by Adam Sutton
, Senior Reporter
Retargeted display ads had worked well for Corel Software. The consumer software company tailored the ads to recent site visitors based on the products they had viewed. The ads ran on a major network and earned a good profit.
But, not everyone was satisfied. Derek Tucker, Senior Global Marketing Manager of e-commerce, Corel, thought a better approach to targeting could earn more revenue.
"At the end of the day, I was retargeting and spending based on the fact that somebody had been to a page on our site. That is all I knew," he said.
In January, Tucker's team transformed Corel's program into what he calls "hyper-retargeting." The approach used multiple criteria to tailor the ads throughout the buying cycle. Revenue from the ads has since doubled.
"Over the last 30 days, from a year-over-year perspective, the revenue we are pulling in from retargeting banners is up 106%," Tucker said.
Here are the steps the team followed to overhaul the program.
Step #1. Pick the right products
A retargeting program is not for everyone, Tucker said. The companies who are most likely to benefit will have an extended buying cycle and customers who visit the site six or seven times before buying.
"Someone is not going to visit your site six times before they decide to buy a pizza from you versus someone else," he said. "However, if you have a situation where people are making multiple visits, then to me, site retargeting is absolutely a no-brainer."
For example, people who buy Corel's professional graphic design software, CorelDRAW, typically visit Corel's site multiple times and download a trial version of the software before buying.
"To me, retargeting is one of those pieces that helps remind the customer to come back for that second, third, fourth, and fifth visit," Tucker said.
Step #2. Map the buying cycle
For each product the team planned to promote, it poured through Corelís data to answer two questions:
- How long is the average buying cycle (i.e., how much time elapsed between the average customerís first visit to the site and the purchase)?
- How many times does the average customer visit the site before buying?
Find the steps to purchase
Next, for each product, the team outlined the steps on the path to purchase. For example, here are the steps for the team's CorelDRAW software:
- Visit the product page without bouncing (stays for at least 25 seconds)
- Download the trial
Step #3. Segment the audience
The steps on the path to purchase naturally segmented the prospects into groups. For example, here are the corresponding segments for the three steps mentioned above:
- Segment 1: prospects visited the product page for at least 25 seconds
- Segment 2: prospects downloaded the trial
- Segment 3: prospects purchased the product
The team planned to target the prospects in the first two groups with ads that encouraged them to take the next steps. Prospects would see the ads through a display network after they had left Corel's site and browsed elsewhere.
To make this work, the team had to coordinate its analytics platform with its display network. Once a prospect completed a step, the analytics system communicated that event to the network, and the network changed the ads. That way, when prospects downloaded a trial, they no longer saw ads for the trial and instead saw ads to purchase.
Match frequency and timing
Tucker did not want to bombard Corel's audience with ads. His team capped the number of times it showed the ads, and the level corresponded with each personís position in the buying cycle.
For example, the team segments prospects by the number of days they have been in the program, such as:
- 0 to seven days
- 8 to 15 days
- More than 15 days
Prospects who have been in the program for less than one week see the most ads, and each subsequent group sees fewer.
Because the ads are sold on a cost-per-click basis, critics have suggested that Tucker uncap their frequency. He will be able to show more ads for the same price, they say. This makes sense at first glance, but Tucker has yet to see it proven with data, and his personal experience tells him it might be harmful.
"We are certainly on the high-end of frequency in the industry, but there is a limit to how far you want to push that," he said. "At some point, I am going to go beyond the point of just staying in front of somebody to the point of upsetting them, irritating them, and I could actually get to the point where I could create negative impressions on my brand."
Step #4. Create the display ads
For each product, the team created ads that corresponded to each step on the path to purchase. The ads changed when prospects completed a step, sometimes in subtle ways.
For example, one ad
encouraged prospects with a "Free Trial" button as the call-to-action. A later ad
changed the button to "Buy Now" and left the style and colors unchanged.
Constantly test the message
After the team established the program, it continuously tested variations of the ads.
"Don't assume you know what the customer wants," Tucker said. "At any point in time, you should have at least one A/B test going for each one of your audiences."
Tucker's team often tested the text of the ads. For example, a testimonial would be tested against a message based on the features and benefits of the product. The best approach varied by product, Tucker said.
"As a general rule, I've found that the more high-end the product, the better the testimonials work," he said.
- Free trial ads
- Buy now ads
Related ResourcesGoogle Display Network
— ad network the team used for retargetingB2B Retargeting: 5 tactics from burn pixels to FacebookRetargeting: 5 tactics from drip email to lead generationOnline Advertising: Retargeting drives 3% to 7% in incremental topline revenue for CafePressBanner Blindness: Why your marketing messages are hiding in plain sightOnline Advertising: The 3 obstacles you must overcome to create an effective banner ad