Don’t Test. Survey Instead.

Because of its incredibly low cost, email harbors a temptation that doesn’t exist with other media: Forget caution and prudent testing; toss as many emails against the wall as possible to see what sticks.

With the low cost of sending emails, people rationalize, saying, “What’s the harm (besides wasted time and consumer complaints) in just trying everything and hoping that, eventually, some of it will work?”

The ease, speed, and low cost of testing offers and copy on the Internet has many companies basing their marketing decisions solely on opinion, rather than on customer behavior, data profiling, and controlled testing. In other words, they’re approaching marketing as an art rather than a predictive science.

The reality is this haphazard process can actually cost lots of money.

Here’s a quick example: You want to test three different price points for a product — $11.95, $14.95, and $19.95. To do so, you opt to blast 1 million emails for each price point, resulting in 59, 32, and 20 sales, respectively. (It’s not surprising the lowest price point generated the greatest number of sales.) In dollars, that comes to $705.05, $478.40, and $399.00, respectively.

Here’s the kicker: Let’s say your all-in cost of goods to manufacture the product is $6. This means your gross margin at $19.95 is 70 percent, compared to 59 percent at $14.95, and 50 percent at $11.95. From this, you have to deduct the associated CPM and emailing cost to complete the analysis. Ultimately this will lead you to a volume versus margin decision and, in my example, the highest price point will probably win out. Unfortunately, by this time, you’ve burned 2 million email addresses to come to this conclusion.

So how do you determine which offer to roll with or what copy points to emphasize in your email advertising, to produce the highest and most margin-efficient conversion rates? Is it experience gained from other channels, such as direct mail? Do you conduct customer focus sessions? Do you count on creative suggestions from your marketing organization? Or do you simply rely on gut instinct and common sense?

The answer is predictive modeling.

Ironically, because of the ease of using email and the speed by which offers can be distributed and evaluated, predictive modeling is significantly less expensive than in other media channels, such as direct mail.

One of the easiest, most powerful, and valid techniques to help predict responses and conversion rates with return on investment (ROI) metrics is the email survey. Surveys can be quickly and inexpensively constructed to test pricing options, copy, and offers. They can be sent to your existing customer base or to prospects to confirm or refute assumptions.

And the size of the survey pool doesn’t have to be massive to obtain statistically relevant information, either. Just 100 valid responses (not numbers sent, but actual valid responses) from a good list indicative of your customers will provide important data to help your marketing and creative team improve their strategic decision process and, ultimately, the success of your email campaign. That’s right, only 100 responses are required. Think about it this way: In a national election, only 700 people are surveyed to predict the outcome.

When constructing a survey, try to limit it to a few important questions. For example, we recently sent an email survey to find out if more consumers were apt to buy a single product at the lower price of $19.95 versus paying $29.95 for two products. Much to our surprise, 70 percent of the survey respondents selected the $29.95 deal. This shaped our creative and was the offer we ultimately rolled out.

Remember, though, that whenever you implement a survey, you should always tabulate the data and restate it in comparative percentages to provide an easy method to analyze the finding. Compare apples to apples.

Sometimes, to engage consumers to participate and respond to surveys, we find it necessary to incorporate involvement devices such as contests, sweepstakes, or inexpensive gifts. Where we’ve used these devices, we’ve kept our total costs under $3,000 while still producing tremendous participation. Clearly, surveys incorporating involvement devices are an extremely effective way to quickly engage tens of thousands of prospects in, literally, 24 hours.

Think of how inexpensive this is in comparison to the risks of rolling out an advertising campaign with untested creative. Plus, it’s an opportunity to learn how people think and behave by the comments they make, because surveys can be used not only as a way of gaining critical insights and tactical knowledge to help improve creative but also as hybrid marketing tactics.

Survey respondents who identify their interests and behaviors are often very willing to consider marketing opportunities matched to their needs and wants. Sometimes we determine this by ending a survey with an opt-in question asking whether they would like to see an opportunity that addresses their needs. We’ve seen cases where more than 50 percent of the prospects have responded affirmatively.

As a direct marketer who has worked extensively in every other media channel, I’ve learned using email surveys allows you to quickly and inexpensively reach tremendous numbers of prospects to gain instant feedback, test creative assumptions, reduce risks, and increase revenue.

Try it yourself before launching your next email campaign, and don’t be surprised if you see results improve by at least 10 percent. Keep reading.

 

ClickZ Marketing Excellence Awards 2003: Who Won?