You Don't Have to Be Amazon.com to Target Your Emails

To take best advantage of the web, companies must proactively engage and motivate web site visitors. We can now segment web visitors using their expressed interests, behavioral patterns, past purchases, responses to promotional campaigns and more. This level of effort is no longer exclusively for the likes of Amazon.com. If you're not tracking your email marketing to this degree, it's time to start planning for it. You can be sure your competitors are.

All varieties of companies are finding email to be a significant driver of sales, retention and loyalty. Consumer marketers like Scotts lawn care are sending out different newsletters every day to different customers about different lawn care issues in different parts of the country. Business-to-business companies are keeping their clients up to date on industry happenings. This is a branding, retention and loyalty tool, whether you're 100 percent web-based or run a full brick and mortar company.

Being able to calculate how many emails went out, how many were opened and how many people clicked through to the site makes email one of the most measurable promotional and relationship building tools. Personalized, database-driven newsletters can engage prospects through the education and persuasion cycle. Email scenarios (multi-wave campaigns) can automatically and uniquely respond to web events such as white paper downloads and replies to previous email messages

Downloading a file from a web site is one example of an email scenario trigger. An outbound email offer can kick off a series of targeted messages as well. Recipients who click on specific product links contained in an initial email (without making a purchase) automatically receive a second-wave promotion offering added incentive to purchase the particular product they clicked on. This makes the email much more relevant than it would if you simply sent them the same generic email every customer receives.

Email segmentation
Here's an example of how this works.

Once in the opt-in database, visitor behavior can trigger a wide variety of communications:

IF	THEN
A visitor responds to an email offer but does not pursue additional information that day.	Wait two days, then send a survey asking whether he or she was looking for something other than what was offered.
A visitor responds to an email offer, puts the item in his or her shopping cart but abandons the purchase process.	Offer a further incentive to buy today (i.e. free shipping).
A visitor responds to an email offer, puts the item in the shopping cart and makes the purchase.	Send a follow-up cross-sell incentive offer the next day.
A visitor forwards your offer to a friend.	Send a thank-you offer to the customer.

In the case of an email program, the results are assessed beyond simple clickthrough or simple conversion (Figure 1).

Figure 1


The results above show that offer B did very well across two customer segments, but offer A was a major disappointment. With this much data in hand, another round of testing might reveal that B is not necessarily the right message for all customers. The next step is use those results to improve the next round of customer contact (Figure 2).

Figure 2


Offer B is a hit with segments 3, 4 and 6. However, segments 1 and 5 are just not buying it and segment 2 is a complete loss.

Sending out alternate emails is a powerful test of messages.

The Scotts Co. (lawn care) sends email every day depending on recency of purchase, product purchased, geographical location and even the weather. If you grow Kentucky blue grass, Scotts will send you discounts on Kentucky blue grass seed or fertilizer formulated for Kentucky blue grass. If its telephone call center notes a rash of grub infestation in your area, you'll get an email giving you the heads up, telling you how to check your lawn and, of course, advising you which of its products can solve the problem. Scotts' customers are far less inclined to delete Scotts' email, because they know there will be relevant information in each one.

Once a visitor arrives at your web site, you can put the same testing concept into play. Which landing page has the most impact? Which persuasion path is the most powerful? Which sequence of information is the easiest for the visitor to consume?

Results from one round of trials can be used to revise offers in a continuous feedback loop. Multiple landing pages can reveal which combination of message, graphic and offer is the most powerful. In the following example, (Figure 3), Landing Page A had a significant impact in all segments on shortening the number of clicks from first response to final action.

Figure 3


Another landing page is tested to determine the content that will entice customer segments 2 and 5 to buy (Figure 4).

Figure 4


Web analytics is teaching us that your mileage may vary. The gold is not found in the lessons learned over time and the wisdom of our elders. The marketplace is changing too fast for that. Instead, we're learning that, as the people at Amazon like to say, "data trumps intuition." Creating a culture of experimentation is key. Test and measure. Test and measure. And don't stop.

All of the companies I've spoken to and worked with have found a significant lift in their email marketing when they segment, target and personalize. Successful email marketing means delivering email campaigns that are more targeted, content that is more personalized and upsell/cross-sell offers that are more relevant. How will you do it tomorrow? You will follow the lead of those on the front lines. You will adjust your messaging based on each individual's responses, historical clickstream data, purchase behavior and real-time in-site activities.

The tools exist, and the methodology is sound. The ability to raise revenue, lower costs and increase customer satisfaction are in hand. Now it's just a matter of trying it. Those who do will gain an enormous competitive advantage over those who do not.