November 7, 2011
So far in the "Strategies for Successful Testing" series, we've looked at:
In this post, we take a closer look at the third of four questions (Who, What, When, and Why) that help you structure a great test campaign: When should your test run?
I come from a SEO/PPC background and, in my experience, most search marketers use dayparting incorrectly. So the (shaky) logic goes, "Most prospects shop between a certain set of hours, so you should only run your ads during these times." But unlike radio and TV (both push media), search is a pull medium; your ad only displays when someone demands it, so dayparting according to this logic should have no impact on either your clickthrough rate or conversion rate. To put it another way, you shouldn't care if only a fraction of your target audience is online between midnight and 6:00 am.
The best strategy for employing dayparting is to tie start/end times to key events, such as a flash sale. Flash sales are short-lived promotions, typically lasting less than a day. Marketers can use dayparting to run a multi-campaign test around the three stages of a flash sale:
1) Awareness Stage: In the run-up to any flash sale, retailers typically promote heavily. Pre-sale promotions typically run 24 hours per day and, as a result, don't employ dayparting. But marketers can promote the sale in more relevant ways by reacting to the timing of when visitors are browsing the site. For example, if your flash sale runs from 6pm to 10pm on Friday, then during these same hours on Thursday night, serve messaging that says, "Come back this time tomorrow for up to 75% off everything in stock."
2) Sale Stage: This stage refers to the hours of your actual flash sale. During this stage, you'll likely want site-wide messaging (and probably a different homepage hero) that mentions the ongoing sale, links to appropriate site sections, and automatically applies any coupon codes.
3) Post-Sale Stage: For prospects landing on your site after the flash sale has ended (for example, by clicking a link in an email you sent in Stage# 1 announcing the event), offer a simple message such as, "This sale has ended, but you can still find great deals throughout our site." If possible, link to a relevant section of the site such as a "Clearance" page. The post-sale stage typically involves a No Control test that runs indefinitely (i.e., without dayparting).
In the past, two factors have typically made flash sale promotions very difficult:
- Timing a test to specific start/end times was difficult.
- The short-lived nature of the flash sale made it hard to justify the cost of deploying a well-targeted test campaign.
A good tool and a smart strategy at your fingertips can make each of these factors moot. You can even up the ante with pre- and post-sale test campaigns to maximize visitor participation and minimize site abandonment.
In the next post of the "Strategies for Successful Testing" series, we'll tackle the most fundamental question of them all: Why test in the first place? And in the meantime, check out Peter Borden's awesome post on flash sale execution and workflows, with specific examples of how an apparel retailer leveraged one of the moment's hottest trends.