ABT: Always Be Testing Your Marketing Campaigns!
December 22, 2017
What if you found a slot machine that you won on all the time?
Would you play it a lot? All the time?
But maybe there are slot machines out there that would pay even better.
How would you find out? You would have to "test" different machines to see how they perform. And once you found a winner, you could keep playing it for a profit.
Sure, unfortunately slot machines don't work that way, but what if your marketing strategies did?
What if you found several key marketing strategies that (at least for now) consistently give you positive results and sales? Would you implement them often? Maybe even try to improve them?
The only way to find out if a marketing campaign works is to test the results.
As an example, say you have a certain product page on your website that converts at 5%. How could you improve on this %? By "split-testing" your product page against a second page to see which page performs the best. Split-testing is a testing method where you run two versions of your campaign to see which one out-performs the other. Whichever works better, you keep. And then you can try a new split-test with another variation.
With web sites, it's actually fun. Like putting two horses in a race. The winner goes on to compete against another horse (page) until you're happy with the results. However, you can keep testing as long as you're getting improvements.
The same tests can be applied to almost any type of marketing campaign from snail-mail to radio ads, to t.v. Here are a few keep points:
1) Make sure you have a system for tracking the exact results- whether they're sales, phone leads, contact forms, whichever metrics you deem important. Sometimes a separate email or phone number may be required so that you can keep track of which campaign is generating which leads or sales.
2) Only test one variable at a time. If you're split-testing, make sure you only change one variable at a time so you can be certain which variable is causing the change in results. Or, if for example, you're running radio ads, try to keep the changes obvious so that you can be certain these were the causes for any changes in results.
3) Make sure you have a large enough "sample" to ensure reliable results. I.e., you're going to need more that 20 or 30 page impressions to see which web page produces better. Whichever type of campaign (web, t.v., radio, mail, etc) research to find what size sample you should test prior to making any decisions.
Testing is sometimes difficult to implement (and learn how to do,) but once you get past the learning curve, it can be a blast! (And profitable:)
Thank you for reading,
Cheers,
Chris
Leave a comment
Comments will be approved before showing up.
