The Lunchtime Effect

In radio, it’s called drive time, in TV land it’s called prime time, and in the wonderful world of search engine marketing (SEM), it’s called lunchtime. Dayparts are a way of segmenting media by the time of day, because viewer receptiveness to messages and readiness to take action varies based on the time of day. Even the retail business has dayparts — times of the day when customers are more valuable and more likely to buy.

It should come as little surprise, then, that certain search engine inventory can be bought by daypart. Dayparts are one way of selecting out the premium inventory and going beyond the average — potentially doubling your effective SEM marketing budget.

A significant percentage of SEM inventory is sold in auction-style environments. The auction environment may sometimes make your blood boil; however, there are positive aspects to it. The real-time nature of Overture, Google, and other pay-per-click (PPC) auction engines provides the opportunity for savvy marketers to take advantage of the differences in searcher behavior at different times of day, by measuring post-click results.

Of course, the Internet is a national and global medium, and therefore times of day are relative — unlike radio or TV dayparting where the exact time of day the audiences is experiencing can be determined. Online marketers don’t need to know exactly from where each searcher is originating, and even if they did, they can’t target by the click. All marketers really needs to know is their conversion results and how the results vary by time of day.

Questions to ask:

• When are the most orders coming in?
• When are the most profitable orders coming in?
• When are conversion rates the best?
• Are these patterns repeating day after day, week after week?

Although instituting a feedback loop, as I wrote about before, will likely add significant efficiency to your campaign, you can take data analysis even further with dayparting. But you must understand how dayparts work for your site. Not every site can implement the same daypart strategy, because each site has a different target market and a different buying cycle. Business-to-business (B2B) marketers and business-to-consumer (B2C) marketers will have different daypart optimization opportunities. The B2B marketer may also have significant weekday-versus-weekend shifts in value.

Let’s assume after gaining an understanding of the relative value of your keywords/listings in the various engines, you have your campaign humming along. With a feedback loop in place, your listings may even be adjusting based on time of day automatically if the volume of clicks in an hour is high. However, your heat-seeking missile tactic can be even more powerful if it knows where the target is going to be before it gets there.

Daypart analysis will give you a predictive model that will allow you to modify your campaign for maximum daypart-derived power. You do this by collecting and analyzing data about conversion percentages. The most difficult part is deciding how granular you want to go. Most marketers will derive a large percentage of the daypart efficiency by looking at daypart conversion data on the aggregate (mixing keywords, engines, and products together).

As an example, let’s imagine a catalog marketer on the U.S.’s East Coast that ships throughout the U.S. We’ll call the company Yuppiemart, and we will look at the segment of its campaign where the target cost per order (CPO) is $35. Much of the online shopping among the U.S. population goes on during lunch, and, for Yuppiemart, the target market is likely to be employed. The Yuppiemart marketing VP is looking to optimize the campaign by conversion to sale, keeping the CPO under $35. (It would be possible to factor in catalog requests or phone inquiries from prospects not quire ready to buy, but let’s take one optimization example at a time.)

Yuppiemart’s conversion from click-through to sale is 4.5 percent from 11:30 a.m. to 5:30 p.m., then 2.5 percent from 5:30 p.m. to 1:30 a.m., and 2.0 percent from 1:30 a.m. to 7:30 p.m., then back to 2.5 percent till the lunch crowd hits again. If you were to look at premium Overture listings over a 24-hour period, you would probably not see many marketers changing their listings’ positions significantly over the course of the day to reflect a difference in conversion behavior. Perhaps the high bids and positions you see the advertiser using conform with their CPO targets for the full 24-hour period, but I doubt it.

For Yuppiemart, the average conversion rate over the full 24-hour period is 2.875 percent. That would mean for its most important term, “izod shirt,” the company can only afford $1 per click. Thus, Yuppiemart may be stuck in position four. This scenario results in two campaign inefficiencies:

• During the peak in traffic, and when the quality of the searcher is greatest, Yuppiemart loses a huge opportunity. If a competitor has search position number two, the searcher may find an appropriate product at that company’s Web site. The tragic part is for that six-hour lunchtime span, Yuppiemart could make money paying for position number two for “izod shirt.”
• Before and after lunch, Yuppiemart may make money in position five, given the 2.5 percent conversion rate. Then overnight, the only way that Yuppiemart can stay on target with its CPO is in position six, where the CPC is $0.70.

Another alternative would be to simply focus on the lunchtime period exclusively, but then order opportunities during other times of the day may be missed. In addition to implementing daypart campaign optimization, Yuppiemart may find that analyzing the days of the week will also uncover patterns of conversion that can be used to optimize.

Many SEM tools have the capability to assist you in managing by daypart or day of the week. If you have an agency manage your paid SEM campaigns, ask your contacts if daypart optimization is available.

The larger your media spend, the faster you will see a positive return on investment (ROI) on the time and energy spent optimizing by daypart. Is this kind of analysis easy? No. You need the right tools and right strategic mindset. Is it worth it? Can you afford not to find out?