"Lead ad product strategy for the world's dominant social media platform. Scale world of mouth marketing in a way that's relevant to a billion or so users with myriad devices, nationalities, languages, and cultural reference points. Do it in a way that unlocks the spending power of global brands. Oh, and while you're at it, tap into this RTB thing. We hear it's gonna be huge."
Such a ponderous checklist might slow down some product executives. Not Rajaram, nor the engineers, project managers and designers who report up to him. In the past four months alone, Facebook's ad pipeline has spewed forth a continuous geyser of new products and enhancements, including the Facebook Exchange, mobile-only ad targeting, Sponsored Results search ads, and Premium ads in the API.
In this lengthy interview with AdExchanger, Rajaram talks in-depth about those products, and Facebook's strategic vision and development cycle for ad innovation.
On September 20, Rajaram will speak at AdExchanger’s Human Centered Automation conference in New York City.
AdExchanger: You've spoken publicly about Facebook's ability to influence decision makers high in the purchase funnel. How do you make that happen on a product level?
Gokul Rajaram: Facebook Pages work really well to reach people at all parts of the funnel. The Page should be the hub of all your marketing, we believe, because Pages allow you to reach people in the news feed. Pages allow you to publish content. That initially starts out as organic content, and then you can sponsor or boost the content [using our ad tools].
If you think about Page posts as photos, photo albums and videos, all of these help move people from awareness to consideration, and beyond.
For brands, one of the most interesting product offerings we have is Reach Generator, which ensures that every single one of your page posts reaches all of your fans. Organically, you get anywhere from 15 percent to 20 percent of your fans, that you reach organically. In order to reach the remaining 80 to 85 percent, sponsoring posts is important.
We're also building products to make it easy to reach friends-of-fans, because social context is really powerful in upper funnel experiences. We've seen a high increase in purchase intent for ads with social context. We also, a few months ago, enabled marketers to choose feed‑only or mobile‑only bidding, so we can have more control over where they reach people.
It's the combination of Pages, Page posts, and News Feed, which transcends desktop, and is a cross‑platform thing, where you create a Page post and it reaches people not just on the laptop, but also on mobile, wherever they are.
It's very powerful for brand marketing, and large brand marketers are extremely excited about countries with large mobile populations, like India or Indonesia and other countries like that, where they can make buys, now, that can reach mobile populations and not just desktop. They don't have to think of mobile as a separate thing.
How do you create ad experiences that work across all those country populations and their predominant devices?
Our advertising strategy on mobile is all about making sure we are part of the organic content that people consume. It's primarily News Feed. On mobile, it's actually fairly easy for us. We are essentially piggybacking on a lot of the great work that the News Feed and Mobile team has done, because a lot of our mobile advertising strategies are about taking organic content and sponsoring it. The organic content already is being rendered by the News Feed and Mobile teams, together.
Since we don't introduce content which would not be there otherwise ‑‑ like a banner ad or something like that ‑‑ in mobile, it's a much easier transition for us from a pure product development point of view.
You hail from Google, which seems bent on conquering many channels at once. Can you contrast that with what Facebook's strategic vision is for ad products?
As you understand, I can't comment on what Google is doing.
One of the most satisfying things for me about Facebook is that it is not purely about advertising. It is about marketing. We don't refer to our customers as advertisers. I'm going to allude to this at AdExchanger's Conference during my talk. I think marketing is deeper than advertising. It's about creating that emotional connection with your customers, connecting with people, getting people to connect with you, getting them to raise their hands, and then creating content that is compelling enough that people share it.
So when we see Burberry post a photo album post on their page, thousands of people will like that photo album, and that goes to their friends. That is marketing.
So, the last two years at Facebook have been a great learning experience for me. I've spent more than a decade in online advertising. But I think what we are trying to do fundamentally is deeper than advertising, and this goes to the marketing. I think we are trying to reinvent marketing versus just advertising, which I think is very different from what has been done before.
How does your ad product development cycle work?
We are a mobile first company, and this vision and philosophy also extends to the ad product. Ideas for new initiatives emanate from several sources here, engineers across the company, not just on the Ads and Pages teams, but basically engineers across the company and across all teams.
The most important thing for us to observe ‑‑ that has led to a lot of interesting product concepts ‑‑ has been how our users use Facebook. Because we intrinsically believe that advertising and marketing should enhance native‑user behaviors. Things like Offers originated from this belief. When we survey users asking, "What are the reasons you Like a page on Facebook?" one of the primary reasons was people want discounts, coupons, and deals. To build a compelling product it takes a coupon, which is an age-old thing.
We have small teams of two to three engineers and a couple of product managers and a designer on every initiative. So that leads to very fast and iterative product development.
In fact, one of the best compliments for me one of our initial Demand-Side Platform partners for the Facebook Exchange telling me that they were blown away by how fast our team is. They thought Facebook, as a large company, would not be able to move this fast. They were just blown away by how fast our team had executed on and launched a product in Alpha. I took it as a really high compliment because it was coming from a nimble startup.
Our focus is to move fast -- have all teams that work together not encumbered by any constraints. They execute and launch something fast and learn something and move on.
That's what I think also makes Facebook attractive for engineers, designers, and product managers. People feel that they will have the ability to exercise their creative muscles, their execution muscles, and have a sense of ownership. In a small team, you feel a complete sense of ownership.
Does that iterative speed bewilder your Preferred Marketing Developer partners? How do you communicate product and feature changes to all the PMDs, and is more communication needed?
We have a number of communication channels with PMDs. All PMDs have access to a PMD-only community forum. We have several forums, groups on Facebook that they have access to. Then we have product and program update forums. Then we have PMD email support. We also have regular newsletters that we send out, and we have limited opportunities for select PMDs to test products in pre‑launch status so they can test it and give us feedback.
We always launch products in the API first because we want the PMDs to try building stuff around these APIs. They start building the interfaces themselves.
Depending on a number of factors, select PMDs also receive partner management support. So we have people who support them ‑‑ sales training, access to partner engineering, account management, and also other consulting services from our marketing partnerships with our PMD team.
We have a large number of different channels, both community oriented as well as actual support for some of these larger PMDs.
Do you see the PMD program continuing to grow in terms of the number of badges and partners you add?
Yeah, I think we do. For example, we have seen the Facebook Exchange expand the PMD universe. Given that there are dozens of DSPs out there and as the exchange grows, Facebook exchange technically and operationally is a subset of the API. Because of that, it will expand just because people start buying on Facebook Exchange. Many of the FBX partners are interested in the native ads API, because many of these guys have not done much buying through native ads APIs.
Now, as they start figuring out what the capabilities of the ad system is, they want to essentially get into the ad space a little bit more on the Facebook side. So they want to invest further in it. So we do see expansion of the PMD program just based on new products like Facebook Exchange that we released.
How do you evaluate the adoption of Facebook Exchange to date? If you can say anything about performance or marketing impact, that would be great.
I can't say anything about success because it's still in beta right now.
We are working with the initial partners. Right now our focus is to make sure we work with all these DSPs and partners to insure an end-to-end seamless experience for advertisers. We want to make sure the initial data that they get is validated, proven, et cetera, before they go out and essentially do more stuff and we get more partners. So we've been very conservative and careful about having partners in the Exchange because we want to make sure the system works really well.
There are obviously some differences in Facebook creative being different from [standardized display] and so on. We want to make sure the partners understand that just because they are integrated, it doesn't mean the responsibility ends there. Their responsibility is to work with us to make sure the advertisers who are our common client have a great experience on Facebook Exchange.
Can you say anything about the amount of inventory that's flowing into FBX at this point?
I can't comment on that.
Aside from just making sure it's a good experience and that you onboard the partners effectively, what's next for the Facebook Exchange?
There are a bunch of features. I mean, this is our first foray into essentially doing anything with the ecosystem outside of Facebook. We fixed a lot of things that we learned from our partners. Now in the beta, we are essentially launching a bunch of new features. I think right now, the goal is to simply bring it to a stable version where it works… And scaling with the partners, making sure that we fix any of those issues, and get a baseline of features that any Exchange needs to have.
Do you foresee a day when user data on the Facebook platform can be used to deliver ads through the Facebook Exchange? Will there be more of a mingling of data?
Not at this time.
Retargeting was considered the big opportunity with the Facebook Exchange, but a lot of people in the space believe that upper funnel or brand marketing will become more important in programmatic ad buying. Do you agree?
It's hard to say, because we feel the challenge with convincing the big brand marketers is not whether it's real‑time bidding or not, but it's really around the right metrics, showing measurement that enables them to measure this consistently. I think the bidding is a red herring here.
Big brand marketers aren't waiting around saying, "Oh, I want to have real‑time bidding. I will move my brand dollars to online and Facebook," or wherever it is. I think they are saying, "I need to see that online actually works, compared to what I'm used to buying. I need to see comparable metrics."
I, personally, feel that the whole discussion of RTB and upper funnel is just a tactic… Search didn't have RTB. But, it works really well because the metrics are very clear. I think the challenge with brand is not RTB or audience buying, or RTB or non‑RTB. It's really around metrics and measurement.
Since you mention search, Facebook has just launched Sponsored Reults, a search product. Do you think search could eventually become a big part of Facebook's advertising offering?
It's an initial test we've been running, since we have seen that people on Facebook search for pages, apps, and of course people. We've been trying to see what are interesting advertising opportunities there. It's just an initial alpha right now, with a set of marketers, that we are trying to see. It's still too early to see where it takes us, but again, like everything else, we are going to move fast and, as we get learnings, we'll evolve the product.
Another format that people would not ordinarily associate with Facebook is video. Yet Facebook is one of the largest video platforms, in terms of the number of views that happen in the world. Is there an opportunity down the road for Facebook to create video advertising formats, or a video ad experience that's relevant to the platform?
It's something we haven't talked too much about. We probably need to focus and build a lot of stuff on the consumer side, if we want to do something. I think, so far, we haven't talked much about the advertising side of video, at this point.
Online at least, upper funnel advertising is an area historically dominated by content and portal sites. Do you see Facebook competing more directly with those publishers?
It's not just about brand dollars on the Internet. It's about brand advertising, overall. Currently most, if not all, brand advertising is on television and out of home, and things like that.
We feel one of the biggest things holding back brand advertisers has been the lack of standardized metrics for digital, so it can be compared with other media in an apples‑to‑apples fashion. We've been working with third parties, as well as IAB, ARF, et cetera, to advocate for industry standard metrics. As you know, Nielsen debuted a product, OCR ‑‑ Online Campaign Ratings ‑‑ last year, enabling marketers to measure their online campaigns in GRPs.
Is Facebook taking steps to actively support the Nielsen product, or online GRPs in general?
Yeah. We are a big supporter of it. We have seen several large marketers and agencies start to test it, and in the initial phases of adopting it. It's been exciting.
Facebook has added some new ad tracking partners this summer, including Adometry and DoubleClick. Are impression tracking and attribution modeling becoming more important to Facebook?
We feel it's important to think deeply about attribution. We think that the field has an explosion of choice around attribution models. We've talked to a lot of the attribution companies, and there are half a dozen to a dozen different attribution models, depending on how you attribute different things in the funnel, or during the consumer journey, as they say, from first impression to the purchase, or to conversion.
I think our ideal would be to have a small number of standards that everyone agrees on. One of the things we see, as we've been working with these attribution companies, is everyone has different models. We would love to see more standardization, which is certified or approved by standard bodies, and adopted by marketers. The plethora of choice is confusing to marketers, and they don't know which one to use. We would love to, and we are very supportive of having standardized attribution models, that have been vetted by standard bodies, to measure all online advertising.
The bigger picture of the attribution model is definitely important. We feel the bigger dollars, especially in the medium term, is really trying to do standardized measurement across media, versus just focusing on digital and how to move digital dollars between different channels. It's really how do you take these methodologies and have standardization between digital and non‑digital.
What about impression tracking for organic content on Facebook? Is that on the horizon?
Not at this point.
What changes have you observed in Facebook's approach to advertising?
What has been exciting for me [came] after we started working collaboratively with the News Feed team and the Mobile team, and the Platform team. When we launched ads on Facebook four or five years ago, they were these stand‑alone units that basically were some text ads, and some of them could have images.
Advertising has come together over the last year to 18 months. Actually, they did a lot of collaboration with other teams, so now we think of advertising or marketing as not a vertical thing that's just owned by one team. We think of it as almost a horizontal service that, to succeed and to have a great user experience, depends on other teams at Facebook to work together with us.
Sponsored stories, I think, was the catalyst. Where do these stories that are sponsored come from? They come from all the other products that people use on Facebook.
I think that was a spark, an a‑ha moment, that got us to say, "Oh, actually, it's not just that ads or marketing on Facebook should be a great part of the user experience. It's that they need to be, because in order for sponsored stories to work, sponsorable stories need to be generated by all of these other products that people use.
As soon as that clicked, a year and a half, two years ago, I think all the teams and us started working really closely together, because we want to make sure that: (A), the user experience around stories and sponsored stories is compelling, and (B), you essentially don't think of advertising as an afterthought.
Does the rising amount of data from sponsored stories ads help you optimize better for ad effectiveness?
At a high level, we are able to optimize. One of the best things for us around optimization is understanding the advertiser's goal. Advertisers don't necessarily want just clicks or impressions. Those are proxies. They want to get their apps installed. They want a mobile app installed. They want someone to claim an offer. They want someone to fill out a lead generation form. They want to reach a million people.
They're really trying to have better systems in place to take advertisers' goals and then optimize based on the goals. The more understanding of goals we have, the better. We have this product called Optimized CPM which is an API and most of the PMDs now use it.
Instead of saying, "Oh, better CPC or CPM," they basically say, "Give us a goal. Tell us your objective," and we have a few objectives -- for example, get a fan, get an app installed, etc. We are adding more. Give your goal to us, and then we optimize for that goal. Even PMDs that had really sophisticated bidding algorithms realized that layering those algorithms on top of our optimized CPM made them much more popular and effective. So that's really one of the biggest things we found and we want to encourage marketers to tell us more about what they intend to do.
Again, this is something that didn't matter as much on the search front, because everything was a click to a conversion. When you have a marketing environment where there could be a large number of different things that you could want, expressing it just by a CPC or CPM is not the optimal thing. Putting a lot of energy into getting marketers more about what their goals are, what their objectives are, and optimizing and even giving them back information about how we did versus those goals is a big focus area for us.
Where does the upper-funnel opportunity you describe leave lower-funnel, direct marketing and commerce on Facebook?
Actually it leaves it in a pretty good place because it turns out Pages work really well for middle-low funnel activities. I'll give you three examples. One is a product we call Offers. You might have seen some of the stuff we've done with offers. It's a viral take on coupons. A business posts an offer to their fans, and then the fans claim the offers and they go to their friends. It's very powerful, because businesses are able to reach new customers through their fans with new speed. In a large percentage of all offers, we see that the majority of all claims come from friends-of-fans as opposed to fans.
Another compelling use case is mobile app installs. A large percentage of mobile apps today already leverage Facebook organically for distribution. We are starting to see success, and driving both volume and good cost-per-install metrics around that.
Finally, e‑commerce players. For a long time they were trying to do search‑like things on Facebook, and after a lot of education and a lot of working with them, they are finally starting to better understand how to leverage Facebook.
Fab.com's CEO, Jason Goldberg, recently stated publicly that Facebook advertising accounts for 75 percent of all their advertising. Fab, like all e‑commerce players, are only going to allocate that kind of budget to a channel that drives clear and measurable ROI, so that metric was really great to hear.