How to Grow Your Opt-In List More Quickly and Easily
This past fall, two marketers from Siemens presented exclusive results from their recent multivariate Website tests at MarketingSherpa’s Demand Generation Summit. They tested several types of Web pages to see what design changes would improve response rates, including corporate home page design, product info page photos and copy, and registration forms to gather opt-in e-mails.
The test the creative team was most excited about was the product page photos. They’d had lots of internal debates over whether the happy-people shots that the PR department adores would pull better or worse than technical product photos that turn on current customers.
Happy people won, much to the shock of the marketers, but the biggest shock was actually something completely different. Turns out that graphics, Web page design, and copy tests all trailed in the dust of the big, BIG test results. Yes, the tests the team ran on opt-in forms wound up providing a far higher response lift.
The big lesson: If you want to improve Website results, don’t waste time debating “creative.” Instead go for the high-impact stuff: tweaking the design of your opt-in form.
A tiny opt-in lift means your e-mail list grows bigger. And those names are better than any list you can rent or advertise on because they came from your Website, so they’re already self-selecting. More high-quality names on your e-mail list equals more responses to impress the chief marketing officer or the chief financial officer.
Inspired by this real-life story, I called up the folks at multivariate-testing firm Optimost to see whether its clients were testing registration and opt-in forms. Turns out it had run tests for dozens of clients over the past year. Based on those tests, MarketingSherpa’s research team worked with Optimost to develop a series of “rules” for opt-in forms that generally tended to lift or depress response. You can see an illustration of one such rule immediately below and a link to a report with the rest at the bottom of this column.
Source: Optimost for MarketingSherpa, November 2006
As you can see above, one of the rules was that vertical forms work the best. If you have to ask many questions, don’t give folks two columns of forms to fill out. (By the way, if you have a chart on your Website that uses columns for shipping vs. billing address, you may want to test changing that.)
The common key to all the lessons was that testing your opt-in form works gangbusters. So if you have the budget for a multivariate study, that’s a great page to focus on.
But what if you don’t have much in your testing budget? You can still tweak your form. At the very least, you can probably work with your Web design department (or e-mail service provider if they “power” your form) to try several versions, either as an A/B test if your servers can handle it or as a week-by-week test (one week is one form, the next a different version). You need at least 100 form fills per test cell for useful results.
Five suggestions of possible opt-in form tests worth your time:
1) Offer. Should you offer a discount coupon or a white paper to be e-mailed out to everyone who fills out the form? Don’t assume an added inducement to sign up is going to increase response rates. The Motley Fool told me it tested giving away an e-book with a free subscription vs. no giveaway. The no-offer test won.
2) Image of offer. This is often very powerful. Should you have a small thumbnail of what looks like a newsletter or a sales-alert issue? If it’s a white paper or an e-coupon, should you “show” an image of that?
3) Amount of information requested. You probably know that less info means you’ll get more form fills. On the other hand, some bits of data don’t make a big difference depending on how trusted your brand name is. For example, our own house tests have shown that asking for first name and last name in addition to e-mail address doesn’t hurt results. A bridal photographer told me his online e-mail lead generation form gets better fills when he adds a “your wedding date” field, even though that won’t affect what he sends the bride-to-be.
Phone numbers, however, always crush the reply rate.
4) Description of e-mail that will be sent. Don’t assume the terms “e-mail newsletter” and “news” and “get e-mail from us” are remotely enticing. You can name your regular e-mail program anything you want, but each name carries different shades of interest and meaning. Some examples:
* Hot sales alerts
* Monthly Journal
* Success Profile Series
* Private sale announcements
* Tips
* Video magazine
* Ten-step eCourse (value $29.95)
5) “Submit”-button copy. Definitely worth testing. Don’t assume “submit” or “subscribe” is the killer copy here. And do assume you have more space to play with than you may think. Buttons can widen to contain several words. Whether that helps response or not depends on the words themselves.
Anne Holland is president of MarketingSherpa, a research firm publishing buyer’s guides and benchmark data for its 237,000 marketing executive subscribers. For a copy of MarketingSherpa’s new E-mail Marketing Benchmark Guide for 2007, which includes more results from the Omniture opt-in form study, go to: go to: www.sherpastore.com/email-benchmark.html?8966.
© MarketingSherpa, Inc. 2007
Other articles by Anne Holland:
Revealed: Which E-mail Tests Work Best New Eye-Tracking Test Results: E-mail Campaign Click Patterns Surprise
TV Blitz Helps Ask.com Grow Market Share
New Data on Search Marketing Click Fraud: Three Action Items
Newest Eye-Tracking Study Results for Google
Absolutely Pitiful E-Commerce Shopping Cart Abandonment Stats…and Four Ways to Improve Yours
Study Data: Reasons to Get Evangelical About Evangelism Marketing
The Year’s Best and Worst Lead-Generation Offers
The Two Easiest (and Most Overlooked) Ways to Improve E-mail Response
1,120 Shoppers Say Why They Abandon E-commerce Sites
