Paul McDonald has been with Google for 10 years, most recently as a product manager for Gmail. Prior to working on the e-mail service, the 33-year-old has worked on Google’s commerce, advertising, and cloud computing efforts; launching Google Checkout, Google App Engine and several optimization products for AdWords. Today marks the one-year anniversary launch of Google Consumer Surveys, a product created with software engineer Brett Slatkin. Completed micro surveys grant readers access to a website’s content and earns publishers $0.05 per response. Earlier this week McDonald opened up about the past, present and future of Google Consumer Surveys. Here’s an edited version of that conversation.
Q: When did you come up with this idea?
A: It was really a quest to save the newspaper industry. In 2010 I was product manager for Gmail, and so like all employees at Google we get 20 percent of our time to work on things that we’re passionate about and I’d been working with a former colleague of mine from another team on ways to monetize content online. Around that time I also saw charts of revenues from the newspaper industry and you could see around 2005/2006 that the revenue from print subscriptions sort of fell off the face of the earth and the revenue from online advertisements hadn’t made up for the loss of subscriptions.
We noticed that a lot of these publishers were putting up paywalls, like The New York Times and The Wall Street Journal, trying to compensate for that lack of revenue from the print subscriptions. We didn’t quite like that experience, but we knew that journalism was expensive, especially really good journalism and so we wanted to find a way to compensate those publishers.
We hit upon this idea that instead of paying for content with your money, you could pay with your time for answering a one- or two-question market research survey; or at the time we had another idea, you could do what we called the human computation task, like labeling an image or categorizing a product. So we did this test in 2010 with the Huffington Post and we saw that people were happy to answer the market research questions and they weren’t so pleased with doing the human computation tasks.
We had an opportunity around July 2011 to form a team to see if we could build this product and make it successful and that’s really what we’ve been doing ever since.
Q: Why is saving newspapers so important to you?
A: I have friends who are journalists. The co-founder of this project, his wife is a journalist. I appreciate high quality news and I don’t want to see that go away because news organizations can’t pay for that content or can’t get the journalists to create it. So it’s important to me and I think it’s important to Google to have a variety of high quality content and design, whether that’s from paid news organizations or from bloggers or people who are professionals. We’re trying, basically, to create a way for them to be compensated without the users having to pull out their wallet and pay for that content.
Q: What was it about Google Consumer Surveys, do you think, that earned it the green light to become more than a 20 percent project?
A: I think it speaks to Google’s core mission, which is to organize the world’s information and make it accessible to users. Since publishers were putting up paywalls we didn’t have access to that content. So we were trying to make it available for Google to index, but also for users to access without cost.
Q: How many publishers are participating in Google Consumer Surveys?
A: We have over 130 publishers in the U.S. Three of the top 10 newspapers, seven of the top 15. So sites like the New York Daily News, Christian Science Monitor, and [yesterday] the LA Times. There’s a bunch of high quality papers; Bloomberg is another one.
Q: Do you think a system like this or other paywalls will become commonplace in the near future?
A: Every publisher we had a conversation with—and we talked to every single major publisher in the U.S.—were all considering paywalls, have considered them or tried them. For us, I think this is a way to ease into protected content.
With our survey wall, as we call it, we’re getting completion rates of 30-50%. If you look at a site like the Newton Daily News, which is in Jasper County Iowa, the advertising they could sell on that site was all local and they were getting CPMs (cost per 1,000 impressions) on the order of $2. With us, our average CPM is $19 and our top 30 publishers is something around $38 CPMs. I think the major publishers and the regional papers will have to move to some form of protected content or subscription to compensate for the lack of revenue they’re getting in other places.
Q: What are the capabilities of the surveys?
A: We have 12 formats right now. They go anywhere from multiple choice to image-based questions that allow you to select from one of two images; then all the way to open-ended free form survey questions, which has people type in words or numbers into the surveys and we then do automated analysis on those to create word clouds and group them by similar phrases and sentiment or create average or numerical analysis.
Q: Can you explain the alternate action option?
A: The alternate action is one of the really nice parts for publishers because each publisher wants their reader to do something for them. Sometimes that’s signing into the site, sometimes that paying for the content, sometimes that’s sharing the content on a social network. So if the reader doesn’t want to answer the survey question they have an alternate action to get access to that content and the publisher defines that alternate action.
Q: It’s my understanding that very few people choose the alternate action compared to taking the survey.
A: Yes, the overwhelming majority take the survey.
Q: Why do you think that is?
A: One thing is people like to give their opinion. They know that these responses are actually being used to build better products for them, they’re interesting questions, and they like answering them. Secondly, it’s so easy to do. Once you’ve done it once you understand the process and it’s easier, actually, to answer the question than to do the alternate action most of the time.
Q: How do you get the reader comfortable with the surveys?
A: The first time a user ever comes across one they get a little message before the survey defined by the publisher that explains what the survey is: that it’s anonymous data used by companies to make better decisions and build better products for them. I think that helps when they first see the survey. Of course once they answer one of these questions they understand how it all works. It’s no big deal and we actually see the completion rate go up significantly for second-time users.
We have also seen from publishers that the abandonment rates on their pages don’t change and the time on site doesn’t change. So we’re pretty comfortable that we’ve created an experience that both publishers and readers alike enjoy or at least tolerate.
Q: This also allows publishers to continue to serve the ads on their site.
A: That’s right, they continue to make money from the ads they are serving on the site. We don’t block any of the ads, we just block the content.
Q: Are the surveys available on mobile devices?
A: Today it’s not available to readers on mobile devices. Well, it is available on tablets, just not phones.
Q: But when you say tablets, it doesn’t work with Safari browsers and therefore would not work on iPads, right?
A: That’s true, but you could use Chrome on an iPad and that would work.
Q: Are Google Consumer Surveys open to all marketers and publishers?
A: For marketers it’s open to everyone and for now marketers can only serve users in the U.S., U.K. and Canada. For the publishers we have kept it to an outreach process, where we’re going out to publishers and contacting them so we can control the inventory a little bit to make sure we have high quality publishers in our network.
Q: How much does it cost for a marketer to get started?
A: We have two surveys for marketers. For a general population survey, which can be a representative population sample, we charge $0.10 per response. For a survey where you are targeting based on age, gender or geography, or are screening people out, those are $0.50 per response. For example, “Do you own a dog?” “Yes” or “No.” And since you only want to hear from people who own dogs people who select “Yes” will see questions from the rest of the survey and those are $0.50 per response. That’s the price we have today, but that might change with more functionality, longer surveys or as different types of multimedia are added to the surveys.
Q: Can you explain a little more about who could use the surveys?
A: For the most part it breaks out into four different areas. The first is around the tracking of a brand. We just introduced a new format which allows you to track over time, basically asking the same question so that you can see responses and how they change with time. A lot of smaller companies are using this as brand trackers to understand the awareness of their brand as they gain more marketshare from their competitor.
A lot of companies are testing products. So understanding what are the features that users want to see in products, the colors and logos that they use, and there are a lot of app developers that are testing characters for their games.
There’s another set of users who are measuring the effectiveness of their advertising through our survey. So we show an advertisement to a user and then find that user and ask them about that advertisement. So here we can measure the effectiveness of a brand campaign where normally it would be much harder to do so.
Finally we see a lot of socioeconomic research. There are economists at universities who are running surveys to understand how people feel about different types of questions and socioeconomic status.
Q: When you’re measuring the effectiveness of brands, those are called Brand Lift Surveys, correct?
A: That’s right.
Q: So what are you working to accomplish in the next 6-12 months?
A: I guess I can talk about what we’ve heard our limitations are to our product and what we’re doing to adjust those. The biggest limitation is the one or two questions at a time. There are a lot of researchers who want to do cross tabs or compare the answers of one question to another question in their survey and they can’t do that today because of that limitation of 1-2 questions at a time.
We also want to expand our publishers base, both here in the U.S. and internationally to all the countries that Google supports. To do that we either have to sign up publishers or become our own publisher through applications on mobile phones or use our own properties to do the surveys, so you may see some of that in the future as well. Of course the core business of Google is ads and we want to make sure that advertisers feel they’re getting their money’s worth and they’re spending their money with us; and so we’re working to improve the ad experience and we’re measuring the effectiveness of those ads and helping advertisers build better campaigns.
Q: Anything else the readers of Forbes should know?
A: Really we’re just trying to build a product where businesses of all sizes can use data to make data-driven decisions for their companies, whether that’s a Mom and Pop shop on the corner or a big company, everyone can use this service. I think once you understand the power of that data and how to use it, you can build better products.