October 10, 2001
Case Study
|
SUMMARY:
Despite all the media reports of banners "not working anymore", marketers at Fleet Boston's Consumer Lending Group knew their target demographic for home equity loans was spending more quality time than ever online. Consumers routinely used their work Web connections to surf during the lunch hour, and moms at home surfed late at night after the kids were in bed. So, Fleet devised a series of three test campaigns to learn what would work for them online without too much risk. (BTW: Big media buying surprise in the results.)
|
|
CHALLENGE
Back in early 2000, the media was packed with stories about how online advertising didn't work anymore. Banner click through rates were dropping, etc.
But the marketers at Fleet Boston's Consumer Lending Group knew their target demographic for home equity loans was spending more quality time than ever online. Consumers routinely used their work Web connections to surf during the lunch hour, and moms at home surfed late at night after the kids were in bed.
So, Fleet tasked online ad agency Slingshot LLC with the challenge of figuring out how Fleet could generate more home equity loan applications via online advertising, without too much risk.
CAMPAIGN
Kristy Ladner, Slingshot's Senior Account Exec, told us that instead of rolling out one massive online campaign they decided to run a series of three tests from early 2000 to mid-2001.
Test Campaign #1 - Find out which media and messages work
Slingshot created 10 different banners for the first campaign in early 2000. These ranged from banners that offered general information on home equity loans, to banners that named specific goals consumers could accomplish with a home equity loan such as buying a new car.
Initially Fleet ran all ten banners across a dozen Web sites and ad networks -- from financial sites to women's sites. Then every two weeks, the Slingshot team studied results and tweaked banner creative and media buys for better results. Sites and banners that didn't work well were dumped. All media buys were geo-targeted when possible to reach consumers in the New England area where Fleet is located.
By the end of the campaign, they ended up with four winning banners and, coincidentally, four winning sites.
Test Campaign #2 - Test Pricey SUPERSTITIALs(R) vs. Cheap Banners
For the second test campaign, which ran from October 2000 - February 2001, Slingshot decided to throw a new creative test into the works -- they created a SUPERSTITIAL ad to test against the winning banners from the previous tests.
This SUPERSTITIAL ad was essentially a 20-second online commercial with Flash and audio, which appeared fully loaded when consumers moved between pages in a particular Web site. (In comparison, interstitial ads load at the same time as a Web page, which slows down viewing and gives the consumer time to click 'close' before being persuaded to watch the ad. Pop-unders, such as the infamous X10 micro-camera campaign, open up behind the scenes so consumers only see them when they close Web pages entirely. Of the three formats, SUPERSTITIALs are widely agreed to be the least annoying.)
The SUPERSTITIAL's creative (see link to sample below) used a strongly emotional pitch focusing on the psychological benefits of home equity loans. 'Stop worrying about how to pay for things and focus on the more important things in life.' Ladner says, "There were school kids on swings in the background. It was very pulling on the heartstrings."
Along with geo-targeted selections on their tried-and-proven sites, Slingshot also tested a few new sites in niche categories such as regional news and gardening.
Test Campaign #3 - Goal test landing page options
As the results came in from test #2, Slingshot began to plan the next test campaign, which ran from March - June 2001. Again they kept what worked -- in this case the best of the banners, the best of the media buys and the SUPERSTITIAL.
Along with a few new test media buys, they threw a new additional element into the mix -- a landing page with two options. Prior to this, the click through landing page simply offered visitors some benefit copy and a link to fill out an online registration form for a home equity loan. But, now the landing page had two options. Consumers could either fill out an online form, or they could click on a link to get more educational information about home equity loans. This second section included a prominent locator device to help visitors find the Fleet Bank real-world branch nearest them.
RESULTS
Surprisingly, ad network and site specific buys for topics such as women's sites outperformed both financial and regional news sites. Ladner says, "It was not at all what we were expecting. We thought people interested in money would be interested in online loans. But niche sites targeting women, or gardening, or travel worked well because they were emphasizing end-uses for a loan. If I'm looking at a travel site, I may want to go on a vacation and here’s how to pay for it."
As you might guess, the banner creative that worked best also spoke directly to end-uses for loans, rather than loans themselves. Ladner says, "'Use a home equity loan to pay for your child's education' had a better click through rate than ones that just said, 'Learn more about home equity.'"
As you also might guess, the SUPERSTITIAL pulled substantially more clicks than the banners did. However, SUPERSTITIALs can cost at least ten times more than a banner to create (which is why Fleet didn't test a variety of creatives with this format.) Plus, while it can cost about $5 CPM (cost per thousand) to run a B2C geotargeted banner campaign, SUPERSTITIAL media buys can cost $35 CPM or more. Last but not least, high clicks don't always mean high success. Ladner notes, "Occasionally we'd see banner creative with a relatively low click rate have one of the highest registration rates, and flashy banners with high clicks do vice versa."
Despite all these factors, the SUPERSTITIAL proved to be both the most economical and powerful format. The first time Fleet tested the format in Winter 2001, the SUPERSTITIAL ad generated 20 times more loan registrations per impression than the next-highest performing banner ad did. The second time out, the SUPERSTITIAL was responsible for 87% of all loan registrations despite the fact that it accounted for only 30% of impressions purchased.
Fleet also learned that it's worth it to offer an informational option on a landing page. On top of the normal number of registrations, an additional 50% of click throughs chose to take the informational route rather than immediately applying online. Of these, a full 27% used the branch locator to find a brick-and-mortar Fleet office near them.
Useful links related to this story
Fleet
http://www.fleet.com
Slingshot
http://www.davidandgoliath.com
Link to sample of Fleet's SUPERSTITIAL creative:
http://www.unicast.com/superstitial/?type=cs&id=20&cid=68
