Google Tested 3 Versions of This Honey Maid Ad to See Which Worked Best Online

How a :15, a :30 and a 2:17 compare in boosting certain metrics

Headshot of Tim Nudd

It's a fundamental choice for online video marketers: Should you hit viewers with a quick, memorable pitch that's (hopefully) less skippable? Or should you engage them with longer-form content that's more captivating but runs the risk of being tuned out?

Clearly, it depends in part on what you're selling, to whom, and what your immediate business goals are for the content. But are there general lessons to learn from shorter- vs. longer-form online video?

Google thinks so, and recently partnered with Mondelez International to try some tests.

They took three different cuts of the same Honey Maid graham cracker commercial—a 15-second version, a 30-second version and a 2:17 extended version—and tested them using TrueView, YouTube's skippable preroll ad format. Then, they measured how viewers responded in two separate ways—how much of each version of the spot they watched; and how each version affected ad recall and brand favorability.

The spot being tested was made for National Hispanic Heritage Month in 2015 and explored what it means to be American through the eyes of a Dominican family of immigrants. Here's how Google described the three versions of the spot, and the theories, going into the experiment, about how each one might perform: 


The 15-Second Cut:

The shortest version of the ad, with a voiceover from the father, includes scenes of the family together and ends with the brand's logo and tagline. The product appears at the six-second mark, and either product or logo is present for a total of five seconds, or 33 percent of the total runtime.

The Theory:

The short length of this YouTube ad will make it less skippable, without sacrificing the narrative or brand lift effectiveness. The balance of story and brand in a concise format will hold the viewer's attention and create a connection to Honey Maid.


The 30-Second Cut:

The longer cut gives more detail and dimension to the story, with scenes of the father headed to work and the family playing together. While the longer format adds more Honey Maid product shots, the relative amount of explicit branding is roughly the same as the 15-second version. The product first appears at the 11-second mark, and either product or logo is present for 10 seconds, or 30 percent of total runtime.

The Theory:

This video will draw viewers in with a more in-depth story, and is still relatively short. It is the best of both worlds—short enough to keep viewers entertained and long enough to create a meaningful impression. 


The Long Cut (2:17 runtime):

The longest version adds further depth to the family's story. In addition to the father, the viewer hears from the mother, daughter and grandmother (who speaks in Spanish). Like the other ads, the themes of family and celebration are highlighted. The product does not appear until 1:17, and either the product or logo are present for only 12 seconds or just under 9 percent of the overall runtime.

The Theory:

The long-form version builds tension by illuminating some of the Gomez family struggles, which adds a richness to the final scenes of celebration. The layered story that reveals more facets to the family will pull viewers in and keep them engaged. 


For years, Google has preached about power of longer-form content on YouTube, pointing out that there's a consistent relationship between how long an ad is viewable and increases in brand awareness and consideration. Also, viewers are clearly willing to watch long ads on the platform, as Adweek and Google's monthly YouTube Ads Leaderboard regularly shows.

But the Honey Maid experiment offered more nuanced results.

• The :15 was the most skipped of the three versions, and the least effective in lifting brand favorability. But it was the most effective of the three in driving ad recall. Thus, this kind of short format would seem like a good choice for brands focusing more on awareness and less on favorability.

• The :30 was watched all the way through more than either of the other versions. (It was 30 percent higher than the :15 in that regard.) It was also better than the :15 at lifting brand favorability. As the initial theory suggested, the ad was short enough to keep viewers entertained yet long enough to create a meaningful impression.

• The 2:17 was skipped less than the :15 but more than the :30. Only about 15 percent of viewers watched the 2:17 all the way through. While this about double the typical benchmark for consumer-packaged goods videos of this length, it still meant many viewers never saw the Honey Maid branding at all—since it didn't appear in any form until 1:17. Like the :30, the 2:17 also was good at lifting brand favorability. But Google suggested weaving the brand into the storytelling earlier for best results.

In its final analysis of the experience, Google still argued more for value of longer-form video, despite the bright spot in the research for the :15 (its stronger ad recall) and the limitations evident in the extended cut (the loss of branding opportunity in this example.)

Google summed up the tests this way: "The challenge is the same now as it was in the beginning of advertising: to figure out how to blend story and brand successfully. This has never been tougher, as we're now competing to reach people who are hit with thousands of messages a day from every direction. This media pressure can lead brands to feel like everything needs to be faster faster faster. But, as this experiment showed, making ads shorter doesn't get them more attention—it may get them even less. With a great story, brands can take the time to create a connection, and change a mind." 

@nudd Tim Nudd is a former creative editor of Adweek.