Whatever else it may be remembered for, last February’s Super Bowl will be recalled for the war of TV commercials waged by makers of free mobile games to drive installs and get new users.
Although model Kate Upton’s fetching appearance in an ad for “Game of War” got much of the attention, Super Bowl ads for “Clash of Clans” and “Heroes Charge” also convinced large numbers of viewers to get those games.
This week, mobile ad agency Fetch is out with a new “TV Uplift Dual Screen Optimization” report that indicates you don’t need Kate Upton or the Super Bowl to move the needle on app engagement driven by TV ads. The apps also don’t need to be games.
It found that TV ads have an immediate effect on installs, plus some extended decay. Although Fetch declined to provide the names of its clients’ apps used for the report, it did say they were mostly transactional rather than games.
The report said that app installs in the viewing area of a TV commercial demonstrate uplifts ranging from 56 and 74 percent during the ad airing. For the next 10 minutes after the airing, there’s an increase of 24 percent, which then continues at about that rate for about two hours afterwards.
This causation points to the connection between TV and mobile behavior: many people watch TV these days with smartphone or tablet in hand, and they tend to use them during the commercial breaks. The Fetch report theorizes that TV ads may be “the final tipping point” in app marketing efforts.
Although the report is not exhaustive, it does show that some times and some techniques are more effective than others.
For instance, frequency in a short period of time can be very effective. If four ads are shown within two hours, there’s a whopping 467-percent increase in installs during that time.
The report did not specify if the ads were shown on the same channel or in the same program, but it did recommend shorter TV ad campaigns of higher frequency. Fetch Head of Data Dan Wilson told me via email that the frequency effectiveness appears to peak at four ads within two hours.
That four-in-two frequency can also boost in-app conversions — where users make in-app purchases, like buying higher levels or app-related products — by 316 percent within those two hours.
Time of day can also be a big factor. Six to 7:00 p.m. generally has a high level of viewing, and ads shown then result in a 500-percent increase in app installs. This drops by 50 percent within an hour and then decays rapidly.
But there’s also a huge increase of 650 percent in app installs for ads shown in the midnight to 1:00 a.m. slot.
This finding that the highest uplift is outside of peak viewing hours is one of the surprises in the report, Wilson said.
He added that, in part, this is because the “the baseline is very low.” While the actual numbers of installs weren’t given, a relatively small number could represent a large percentage increase.
Wilson surmised that users watching TV during the wee hours have “a high response,” which assumedly is because there are few distractions — other people, phone calls, work and the like.
The report did not look at types of ads, lengths or the channels/programs where the ads were shown.
Based on 237 TV ads and more than 50,000 mobile actions, this report is Fetch’s first on this subject, although the firm has previously released data-driven studies of other mobile behaviors.
The app install data comes via ad measurement firm Mobile Measurement Partners from Fetch clients using Tune and AppsFlyer attribution services. Fetch matched TV schedules with first-party app data for the same location and time.
A baseline was established five minutes prior to the TV ad airing for app installs in the same location, on the idea that external factors were not likely to change within those five minutes.