A question from someone who was kind enough to read my latest Deliver column, followed by my answer:

Q: Is an order form equally important when it’s a nonprofit arts group selling tickets? I always push for including a user-friendly order form or invoice for subscription brochures, but am getting increasing pushback.  Clients feel they’re unnecessary, take up space, people go on the internet anyway, so why bother? I’m assuming we should include them for the same reasons you cite for reply cards. I don’t have pushback about using reply cards when requesting a donation, but subscriptions are much more complicated so order forms take up more space.

A: The most responsible answer to any "what works best" question is, of course, test. That said, my own experience is that you should ABSOLUTELY include a reply card, invoice or order form (I'll just say "reply card" from here), even when people can respond online. My reasons:

• Even the technologically adept follow the course of least resistance. If checking YES on a card and dropping it in the mail is faster and easier than logging on, people will do it. We recently mailed a subscription offer for an email newsletter. Online sign-up was easy, yet half of the response came from people who checked YES on the reply card and snail-mailed it.

• The reply card does more than enable snail-mail and phone replies. It is often the first looked-at piece. A good one impels readers into the other materials. They may reply online, but the reply card set the sale in motion. The trouble is, if that's happening, it's invisible to you and the client. Only split testing will tease it out.

• People often discard the rest of the mailer and hang onto the reply card to revisit later. Omitting the card takes that option away, possible at the cost of sales. Again, the card may drive them to the web. But with no card to hang onto, they may do nothing at all.

Some quick notes on split tests (at the risk of repeating what you already know):

To split-test your mailings, send half of your list a package with the card and half without, and see if there is a difference in response over time and over multiple tests. Over time and over multiple tests, because flukes happen. Allow time for all responses to drag in, and be wary of drawing a firm conclusion from one or even two tests. I'd also be wary of generalizing for all clients based on test  results for one. And, be wary of carving test results in concrete. I'd retest from time to time, as behaviors sometimes change.
 


Comments


Your comment will be posted after it is approved.


Leave a Reply