Testing is one of the most-preached and least-practiced activities in marketing. What is your excuse for not testing?
The only marketers who actually use testing on a consistent basis are those who depend on direct response (including advertising, direct marketing, online marketing) to stay in business.
These are catalog companies, mail-order and online marketing companies, publishers, membership organizations and non-profits.
For them, direct response (regardless of media) isn’t just another advertising channel – it’s the lifeblood of their business. And they’ve seen firsthand how important testing is to their success.
Testing isn’t a high priority for most
But for many more businesses, advertising, direct marketing and online marketing are just part of the marketing mix and not seen by management or marketing as absolutely critical to the success of the business.
For those businesses, testing is very low on the priority list.
They may see advertising, direct marketing and online marketing as useful, maybe even important, parts of their marketing program, but they don’t see the value of going beyond the basic applications.
Much of this is because they don’t understand what testing is (and is not), how much time or money is involved or how to do it correctly.
Regardless of the reason, here are some of the excuses we hear:
1. We need to get moving. We don’t have time to test.
Many people believe that testing is something you do before you actually begin your campaign. This could be true when you’re just starting out (much like market research), but testing is more of an ongoing process. It’s a process that is built into every marketing campaign in an effort to continually improve performance.
When you are planning an entirely new campaign, you have no baseline of results. So you will probably need to invest a larger percentage of your budget (perhaps your entire budget) in testing different elements.
For example, you might need to test several media options (for advertising), several lists (for direct mail), several offers, several messages and several landing page/response channel options. You might even need to test your sales process.
This doesn’t mean your marketing campaign is on hold. Your tests will produce results (leads, traffic and customers) just like any campaign. The only difference is that you will have structured your campaign to produce comparative results. That’s testing.
Over time, you will know what’s working for you, but you should continue to use a smaller percentage of your budget to test improvements.
2. We don’t have a budget for testing
I hear this all the time, but testing doesn’t have to be a budget breaker.
Mailing lists and media options, for example, can be tested with no increase in your budget.
Consider this: Instead of using one list of say 20,000 names, why not get four lists of 5,000 each? The cost differential is negligible, but you will come away with comparative results on four different lists.
Instead of running one big ad in one publication, try a smaller ad in four publications. It’s not exactly the same cost, but it’s close.
It’s true that if you decide to test offers, headlines, graphics and different creative approaches, that may require additional creative and production costs.
Landing pages will cost extra to create, but you can run alternating landing pages for virtually no additional cost..
3. We know what works for us – we don’t need to test
Maybe you do – but how will you know for sure?
No one can be so arrogant to assume they know everything about their advertising. Too often testing produces results no one could have intuitively expected.
When it comes to testing, personal opinions and intuition don’t count. The only thing that matters is the numbers. And the numbers may surprise you.
4. We are happy with our results
It’s great when you can make your marketing results work for you … when you can deliver a sufficient number of leads to your sales team with your current budget. In that case, you should be happy. But could you do better?
- What if you could increase your lead output by 10% without increasing your budget.
- What if you could reduce your cost per lead by 20%?
- What if you could increase your lead conversion rate by 30%?
Think about what those changes could do for your bottom line. Wouldn’t it be worth testing?
5. We don’t know what to test
This is often the situation but few will admit it.
Your best bet, at least to start, is to test the big stuff – things that will have the greatest impact on your response.
And the big stuff would be (1) your list (or media) and (2) your offer. Everything else is secondary.
Only after you have settled on your most productive lists/media and offers do you start testing things like format, headlines, message, graphics and other smaller elements.
6. We don’t know how to test
Nothing wrong with that. If you haven’t done it, how would you know?
Here are the basics: First, understand that when you test, you are usually testing against a control. A control is your current winning component – whether it’s your best list, your best offer, your best creative.
When you test, you isolate the element you want to test while keeping the rest of the campaign exactly the same as the control. This allows you to see how that one element performs in comparison with the control.
In the end, after you test, you will come with results that are either better or worse than the control’s results. If the results are better, you have a new control.
Many people will compare the results one campaign against another and call it a test. Yes, one campaign will have better results, but what part of that campaign made the difference? Was it the list, the offer, the format, the message, the graphic? You don’t get those answers with this type of broad-based test.
7. Testing takes too long
It doesn’t have to take that long. Sure, it’s something else to think about – and your time is already stretched pretty thin.
But if you consistently add a new testing element into every one of your campaigns, you come to expect it. Once the campaign is executed, there is no extra time involved. You just track the results and see how your test compares with your control.
8. We don’t have enough quantity to test
That can be a problem – but not an insurmountable problem.
Traditional testing amounts for direct mail are 5,000 names for lead generation and 10,000 names for order generation. This is the kind of quantity you need to have a statistical confidence in your results.
But for some companies, their entire universe may be just a few thousand – or even a few hundred – prospects. In this case, you’re not going to have sufficient quantities to conduct a valid test. But you can look for indicators.
If your list is 2,000, split it in half and test two offers. A test cell of just 1,000 names isn’t going to produce a statistically valid result, but it might give you some direction, some idea of which offer is more popular. Next time out, switch your offers so everyone gets a new offer and see if the first winning offer continues to pull better. This may take a while, but it could be well worth the effort.
9. The results will be insignificant
Sometimes you test and you don’t see any definitive results. The response rates or cost per response is so close that there are no clear winners.
From a testing standpoint, this can be frustrating.
Of course, if the results are good across the board, this is good news knowing you have many options ahead of you. If the results are poor across the board, it may be telling you something you didn’t want to hear.
Keep testing though. You may come up with a real breakthrough winner.
If you’re not testing, you can take comfort in knowing you have a lot of company. If you are testing, or planning to test, you may just be able to break out from the pack and gain the marketing advantage we all want.
Written by Bob McCarthy
This article may be reprinted without permission as long as the article includes the following credit:
Bob McCarthy is a direct response consultant helping businesses develop measurable marketing programs across multiple marketing channels. You can read his blog, The Direct Response Coach, on his website at www.mccarthyandking.com. Bob can be reached at 508-473-8643 or at firstname.lastname@example.org