| If you can measure it, you can improve
| |
| | effective, it also has drawbacks. Number
|
| it. That is the maxim of online testing.
| |
| | one is that you need thousands of
|
| Those who test their advertising
| |
| | impressions and hundreds of conversions
|
| campaigns can see ROI double, or even
| |
| | or click-through actions in order to
|
| triple. Testing will increase conversion
| |
| | achieve statistical confidence. The
|
| rates. There is no such thing as "good
| |
| | other setback is the ability to test only
|
| enough." You should always be testing and
| |
| | one element at a time. Testing more than
|
| improving your campaigns.
| |
| | one element can take a lot of time and
|
| Every element of an online marketing
| |
| | traffic.
|
| campaign is measurable. Test your PPC
| |
| |
|
| ads, landing pages, and email campaigns.
| |
| | Multivariate Testing
|
| Testing results in actionable data that
| |
| | Multivariate testing is the testing of
|
| will show up in your bottom line. Split
| |
| | more than one element or variable at a
|
| testing and multivariate testing are the
| |
| | time. This form of testing allows you to
|
| two main forms online testing will take.
| |
| | test several elements at once and find
|
|
| |
| | the individual elements that bring the
|
| Split Testing
| |
| | best results, as well as the combination
|
| Split Testing, also known as A/B testing,
| |
| | of elements that will give you the best
|
| is the simplest form of testing. Google
| |
| | possible outcome. It allows you to find
|
| AdWords, Yahoo Search Marketing, and most
| |
| | which elements matter and which
|
| providers of email marketing services
| |
| | combinations are strongest in a
|
| have split testing capabilities built
| |
| | relatively short amount of time.
|
| right into their campaign interfaces.
| |
| | Multivariate testing can give you the
|
| Split Testing is randomly dividing
| |
| | best bang for your buck, and the testing
|
| visitors into two groups and showing each
| |
| | algorithm automatically determines the
|
| a different version of your PPC ad,
| |
| | best page design for all possible
|
| landing page, or email. The objective is
| |
| | combinations of elements.
|
| to see which version performs better in
| |
| | Testing at this level requires
|
| terms of conversion, CTR
| |
| | sophisticated statistical analysis.
|
| (click-through-rate), lead generation, or
| |
| | Unless you are a coder with and advanced
|
| some other target.
| |
| | math degree from MIT, you are better off
|
| Email campaigns are easily tested. Send
| |
| | using an off the shelf application or
|
| one version of email to half the leads on
| |
| | provider. Most providers of multivariate
|
| your list, and another version to the
| |
| | testing are proud of their work and cater
|
| other half. Use a subset of your list
| |
| | to enterprises with deep marketing
|
| for testing, and then apply the results
| |
| | pockets. The exception is the Google
|
| to the rest of the list. You can use
| |
| | Website Optimizer. It provides all the
|
| split testing in this way to find the
| |
| | power of these other applications but
|
| best layout, ad copy, and headlines for
| |
| | beats them on price. It's free.
|
| you email.
| |
| |
|
|
| |
| | Conclusion and Resources
|
| Split Testing an AdWords Campaign
| |
| | Testing takes the guesswork out of
|
| Google AdWords makes split testing of ad
| |
| | marketing campaigns and allows you to
|
| copy easy. Simply create two different
| |
| | make continual refinements that will
|
| ads to driving traffic to the same page
| |
| | increase your ROI. Implement a
|
| and see which one gets a higher CTR. It
| |
| | thoughtful testing program and you'll be
|
| is important to turn off the optimize
| |
| | amazed at what works and how little
|
| feature in the AdWords campaign settings
| |
| | things can make a huge difference. But
|
| so that both ads get an equal number of
| |
| | most of all you'll be amazed at the
|
| impressions.
| |
| | difference it can make in your bottom
|
| Another way of to AB test with AdWords is
| |
| | line.
|
| to have the same PPC ad go to two
| |
| | See my article on increasing conversion
|
| different landing pages. This allows you
| |
| | rates with Google Website Optimizer for
|
| to split test elements in the landing
| |
| | more testing tips. GrokDotCom has a
|
| page for conversion or some other
| |
| | wealth of information and resources on
|
| performance indicator. Make sure the two
| |
| | the topic of online testing. Be sure to
|
| pages are significantly different if you
| |
| | see their 7 free Google Website Optimizer
|
| use this method. Search engines frown on
| |
| | resources. Another good testing resource
|
| duplicate data and may penalize you for
| |
| | is Marketing Experiments. There you will
|
| it.
| |
| | find online testing case studies and
|
| Although split testing is simple and
| |
| | research.
|
