5 Ways to Test B-to-B Offers

Apr 05, 2012 6:13 PM  By


Testing one offer against another, one list against another, one price point against another: It could be argued that testing is the number-one job for successful multichannel marketers. Like many other critical marketing functions, testing in business-to-business marketing presents its own unique challenges.

Testing addresses the need for continual improvement in your marketing efforts. As long as you are not receiving a 100% response rate, you suspect that you could test to find a way to improve response or lower costs or increase sales, etc., whatever your company’s goals are. Even if you had a 100% response rate, new technologies and products evolve every day that you want to try (test) before you commit to investing large sums of money into them.

While the potential exists to improve, many tests actually cost money because they do not even replace the sales of the traditional method you are using, leave alone improving them. So the question is: how can you minimize the risk and maximize the success of your testing efforts, especially in business-to-business marketing? Here are five tips to get you started…

Test one thing at a time


Sometimes business-to-business marketing efforts are small or infrequent, so the temptation exists to try to test a large number of variables at one time.

The key to testing is to control all the variables so the only thing that can be producing the result is the one, single variable you are testing.

For example, in a test of 50% off a product with a full year warranty against 25% off the same product with a six month warranty, was the winner decided by the length of the warranty or the amount of the discount?

Test a statistically significant number


Some audiences are so small, especially in business-to-business marketing, that whatever results you get are not strong enough to project them onto a larger effort.

Many of us learned this lesson in the world of mailing and emailing lists. We would test 5,000 names or email addresses, and because our products had a high price point, we would get a small response: four purchases on one list vs. six on the other.

Is this difference large enough to give you confidence to go out and buy 100,000 names from the list with six purchases and none from the list with four?

Re-test to confirm findings


So, what’s a business-to-business marketer with high price products but a small response to do? Well, you can re-test to confirm findings.

No, you cannot conduct test panel A in March (e.g. 50% off) versus test panel B (e.g. 25% off) in June and compare the results as if they were conducted at the same time. Tests need to be conducted simultaneously to lessen the impact of time/season as a variable on response.

However, you can conduct the same test, A vs. B, both at the same time and twice – once in March and once in June. If you find the results to be consistent, they confirm each other and your confidence in rolling out to a large effort can be more justifiable.

Size does matter


Likewise, you need to make very sure that the groups you expose the tests to are the same. You cannot send one offer to large companies and another offer to small companies, and then compare the results to decide which offer to use for all sizes of companies.

The results may be different, not just based on the offer, but on the size and differing needs of the companies. When you test something, it is critical that you identify a larger group/segment that will be the same in composition.

Often, a test that worked on a segment like small businesses will prove successful. Then, you apply it to large businesses, and it fails miserably.

Read results properly


If a test requires the marketer to spend more money, any increase achieved by the test in sales or profits should be viewed incrementally, e.g. how much more did you spend than you would have on the control and how much more did you get than you would have received?

Remember the control is the promotion you have been using traditionally to generate sales and profits. Also, test only what can really matter, what your company is willing to sustain over time.

If your operations department is categorically unwilling to give away free shipping, what does it matter if you perform a test and it wins?

Mary Ann Kleinfelter is president and owner of marketing consultancy Marketing Solutions Today.