Using Segmentation to Improve Site Conversions

• Tweet
• Buffer

Companies are trying to squeeze more out of their Web sites than ever before. We're willing to try new things that we thought we didn't have time for in the past. Web teams are more focused on the ROI (define) of Web investments.

Many of the measures to increase conversion rates require a significant amount of money. For instance, you might seek to increase traffic through paid media or search by 10 percent in an attempt to get more raw conversions on your site. But it's a tough time to get incremental funding for new approaches.

An often overlooked option that could have a significant impact on any Web site and isn't that complex (read expensive): segmenting your messaging.

Whenever we hear "segmentation," though, it sounds like a big, slow, time-consuming project. But it doesn't always need to be. There are some basic or easy ways to start.

What are the benefits of basic segmentation?

• Targeted, more applicable messages to site visitors.
  
  
• Anticipating the needs of visitors.
  
  
• Avoid marketing something to someone with zero interest.

Let's consider a financial institution's Web site with an account login. One basic way to segment your audience is to key off of login cookies. You can split your audience by those who do and don't have the online account login cookie. You can then tune the messages to each of those audiences. We could tune some of the promotional or call-to-action space on the site to promote different things to each audience:

• Visitors with an online account login cookie. If visitors have this cookie, we can assume they're an existing customer and have logged into the account information before. So we would promote the cross sale of new or complementary offerings or special promotions targeted toward existing customers (i.e. "Special Promo for Existing Customers...").
  
  
• Visitors without an online account login cookie. For this audience, we might want to promote "Open an Account" or "Benefits of Working With...".

More importantly, we would waste the space, time, and brain share by promoting things to these two audiences that don't make sense. We wouldn't want the number-one promotion to be open an account to someone who already has an account. We wouldn't want to promote (as our number-one promotion) change from paper statements to electronic statement.

These are the wrong messages for each audience. They convey irrelevant messages, prompting people to tune out because they aren't applicable.

In many cases, it's a matter of real estate. If we think about the home page being focused 50 percent on existing customers and 50 percent on prospects, we might want to swing this percentage based on what we know about our visitor (in this example, if they have the login cookie).

For the audience with the login cookie, we might swing the home page to being focused 80 percent on existing customers and 20 percent on prospects, allowing us to deliver a much more targeted experience. If they don't have the cookie, the balance would reverse to 20 percent on existing customers and 80 percent on prospects.

Remember, there's a margin of error in this approach. A prospect could use a public computer where an existing customer had previously logged into the same site, storing a cookie. Or, an existing customer could be using a different computer or has simply cleared the cookies.

You should still segment and deliver different experiences for both the new visitor without the cookie and the visitor with the cookie -- even though some percentage of each group could be miscategorized. Say, 2 percent of each group falls into the wrong category, you should still segment your message as it will better convert the 98 percent in each group. And, of course, you wouldn't hide the login to the account information for the people without the cookie or hide the new account information for those who don't have the cookie.

While this example revolved around a financial services company with an account login, you can segment your audience any number of ways, drop a cookie, and then key off of that cookie. While this isn't the ideal way to segment your audience, it's a great way to start. Later, you can move into more advanced segmentation, including the use of automated behavioral targeting tools like Omniture's Test & Target (the portion formerly known as Touch Clarity).

I can't talk about places to start maximizing conversion without touching on the importance of site optimization (A/B, multivariate testing, etc.). If you aren't doing this now, find a way to get started with a proof of concept or small basic tests.

Shoot me a note or comment below on how you're looking to maximize the value of your Web site in these tough economic times.