Push ads across devices to drive click-throughs, conversions: NYU professor
November 18, 2013
A location-based ad from Best Buy
NEW YORK A New York University professorat the Mapping Mobile @NYUStern conference said that pushing ads to consumers on both desktop and mobile can increase click-through rates and conversions.
Theprofessor found that increasing a discount by 1 Euro has the same effect as decreasing the geographical distance of a consumer to a store or restaurant by 83 to 210 meters. He also found that advertising on both desktop and mobile greatly increased click-through rates and conversions for retailers.
Mobile ads often get a bad flack, said Anindya Ghose, professor at NYU Stern School of Business, New York. Every now and then youll hear that mobile ads dont work.
Oftentimes were missing the fact that mobile ads can lead to clicks on mobile devices but the final conversion might happen on PC, he said. There are these spillovers that are important for us to quantify. If you only look at attribution through one channel you dont get the whole story.
We have a model now that can tell us guidelines. If I have $10,000 to spend how do I allocate my budget. With retailers in Europe we are working on dynamically generating these offers and looking at tradeoff.
Mr. Ghose shared findings that show how retailers can make their mobile ads give them the most bang for their buck.
Mr. Ghosesfirst study looked at the impact of showing consumers the same ad on both desktop and mobile.
While this may entail a risk of irritating a consumer by showing him an ad too often, the study showed that it can actually have a positive impact as well, by reinforcing a brand.
Mr. Ghose tested this theory in South Korea by manipulating ads for ebooks on a Web site over six weeks. He looked at the differences in click-through rates and conversions based on whether or not a consumer saw an ad on a PC, on a mobile device, or on both.
The results showed that when an ad was only on desktop, there was a 1.18 percent click-through rate, whereas when the ad was displayed on desktop and mobile there was a 1.59 percent CTR. Conversions on desktop-only ads were 3.52 percent, whereas on desktop and mobile they were 4.78 percent, 36 percent higher.
There was a similar increase in overall sales, with desktop and mobile ads showing a 97 percent higher outcome than desktop-only ads.
This supports the idea that pushing ads across channels can reinforce a message and drive to higher click-through-rates and conversions.
Going the distance
Thesecond study looked at a German app that shows deals at various retailers. Mr. Ghose ran a number of different focus groups to test the effect of location, ranking and price on a consumers likelihood to redeem an offer.
The test spanned 14 weeks and looked at 3,900 participants. For each variable location, ranking and price Mr. Ghose created four different focus groups.
For location, for instance, there was a group that saw offers sorted by distance with distance information, another saw randomly sorted deals with distance information, a third group saw sorted by distance without distance information and a fourth saw randomly sorted deals without distance info.
The study yielded three significant findings.
If a retailer increased the discount by 1 Euro, it had the same effect as decreasing the geographical distance by 83 to 210 meters. Improving display rank by 1 unit, meaning moving the offer up to the top of the display, had the same effect on coupon redemption as increasing the discount from 2.25 to 3.1 euros.
Lastly, reducing display rank by 1 unit had the same effect as increasing distance between the consumer and store by 130 to 190 meters.
By looking at actual data and creating a scientific model, Mr. Ghose is now able to advise retailers on what is the perfect mix of location, price and ranking for them within a similar app.
This can also be translated to mobile ads outside of an app, by helping them determine how much of a discount they want to offer consumers based on how far away they are from a store or restaurant.
The Mapping Mobile session
Miklos Sarvary, a professor at Columbia Business School, New York, ran a study that looked at how retailers can leverage a consumers location history to better target ads.
The classic example of leveraging location for mobile ads is pushing a personalized message to a consumer when he or she is nearby a store to drive them inside. However, Mr. Sarvary is interested in using location to target consumers that are beyond 100 feet away from your store.
We know a lot about consumers carrying their mobile phones, Mr. Sarvary said. Can we extract some information about you and make the advertising more effective, understanding your behavior, data about your location, customer lifestyle, and maybe we can improve the segmentation and targeting of our advertising.
Mr. Sarvary created a model by running a small experiment in 2012. He asked consumers to download an app that gave them coupons and asked them fill out random quizzes.
The study had 265 subscribers, but they could also refer the program to a friend.
Mr. Sarvary used a regression model to trace the subscribers location from January to April to find colocations, when a consumer was in the same place as another consumer at the same time. He used this data to draw conclusions about consumers and their behaviors.
For example, if five consumers were at the same location for a long period of time every night, they may be a family. Or if a group of consumers visited various locations at the same times, they may be a group of friends.
According to Mr. Sarvary, information derived from colocation data can be more useful to a retailer than simply knowing that a consumer is right by your store at one certain point.
The logical thing is I attract you to a store if you pass by, Mr. Sarvary said. But we have much more data with people carrying their phones around.
When you move around, youre not just hanging around, you have an objective, youre walking to school or work, so its difficult to engage you and make you depart from your original objective, however just by moving around you generate a lot of information about yourself, he said. I think where there may be a goldmine is learning how to use this information much more so than asking you to do things.
Rebecca Borison is editorial assistant on Mobile Marketer, New York