LOS ANGELES Last week the U.S. Census Bureau awarded DraftFCB its $200 million contract for marketing the 2010 census.
Adweek talked to Jennifer Marks, chief of the Census 2010 Publicity Office, about the future of counting Americans. Marks, a demographer by training, worked for the Population Reference Bureau just out of University of Maryland grad school before joining the Census Bureau 29 years ago.
Q: Why does the Census have to advertise?
A: There are a couple of answers to that. Historically, we saw that the 1990 [census] mail response had dropped and that the undercount had increased. And when we looked at our PSA-based campaigns, we could see that we were reaching a lot of people in the middle of the night, not in the key viewing or listening hours. So we went to paid advertising for the first time in 2000.
Was that a reflection of how PSA time was allocated to the bureau specifically?
It was increasingly difficult to get free time. And many of the media channels, such as radio stations, were dedicating public-service minutes to their pet causes.
Why did the marketing budget rise in 2000?
It started at $100 million and increased because of the sampling controversy. And we started to use a representative sample until the Supreme Court said we had to go to every household. In order to be effective, we expanded the advertising effort from $100 million to $167 million.
Did it work?
It was very successful, and that was not just attributable to advertising. There were partnership efforts. We got earned media. We had a "census in schools" program and road-tour vehicles. All together, we reversed the decline in mail response dramatically. Because of the trend, we were headed for only 55 percent response and we were planning improvements that projected a 61 percent response. We wound up with 67 percent. Of course, proving return on investment is as difficult for us as anyone.
What's your overarching strategy?
We're doing this planning a little differently. We want to have DraftFCB be responsible for an integrated communications plan as its first task. The last time it was just an ad contract and we tried to do the integration internally. This time we want to give DraftFCB a lot of time to interact with us and with stakeholders [such as target audiences and the American Marketing Association advisory board].
What are your goals for 2010?
We have three: To increase mail response; improve accuracy; and improve the differential undercount.
What are the marketing tactics you plan to employ in 2010 that you did not, or could not, in 2000?
We definitely want and expect DraftFCB to recommend new technologies, but one of the things I mentioned to them today is to be careful, because the technology is changing so rapidly things could be dramatically different two years from now when the campaign begins. So we asked them to build flexibility into it. Obviously, there is a lot of technology that's new for us, at least—addressable TV, addressable media, for example. We didn't really know about [all] that last census. But cellphones, podcasts, blogs...will that be happening two years from now? And we're considering the extent to which consumers should create the content. A friend of mine just told me her [child], a high school senior, should be doing a YouTube video for us. That's something we will have to think about.
Because of the controversy over undocumented workers in recent years, how will you market minority outreach?
It will be a huge focus for us. We are very concerned about it as a hot-button issue and how that will affect the census. It has the potential to have a huge effect on people's willingness to participate. So we're doing a lot of research and it's intensive with the Asian, Hispanic and black populations. We're determined to accurately portray the diversity of the nation. We see that with some racial and ethnic groups, the marketing has to be like a personal invitation.
How is that number determined?
In terms of advertising, we first look at our own data about population groups and linguistic isolation. But also you have to have a media channel [in that language] or there is no point.
What did the Census Bureau learn in 2000 that it plans to apply to the new census?
We learned that advertising and partnership works. We have had some statistical evaluations of this, but we're back to the problem of proving anything. That being said, we feel that partnership has had the most impact on racial group marketing. Again, it's hard to prove, but there is some evidence that partnership marketing had an impact on racial participation.
Did it work particularly well with any one group?
With Hispanics it worked well. Many regional offices had partnership specialists. We worked with Target, Wal-Mart and neighborhood associations. In Orlando we partnered with the Linx bus network with bus wraps on heavily Hispanic routes.
What was the aspect of DraftFCB's presentation that distinguished them?
All of the companies were very strong. It was an outstanding set of companies. Ultimately, the decision was based on the best value to the government—a combination of Draft's technical approach and cost proposal.
That sounds procurement-oriented. Was there no room for creative?
We let them submit creative, but we didn't evaluate it a lot per se, because it is very subjective. We were more interested in how they'd target segments. We wanted to know their process. We asked about their management approach. And their subcontracting plan was very important.
How do you plan to allocate media? What is the split between traditional and nontraditional that you anticipate?
It's still too early in the process. It will depend on the research, on what characteristics of the population and geography might merit consideration. I don't know of many companies that have as their goal to reach everyone in America, in 20 categories and 17 languages.
Is there a tagline or theme you favor?
We want Draft to gather information first, come up with an integrated solution, and then execute it. But procurement is a long process. To have that done and be ready to move on is exciting.