 Ad scheduling is a relatively easy path to profit improvement because it’s all about listening to your customers. In fact, it helps you increase your search marketing ROI without requiring you to spend time creating new ads and landing pages, finding new keywords, or executing fancy bid rules.
Ad scheduling refers to weekparting, or varying bids by day of week, plus dayparting, or varying bids by hour of day. Varying a bid can mean turning it off, or bidding more than normal, or less than normal.
Ad scheduling is not very complicated. It's key techniques are the same basic analytic techniques that you use to improve any of your online marketing. The data is there - if you don’t use it, your competitors will, and then they’ll eat your lunch.
Analytics Techniques for Ad Scheduling
-
Value not volume. Just because there is more traffic at a certain time of the week doesn’t mean you should bid higher during those times to get a higher share of it. It’s about profit, not clicks.
-
Cookie / Attribution Windows. Just because traffic does not convert in the same session it arrived does not mean it is worthless traffic. You need an analytics solution that matches conversions back to a click on an ad that drew the visitor to your site many days ago. How many days depends on your business, 30 is a standard number to use but longer can help for certain types of businesses.
-
Non-Brand Versus Brand. Again: just because traffic does not convert in the same session it arrived does not mean it is worthless traffic. You need non-brand traffic to get people interested enough in you to type in a brand search keyword and become brand traffic. Track whether your non-brand visitors eventually convert via sessions initiated by a brand search keyword, and make sure you allocate budget to that non-brand keyword traffic that evolves into customers.
-
Time Zones. This is really basic so I stuck it in the middle. I mention it though because you should just confirm what time zone you are using to set rules.
-
Data Recency and Seasonality. Behavior patterns change so weighing more recent data over older data makes sense. However, keep in mind that certain businesses have highly seasonal affects where data from a couple months ago is less valuable than data from one year ago. For example, retail businesses sell a lot during the end of year holiday season.
-
Forest for the Trees. You can waste time analyzing very small windows of time.
-
Data Volume Required to Make Decisions. You can’t make decisions based on just a couple clicks, a traffic source needs time to prove itself - without busting your budget. As always, the trick is knowing exactly how much data you need before pulling the trigger. With conversion rates in the 3-5% range for search keywords, you need to wait for 30,40, even 50 clicks to see results.
-
Test. Ah, save the best for last. Testing is what brings profits in all online marketing. First, crawl before you run. Don’t spend a lot of money at first, spend a little bit of money on a number of different things to see what works. Second, be methodical. Allocate more budget to what works.
One last tip: make sure you understand the delay, if any, between when you request or schedule a bid change and when it gets executed.
###
Related Resources:
|
