little number at the end of it. That is the ID that you want. If you copy that, go onto the next ad set where you want to run this ad, let's say there's five other ads, that's however many others you want to run into.
If you go into the next ad set, so you go into ad set B, and you say create ad, it's kind of hidden, it dumps you on the create ad option, but there is this little button for use existing post. If you click that, and click enter a post ID, you can paste that ID there, and then all of the ... everything that accumulates on that one ID will show up in every single ad set that you put that ID into. It's awesome, it makes your social proof stack up way faster than having like 10 different silos, it takes a couple extra steps, I know it's kind of a pain in the ass. The good news is once you set it up one way, if you take that and copy it, and it's as existing post, it'll copy this to existing posts everywhere else. You only have to do it one time, and it'll stack it all up.
If you really want to get fancy, you can actually go forth and try and specifically get engagement on it. Let's say you run a conversion campaign that you stick that ad ID into it. It's a brand new ad, it has no social proof on it yet, and you want to jump start it. You can also set up an engagement campaign, use the exact same ID, it'll target that ad towards people, the most likely to comment and share, and you can run them simultaneously. Now, you've got the one ad ID working for you in conversion, same ad ID over in engagement that racking up all the social proof that's showing on your conversion campaign. It becomes this nice, big cycle all feeding through the one ad ID, shoots up your relevance score, which is amazing, drops your CPM, and everybody is happy.
Here's the bad news. Facebook sucks at creative testing. Have you guys ever tried to A/B test in Facebook ads? It is a joke. It is awful. Here are some examples, because I love to rant about this. I don't understand how they don't have this worked out. First, the really bad, if you do multiple ad units in the same ad set, they will always give everything to video. The only reason why carousel has more here is because I started it first. As soon as I added the video, it was all over, everything was all about video. Look at the cost per acquisition for video versus everything else. But it just keeps showing video, because Facebook is like totally going steady with video, it's all it cares about right now.
That doesn't work well. Whenever you're try and mix ad unit types, assume it will favor video over everything else, it does not care what's actually performing, it's going to pick video.
Then, let's say you choose the same type, it's 50-50 rotation. We can just pick a winner after 200 impressions. Here's a couple examples. Both, these were just static images, they're exact the same ad type. It decided to give all the money to the one with the lower return on spend, and it picked the winner after only 12,000 impressions of one versus it gave 144,000 to the other. There's like no science to this, I don't understand how they do this. But it picks a winner really, really, really fast.
The same thing down here, "Hey, look, let's spend $856 on the thing with the higher cost per download. That's a great idea." This is what it does, it just kind of arbitrarily picks the one that it thinks is the winner, usually is picking the one that has higher social proof, because it wants people to stay on the platform, or it doesn't want them to leave. You'll usually see it favor stuff like this. This has been a complaint for a very long time for me and everybody else that runs Facebook ads, because we're like "I want to creative test, why are you making this so difficult?"
They created solutions. Hooray, I guess, because they kind of lack some of the stuff we need. The first kind, they just recently added this actually. Split testing was something they added a year ago, the reason why it's awesome is because it makes sure that it runs it in your target, but it's two totally separate audiences that do not overlap. It's a true A/B test, whereas if you try to run it yourself, like if you set up ad set A, and ad set B, and try to do it yourself, you're going to have people in both those ad sets overlapping and seeing the ads, it's not really a true separated test.
In the split testing, it actually is two completely separate audiences. In this example, let's say we pick the conversion campaign. When you pick your campaign type, you'll see there's this little itty-bitty checkbox, that says create split test. If you check that, you are taken to a new screen. You can choose what it is you actually want to split test. You can try different delivery optimizations, different audience types, different placements, that part is actually really cool, and they recently added creative. That's awesome.
Here's the bad part, and this is the complaint with all their split testing. When you run a split test, you have to pick a set budget, and it has to end. If I have a winner, I can't just run it forever. I have to be like "Oh, great, that one, hopefully it'll do it again when I put it in the real ad set." It doesn't let you run it forever, which drives me crazy. I do have a long walkthrough about split testing, so if you're interested in learning more about split testing, feel free to go there, I have a whole walkthrough about how it works. That's kind of my biggest complaint with that, is that you can't just run a winner forever, you're kind of stuck doing a little episodic test.
They recently unveil dynamic creative, which is really great in theory. When you create an ad set, at the ad set level, there is an option for dynamic creative with a little toggle button. When you turn that on, the rest of the ad set will look exactly the same. But, when you go to create an ad, it's going to look different. It'll give you all these different fields, so you can add up to five different texts, five different headlines, five different news descriptions, you can do different call to actions, you can add different videos, different images, all that stuff, and it'll automatic test, and rotate all of that for you, so you no longer have to do it. Sounds great.
Here's the problem. See, there's always a problem. I'm telling you. Where is the winner? We don't know, because when you run e-commerce, it doesn't actually tell you which image is the winner. It doesn't tell you specifically each one. It will find the best combination, but it's a secret. It's not going to tell you which one is the actual winner. Isn't that great? As marketers, we're like "That's awesome, we don't know who the winner is, but Facebook does." Which, as you saw from my previous nitpicking about how they test things, makes it dubious to me that they're actually running the winner, who knows?
But, yeah, this is an option. The way I've been using it, I'm going to have a longer walkthrough on my site if you want to go through it step-by-step. But, the way I've been using it is kind of doing more manual testing, and then narrowing it down to the two to three winners, and then just throwing those into the dynamic creative, and letting Facebook work out the rest. I won't do like five of everything, it's too many combinations, and you'll actually find it doesn't seem to able to find a winner at that point. That's their solution to creative testing right now.
At the end of 2017, I think the number one biggest question, other than "What the hell do I do about bidding?" Because everyone is obsessed with bidding, was "What's going to happen next?" Here's the thing with Facebook. They need to be more than your news feed. They already, as of a year ago, this past Q4, they ran out of advertising space in the news feed. They need to find other ways to make money. They need to find other ways to be in your life, and to know everything about you. Facebook Messenger is one of those big ways.
1.3 billion users. It's a lot of people. It's a lot of potential ad space. They did start opening it up a couple different ways. They started adding a Messenger line-item, which you've probably seen if you've been in Messenger, there'll be like these random ads that will show up sometimes. Surprisingly, they do okay, generally. There are exceptions where they don't do well, and you just kill it. But I've actually started running it, just as default, because it tends to actually do pretty well. And if you let it run and optimize a little bit, it'll start figuring it out. That's one way that they're starting to monetize that.
Another thing that's starting to happen is chatbots are becoming a thing. I don't know if you've encountered any of these yet. It's basically automated messaging, but it makes it feel like a very personal experience, which is kind of creepy sometimes. But, here is one example, so it's Shopspring helps you literally shop.
"Hey, Alan, what are you looking for today?" "I want men stuff. I want women stuff." "Cool, what kind of shoes do you want? Sneakers, loafers, got it. What's your price range?" It walks you through the whole entire thing all the way to purchase.
Another thing, help me choose something, even if you're not necessarily ready to buy, but you want help. LEGO unveiled Ralph, the LEGO Gift Bot for Christmas this past year. Ralph actually had a personality. They would say things like "How old is our builder?" Instead of like "How old is the kid?" They used the LEGO vernacular. "How old is our builder? Oh, you should see the thing that I was building when my circuits were that old." It make it like a very personalized little character that takes you through all the different ways, and all the different things to buy, and they help you narrow it down, and how to purchase. Plus, bonus, they give you a free shipping code at the end so you go through the whole thing. People are trying to find ways to leverage Facebook as more than just an advertising platform, and turn it into a customer service platform, which of course Facebook loves, because it's more people using Messenger, more people interacting with them.
Another thing that started to happen is that they're taking on Craigslist. Facebook, again, wants and needs to be more than a news feed, and they're becoming a market place. I think it was, this has actually gotten bigger even since then, but last year 18 million new items were posted for sale in the market place in the US during one month. It's growing exponentially. It's definitely starting to become way more of like a traffic place as opposed to before I think people didn't know what it was.
Again, this is something that doesn't actually do all that bad. It depends, again, on the customer, it's not high volume, as you can see, it's lower volume. But, it is something that we've seen decent return on, and a lot of it is because, again, look at the CPM cost, $1 for 1000 impressions. That's insane. That is so, that is like the cheapest media that they offer right now. Even if you get tons of wasted 573 link clicks, it doesn't matter, because we got six sales for $1 CPM, which became like $1,000 return on spend.
More and more you're going to see that the news feed kind of is what it is, it's expensive, people have to get smarter about it, but there are these little pockets that Facebook is now testing that are probably going to become larger in the future. The nice part about them right now is that they're not that expensive because people aren't really using them. They're sticking to the news feed, it's what they know, it's what is familiar to them, but I would encourage you to try some of these little strange placements that Facebook has added on, because they actually are doing well, there's not a lot of competition yet. As they get bigger, and there's more inventory, you'll probably encounter more competition, so now is a really good time to do it, because it's cheap.
I know this is a lot of information. People feel like this about Facebook ads on a good day. If there's one thing that you can take that would help you from this that makes me happy. I think a lot of times people are like "What's the silver bullet? Like what the one way to make Facebook ads work?" As I told people, it's basically like 10 little levers that you have to position just right. A lot of it is getting the lever just right on the creative, getting the levers just right on the type of campaigns you run. It's not one thing, it's this whole little daisy chain of things that get done just right, you finally start to see the answers that you want.
The cool part is, they change it like every month. What works this month, might not work next month, but a lot of times they have something that's rolled out that you find starts working better. I always tell people not to get complacent with Facebook ads, because they're not going to stop changing, it's part of the reason why another question I get is what third-party tool should I use? None of them. Because none of them can keep up with what Facebook is doing. Every time they build something, Facebook is like "No, we're going to do this. No, we don't have Power Editor anymore, we just have Ads Manager."
Every time they start creating stuff, the third party tools are like "Shit, we got to go fix this thing, and change it, we got to remove this checkbox." They just can't keep up with what they've got going on. But, what's super cool is that we got to do it all, and there's not really a rule. If you just go forth and test, you get enough data to start figuring out what works, and what your backup plan should be when something doesn't work.
Now that you have some tools, you can go out and rock some data. I totally did it, I totally made a pun. Go forth and test, I'm always available for questions, case studies, anything, I love to geek out about this stuff. Feel free to find me online, drop me a note, and I'm happy to answer any questions. Thank you.