Hollywood dives into data, but decisions aren't easier
The process by which Hollywood conducts market research on its movies has changed little over the past 20 years. Groups of potential ticketbuyers are gathered, shown footage and asked, “What do you think?”
Their answers can dictate how a studio will market a film, at what cost, and what the final film will wind up looking like. Their early input can also influence whether the pic gets made in the first place.
But the Internet was supposed to make things easier for marketers — or at least make the opinion-gathering process faster, more focused and cheaper. And while some of that evolution has happened, the resulting constant flow of information and instant access to market research can now make any executive’s head swim.
Google searches, activity on social-networking sites, Twitter and blogger chatter, website traffic and online trailer views have joined the traditional tools — focus groups, test screenings, tracking reports, Q Score ratings and CinemaScore results — that studios have long relied on.
No one is objecting to the new tools. But few are quick to embrace such new media data as a truly reliable representation of that behavior, either.
“There’s a lot of noise right now,” says one studio marketing topper. “Studios are flirting with a bunch of (those tools), but they’re still trying to figure out what to do with them.”
The reason is that research firms have yet to determine the power of a tweet and the impact of the “like” button on Facebook. Just because individuals are buzzing about a project doesn’t mean they’re going to buy a ticket to see it, analysts caution.
Consider the example of Justin Bieber’s docu-concert pic “Never Say Never.” Online word-of-mouth was so strong ahead of the pic’s Feb. 11 release of that one might have expected it to have grossed $500 million during its opening weekend. It actually did $30 million — a still impressive tally, but second to the $31 million for Adam Sandler laffer “Just Go With It.”
“To date, Twitter has been found lacking as a box office prognosticator,” said Gordon Paddison, CEO of Stradella Road, which develops marketing strategies for Peter Jackson and studio clients.
Internet-centered market research is still finding its footing, but it’s a growing part of the metrics tapped by studios and execs looking to hedge their bets.
Significantly, it’s made the process cheaper. Research firms can sample a larger percentage of the moviegoing public at lower cost, by conducting more focus groups online and having respondents answer questions directly via websites rather than by tedious phone surveys.
“Studios had been suspect of online because it was early and people weren’t online as much,” says Grant Johnson, a former exec at Nielsen’s NRG, who is now president and co-founder of market research firm Interpret, focused on the convergence of new media, entertainment, technology and advertising. “But penetration rates are so high now that your moviegoing universe is online.”
Johnson adds that the latest crop of new-media platforms — such as Trendrr and other conversation curators — are having a significant impact on consumer behavior.
“We just need to know when it has an impact,” whether it’s a week before a film bows or even down to an hour before it hits the bigscreen, Johnson says.
Established social-media sites provided the foundation. Facebook boasts more than 500 million members. And while Twitter is used by just 9% of adults in the U.S. each month (or 16.4 million,according to eMarketer), the chatter on the service is picked up by the media and bloggers as a gauge of public opinion that’s offered up to millions more. EMarketer expects 28 million Americans to be tweeting by 2013.
Marketers must spend more time analyzing those opinions, however, before using them to shape decisionmaking.
“It begins with the question, ‘Is what they’re saying real?’ and ‘Is this truly the topic they’re talking about?’?” Johnson says.
One researcher points to Sony’s “Salt” as a recent case of marketers being misled by Twitter. The film’s title saw a peak in tweets two weeks before its release last July, prompting marketing execs to believe positive word-of-mouth was spreading about the Angelina Jolie actioner.
But when analysts delved into why the film was trending so high, the result was attributed to Bieber Fever and the young musician’s coincidentally timed trip to, yes, Salt Lake City.
“The natural language-processing programs that social media services are using to distill and filter out the crap from the actual meaningful information is improving. But even if it’s accurately coding the language people are using, it doesn’t mean people will be acting on that in a predictable manner,” Johnson says. “The problem we have as researchers is determining when something will be predictive of behavior. We know the old tools tell us that.”
Marketers are also questioning just who’s tweeting in the first place. Many may be overeager fanboys not representative of the larger public.
“The people we’re trying to reach may have jobs and can’t tweet all day,” says one marketer.
With billions of dollars potentially at stake, it’s no surprise studios have conservatively embraced the tried-and-true testing methods they’ve used for the past 20 years. That’s especially true when the success or failure of a film’s opening means a studio’s marketing maven is “considered brilliant or unemployed,” as one researcher put it.
“Every time some new media channel gets introduced, it slowly gets incorporated into the mix,” Paddison says.
But marketing mavens like Paddison say that the studios have mostly relied on the same methods for market research they’ve used over the past decade or longer. They’re comfortable with the data they’ve been getting and are inclined to stick with what works.
One reason for that is it’s been hard for research providers to “make heads or tails of what all this new data means,” Johnson says, and the numbers being offered by social-media measurement companies aren’t yet deemed reliable enough for studios to act upon. “There’s still a ghost in the machine,” says one studio exec.
Instead, studios turn to companies including Nielsen’s NRG, MarketCast, OTX and Screen Engine that process tracking reports or conduct test screenings, focus groups and exit polls.
Nielsen Entertainment Research also launched NRGi, a market research arm for the indie film sector, in 2005. It aims to serve “all of the arthouse distributors and each of the independent-minded minimajors — even the filmmakers themselves — that are trying to figure out what’s working and what’s not about their films,” said NRGi general manager Joseph Craig at the time.
Studios need to attract the core of their ticket-buying audience, 18- to 24-year-olds, who bought an average of seven tickets per person in 2010, down from eight in 2009, and who are often influenced by word-of-mouth, increasingly generated online. At the same time, the studious are simultaneously courting older auds, which have grown 67% since 1995, according to research firm GfK MRI.
More marketers are using the Web to reach out to new targeted focus groups like the teen and tween girls that companies like Alloy Entertainment have learned to speak to, as well as an increasingly vocal and influential group of parents known as “mommy bloggers” that studios like Disney look to appeal to with their films. Other entities like Marvel Studios are known for offering up casting rumors to websites to test whether fanboys would be interested in a thesp in a super suit.
But it’s hard to avoid the ironic fact that despite having the ability to zero in on specific demos with marketing campaigns, the s
tudios largely aren’t creating pics for narrow segments of the population.
Release schedules are now made up of sequels, remakes of previous hits or adaptations of comicbooks, young adult novels, toy lines and videogames — titles considered safe bets at the box office given their familiarity among wide audiences. That’s not expected to change anytime soon, since many of those movies also have the potential to become major moneymakers for a studio’s other divisions, including homevideo, TV, videogames, consumer products and licensing, and theme parks.
The need for better tools is even more essential now that audience attention is being pulled in so many different directions. Where live TV has long been the dominant marketing tool for studios, viewership keeps eroding. Last year, 52% of the U.S. population watched TV live, but the tally was just 41% among 18- to 34-year-olds, according to research firm Morspace.
“It’s clear tracking is an insufficient tool,” Johnson says. “There’s room for innovation. Hopefully social media can help change that. Data fusion is the future, and we need to marry old sources with new methods. If we don’t, it’s clear we’ll be left in the dust.”
Kevin Goetz, CEO of market research company Screen Engine, says studios would be best served by finding a single research provider they can trust, especially given the surge of info and analysts available now. “People don’t follow companies; they follow relationships,” Goetz says.
But, he admits, “It’s all about your interpretation. When a studio comes up with two disparate results, who do they listen to and what should they listen to?
“You become crazy if there are too many hands involved — someone doing biometrics, followed by online testing, then offline mall samples,” Goetz says. “It doesn’t mean any one of those parties is wrong, it just means whoever wins (the ear of the customer) becomes the trusted advisor, and that’s really the crux.”
Even with all of the various forms of data at their disposal, studios can still be surprised.
Few analysts expected Warner Bros.’ actioner “Unknown” to beat DreamWorks’ teen-skewing “I Am Number Four” during its opening weekend, but it did. The Miramax-produced “Gnomeo & Juliet,” released by Disney, wasn’t expected to perform well, but the toon’s short stubby legs have proven durable over the past several weeks. And few would have predicted that Oscar nominees “Black Swan,” “True Grit” and “The King’s Speech” would each gross more than $100 million in the U.S. as they have.
“There’s a ton of data, but it’s still hard to distill it and know what it all means,” Johnson says. “The key is making sure it’s reliable and provides a valid read on consumer sentiment so that we can then use it to predict future behavior. Even though there are pitfalls, there is no question we have to figure it out.”