As Facebook puts more emphasis on monetization and more advertisers are spending on its platform, measurement is increasingly important.
The social network is still so new and comes with unique ad types and its own vocabulary, so many advertisers still aren’t sure how it fits in with the rest of their marketing efforts. The Facebook measurement team is working to put its ads in a more familiar context for advertisers and apply its research findings to offer better ad types and systems.
We spoke to Head of Measurement Research Development and Partnerships Sean Bruich about the questions Facebook is trying to answer, the value of a Like, how Facebook ads compare to more traditional channels and what needs to happen as marketers begin thinking cross-platform instead of in silos.
The following is an edited transcript from that interview.
Tell me more about your role and what your team does.
My team is the research development team, we also work on partnerships with third-party research companies like Nielsen and Datalogix. Our goal is twofold: to help build the right tools to help advertisers buy media to capture the value they’re trying to get, and the second piece to build out the measurement systems that help quantify that value and optimize campaigns, not just for e-commerce transactions, but a broader set of marketing outcomes like offline sales. It’s pretty simple, we’re trying to solve how you value an ad campaign online if one of your objectives is offline sales impact or one your objectives is staying top of mind. And the second piece is how do you translate all of those things that you know about other forms of marketing and understand whether online is accomplishing those same goals and how you would execute against those goals online.
It seems your latest research has really focused on how successful campaigns aren’t just about clicks, but reach and frequency and other factors. Are you finding that’s resonating in the industry and advertisers are changing their strategy?
Definitely. I think what happened was digital came up as this really specific type of marketing and there was a lot of building up of the infrastructure there, the direct response stuff. But it’s getting harder and harder to plan your campaign because there are more ways to reach consumers. It’s not just TV and online. It’s TV and Facebook and tablets and smartphones. I think there’s a growing realization on the part of marketers that they want to reach people across these devices because if they’re reaching the same person on each of these and counting it as a distinct person, that’s not good. It becomes really hard to plan an ad campaign if you’re just buying in these silos.
So that’s what Nielsen’s cross-platform ratings address, and over time you’re going to be able to say ‘OK, for this type of consumer I can reach them more cheaply with one TV impression on this show and then a couple ad impression on Facebook mobile’ versus saying ‘I’m going to do my TV buy, my online buy and I’m not going to know how those overlap.
I think you’re starting to see publishers adopt this way of seeing the world that looks more like traditional media, and you’re starting to see advertisers do this. Unilever has talked about this very publicly. They buy a lot of television, but they want to buy online the same way and they want to use Nielsen to guarantee or at least assess that what they’re getting from publishers what they’re promised, that they’re getting the reach. Because if you’re just buying impressions those could all be going to one person or people outside your demographic target.
So last year was about identifying factors that are more important than clicks, and I remember the year before that Facebook did a lot of research that talked about the value of friends of fans. What’s important for Facebook to measure and get across next?
I think it’s putting all of those things together. Describing the keys to generating offline sales by different audiences. I think so far when you think of mass market campaigns, they’re not really focused on particular subsets of consumers. One of the powerful things about digital is that rather than offering one piece of content that sort of works for everybody, you can separate out your audiences. You’re still reaching everybody, but it’s going to be the right message for the right person at the right frequency. The sum of which should be more effective than a ‘safe’ message or a somewhat effective message for both.
That’s creative, but I think the same thing applies for how many impressions you should serve somebody — are you willing to pay a lot of money to acquire a new customer versus remind an existing customer? I think some of the data and ad serving pieces of the web are going to be really powerful at getting you the scale of mass marketing but actually getting to the precision that has never existed before, whether it’s more or less impressions or different creative or different flighting of the ad campaign.
Are those things Facebook can do for advertisers or are some of these things so specific that it’s more about creating the tools for each advertiser to do them for themselves and with their partners?
I think it has to be both. One of the things I like doing is looking back to how TV came about and how those things were figured out. And they did both for TV. There were some great industry researchers who found some general things to be true, flighting of ad campaigns, effective frequency. That narrows the window of all possibilities to ‘in general we see these things to be true.’ I think we’re starting to see some of those things converge for Facebook. We know that the range for advertisers is ‘this big’ rather than ‘this big’ but for each particular campaign, they’re going to have to figure out exactly.
A common thing I hear from advertisers is that Facebook doesn’t share enough data to make it possible to track and measure the things they want. Is that changing or is it more about moving slowly and working with the right partners?
Certainly we have to be careful at Facebook given the data situation. We’ve had a lot of success working with really trusted and well-respected third parties. They are going to do a good job to make sure data is treated well, they maintain privacy protections and have security things figured out. The flipside is that we also want to partner with people who are going to be really compelling to the advertisers. Why work with Nielsen? Well because Nielsen already works so closely with these advertisers on television media buys, they have a lot of credibility to say ‘if you shift some dollars to Facebook or online or digital, you could increase efficiency.’
The issue with that approach though is that it focuses on a few major brands. How do you take the learnings that you get from big clients and apply it to a broader set of advertisers?
That’s exactly what we’re worried about every day. I think it falls into a couple of buckets. One of the things you can do is provide a combination of scaled products plus case studies or general learnings, which is really important. Not every advertiser is going to need Nielsen XCR (cross-platform ratings), but for Unilever it matters a ton. They need XCR and it needs to be incredibly precise. On the other hand, managing to reach and frequency, that probably matters for every advertiser. Insights can help for someone who just manages on Facebook to make sure they’re monitoring reach and frequency.
So for example a local pub in Palo Alto, we want to make sure they’re aware of our research findings and try to put it in the context of what they already know about marketing, but we also want to provide some very low cost tools that maybe don’t do the same thing as some these bigger research partnerships, but at least allow them to manage their campaigns in a way that drives better ROI. It’s about putting the research things that we figure out into our tools. So with social context, we know that delivers more effective advertising. We know that from Datalogix and Nielsen surveys. That doesn’t mean everyone needs a custom Nielsen study, but we need to make sure there’s a white paper that describes the impact and then we’ve got to make sure that the tools we offer and insights show social coverage and our marketing materials teach them how to do more Sponsored Stories.
And actually the latest version of the ad tool by default automatically creates Sponsored Stories when people go in to make an ad.
Yeah. Another thing we found through research two years ago was that lightweight users respond very strongly to advertising, and so we actually made a change to how our ad server picks how to deliver impressions. We were able to increase the reach of ad campaigns and increase the lift advertisers would see in ad recall or other brand effect metrics based on the fact that we could do a better job of choosing ‘should we serve an extra impression to this user or serve two users each an impression?’ We were able to figure out optimal user targeting and frequency, and we saw substantial increase in value in these ads that we delivered.
It seems like there are a lot of things we could see as a result of your research, for instance in the ad tool, but there’s probably so much more that we won’t see, but that can just be built into how ads are served.
The nice thing about measurement is that it is really accretive across all these groups. Measurement means that advertisers can plan better campaigns. Facebook’s able to build better ad products. And then users are seeing better ads at a more reasonable or more optimal frequency. They’re not seeing ads they don’t want to see, the creative gets better. And so the whole thing rises. I think that’s a really nice piece of measurement. When I go home at night that’s what makes me happiest. You build this system and it’s not a zero-sum game, it’s all summing together making the system better. The more measurement you have, the more you learn, the more it improves and improves and improves.
What are the biggest or most common questions or doubts that people have that you guys are really focused on?
The biggest thing by far is ‘how much do I know about marketing can I translate to Facebook?’ It’s a great question because it’s a totally different way of delivering an ad and communicating with people, but it’s still marketing. You look at an agency and they’re experts in how to do TV or magazine and they’ve been doing that for years, so it’s not really about the data, it’s about instinct: ‘This is a good ad, this is a bad ad, this is a smart way to spend our money, this is a bad way to spend our money.’ And not many people have instincts on digital and certainly not for these big consumer-oriented ad campaigns.
I think that’s one of the biggest questions, so a lot of our research is around trying to create some comparability to the other options advertisers have. So when we did creative research, rather than invent a new scoring system around social, we applied the Advertising Research Foundation’s creative best practices guidelines. We started with that scoring system and asked, can marketers using that scoring system predict which ads would be successful in the Facebook market? We found that, yes, these things were predictive. Strength of creative can be described in this traditional way and actually do perform better and marketers actually fortunately can translate a lot of their expertise.
That’s really interesting. One of the questions I had wanted to ask you was whether Facebook was more similar or more different than the other channels that advertisers and marketers are used to.
It seems like it’s more similar to the type of advertising, to the objective of the advertising you’re doing. So DR (direct-response) advertising on Facebook is a lot like DR advertising in the rest of online. We’ve seen this with FBX. People thought, ‘no way Facebook works as a direct response channel, people aren’t looking for anything while they’re there.’ That was totally the meme, and we launched FBX and turns out it works great. Then for TV and print, which is more the offline sales, brand-focused, awareness raising, it looks like that (on Facebook too). You’ve got to have creative that’s relevant that has a clear visual of what you’re trying to represent, that it’s rewarding to the consumer, there’s a value proposition there, and then you have to reach your target audience at an effective frequency.
When there’s a new form of media, people notice the things that are different, and that’s actually a good thing. I don’t think we want to minimize the importance of some of those things. You can imagine a more personal tone works better on Facebook. Or if you want to run an ad in News Feed, it’s probably better if it doesn’t look like a classified ad. I don’t have data around that, but my guess is that it’s better if it looks like content and you do something cool that’s going to look awesome in News Feed. But I think sometimes we lose the similarities. One of the things we try to do is point out what are the similarities and what are the things you still have to keep in mind when planning an ad campaign on Facebook. You don’t want to get so worried about the tone of the ad being familiar or honest and open that you fail to deliver a value proposition about the product. I think it’s tweaks; it’s not rethinking marketing.
It sounds like it comes down to having focus, and I think that the message you guys are putting out there is more focused in identifying the tools and channels and ad types for direct response, for brand building, for app downloads or different things. As opposed to ‘here is Facebook, here are ads, get Likes.’ So you had a lot of people doing that before but not being clear why they were doing that. Now it seems like you guys are maybe steering people more the right way.
That’s what we’re trying to do. I think all good marketing has an objective, and a big part of what measurement can do is make sure you’re measuring the objective of the campaign. If your goal is to have Likes — and there may be a lot of good reasons to acquire Likes — then you should monitor how much they cost and make sure the people who are Liking your product are the type of people you want. But if the goal is offline sales, then you should be measuring offline sales. If your goal is to extend a TV campaign onto the Internet and add incremental reach or incremental frequency or spark a conversation, then you should be measuring against those business outcomes.
A question a lot of people want to know is, what is the value of a Like? What is Facebook’s perspective on that question?
I think there’s often two questions wrapped up into that. One is, on average, are the people who Like your brand better customers? And we looked at this a number of ways with surveys and Datalogix, and the answers are remarkably similar across categories of pages and ways of measuring these things. They are better consumers, between 1.5 and 2x better consumers. This aligns with comScore’s data in the Power of a Like paper, so we’re seeing a lot convergence around this answer that fans are, on average, more valuable than average. So that’s the audience question.
I think the second thing people often mean when they ask this question is ‘how does that Like change their behavior?’ That one’s very dependent on what the subsequent marketing objectives are and the subsequent execution. It’s much harder to give one answer across all brands because of the diversity of objectives. We see some advertisers where the Like is just an opportunity to do very low cost CRM type loyalty brand building. But for others, maybe an entertainment studio ahead of a movie release, it’s more of a stepping stone to having a more effective advertising campaign. If you add half a million Likes, what does that do in terms of your incremental optimization capabilities? Were you able to deliver more social ads, more viral impressions? So you sort of don’t want to think about the Like itself in that case, but what did the Like enable, which is a much more important question I think.
What are the biggest challenges for your team in the near term and long term?
The biggest challenge is how do you accelerate the learning? We’ve had 40 years or so of television to get good at it, 200 years of print to get good at it. We don’t want this to take 40 years, we don’t want it to take a hundred years. Every ad campaign that’s running that we’re not informing with this exciting research is a missed opportunity. I think that’s the real challenge. What are the really big research bets we need to be making and then what’s the most effective way to get that information out to a large set of marketers and advertisers in a way that’s going to help them build better campaigns? That’s the short term challenge.
I think the long term challenge is really about cross-platform marketing. Think about a world where a given consumer is watching TV, reading magazines on their iPad, checking out Facebook on their mobile phone and checking out ESPN or something. That’s a tough world for media planners. What a marketer now is going to have to do is figure out which creative and which people and which platform and manage that in a way that’s not four times as complicated or two to the fourth times as complicated. I think that’s the long term challenge.
We’re going to have see a change in the way marketing works from being very siloed and platform-focused to very consumer-focused. And sorta say, based on what we know about this anonymous person which is Elizabeth, she does awesome with video ads. It’s inexpensive to buy video ads in this place where I can reach her really efficiently, but I gotta follow it up with a lot of Facebook advertising on mobile as she gets closer to the point of purchase. That requires a big change in advertisers, agencies, publishers, research agencies. It’s going to be really cool, but it’s going to be the biggest shift in marketing since the advent of TV.
I think that’s the fun part about this. This job isn’t going to be done any time soon. I think we’re going to be talking about this stuff for five, ten years as this shift occurs. And it’s only accelerating. You can’t keep adding silos. You have to be able to say, these are the types of people I want to influence, and what’s the right way to reach and influence them.
It’s been interesting to see the type of advertisers who are getting active in advertising in the mobile feed because when we think about who was advertising on mobile before, it wasn’t the same people who were advertising on Facebook. It was developers or people who have mobile things to sell. To see now that you didn’t just get people to advertise on Facebook mobile, but you got many advertisers to advertise on mobile for the first time, has been an interesting thing to see happen.
I think it goes back to this idea that advertising on Facebook must be really different and mobile advertising must be really different. Like you’ve gotta have this SMS thing that’s location-based. Some of that may be true. I don’t doubt there’s going to be amazing SMS-based location advertising in the future. I’m sure somebody will figure that out. But at the end of the day, people who see TV ads don’t buy on television.
So lots of people are looking at their mobile phones and engaging with something on their mobile phones, why wouldn’t it make sense that it would influence people later, just like TV influences people later, just like radio or print ads. It’s another place to reach consumers and we’ve gotta figure out what consumers, what type of content, what type of ad is optimal. The shift to online was almost understood as ‘this is really different’ and now with mobile there’s another shift and tablets are another shift, but I think what we’re actually seeing is that we’re still reaching consumers.