1MPACT LA - Analytics and Optimization Track
With one person and three tracks, as well as keynotes and vendors with some interesting services, I could only cover so much of the show. I was able to make a few sessions in this track. Each session in this track was standing room only with over 50 chairs set up in the room.
I was quite impressed with Young-Bean Song and his in depth understanding and analysis. He had some great data on frequency with some insights that many might overlook. When looking at the average frequency of impressions for a campaign, the results can be skewed by the super active people who see the ad 20, 30, or even 50 times. Frequency capping only works per ad serving solution, and most large campaigns are spread around with more than one server, so you will likely get readers seeing your ad more than your cap amount. Even so, capping does help as it limits the overexposure, as conversion rate goes down as impressions increase.
In addition, he has see that daypart targeting along with frequency caps increase CPA by 10%-30%. All campaigns are different, and you need to compare your own lift against the increase costs for setting caps and dayparts into the ad serving solutions. Song also noted that most capping is not accurate after 24 hours.
The best thing I got out of this however was an observation about delayed conversions. Because we can analyze everything now, everyone always compares each creative and campaign to the previous one as soon as possible. They often find that the new one is not performing as well as the one that has been around for a few weeks, even though it tested well and should be doing better. The reason is that they are getting delayed conversions sparked by the previous campaign. Users set bookmarks, do some research, procrastinate, or whatever and then go ahead and pull the trigger and take action days and sometimes weeks later. This can make the new campaigns that do not have any history appear to be worse than they may actually be.
The next talk that had some great new information was from the Consumer-Generated Media talk. When Petco added consumer reviews (powered by Bazaarvoice) to their site (good and bad), they were able to increase sales and conversions. The idea it to align marketing priority with the customer voice. Thus, they were able to find the "hidden gems" that consumers really loved so they could highlight and promote those products. They could use reviews for highly rated products in email campaigns. What they found in an A/B test was that the people who followed the "ratings" path had a 35% higher conversion rate and 40% increase in sales per visitor than this using the traditional product navigation path.
