• by editor , David Baker , January 29, 2007

Anyone who has worked with me knows I'm a huge fan of balancing observation with what we already know works. Here is a case in point.

On the surface of it, measuring the value of email marketing is simple: 1,000 emails go out, 995 were delivered, 300 unique people viewed the message in HTML, 100 clicked on one of more links, and 30 purchased something. Anyone can make sense of this logic. Thus, based on a causal relationship to direct response, anyone can demonstrate the value of the email channel to a business.

However, few people in our space have found a formula that represents the true or implied value of an email address to a business. This is why many will ask, "how should I build an email database?" I've written a few articles on this concept of the value of an email address and talked with a great many others who struggle with this type of valuation. We developed a survey at the Email Experience Council on this exact subject. I even spoke on this at the Email Insider Summit. At this point, there are only a few instances where the email channel can be isolated and a valuation can be put on the channel. One is an ROI calculation, and another is creating a proxy to understand the incremental value to a customer.

If I asked you to scrap everything you are doing with email and start fresh, how would you budget for your channel? How would you justify it when you have NO email addresses? How would you dissect your media budgets to understand their influence on sales and the relationship value email will bring to that? I know many of you would struggle with this, especially if there isn't a causal transactional event. Fortunately, I can help you make some useful calculations. Here are some areas to consider:

1. Wherever there is a finite comparison of cost savings (efficiencies gained through the use of the channel). An easy area to isolate is where you are replacing print or fulfillment costs with the distribution costs of email. This is a direct cost savings, yet also has a value that isn't as finite as costs -- the value of efficiencies gained in being real-time in the fulfillment function -- good customer experience.

2. Wherever there is a finite transactional event tied to email. This is the easiest to calculate, as you have a direct-response link to conversion.. You know which emails went out to whom, and whether they bought something. The challenge is, attribution modeling. What credit do media, search and the site get in this conversion? Should it all go to the email channel? For instance: a consumer receives an email and goes to the site but doesn't buy, but then goes back the next day and buys. How do you attribute the revenue? Let's say you track the customer's activity through site-side cookies, so you know he/she was exposed to two ads in that time span and four support pages on the site, plus he opened and clicked through from the email and potentially accessed search and came to another part of the site in that time. Who gets credit for the sale, and what attribution should be given to the media's influence and the site's influence on the sale?

3. Taking this a step further is creating a proxy that is an extension of lifetime value. So you are predicting the future profit value of a consumer based on today's dollars. Consumer would each seem to have a similar value based on their affinity to purchase, how often and at what discounts, right? But do they? What if they are engaged with you through many channels, such as Web, media, search, email, on-premise, and catalog? The challenge is to understand how the consumer is influenced by each channel and to create a proxy that helps you understand which channels offer the most influence and efficiency in creating and sustaining loyal customers. This can get quite difficult to calculate, since an average customer email file loses 2.4% of its database each month.

To read the rest of this article and expanded thoughts, including ideas for how to get this rolling, go to my new blog at http://whitenoiseinc.com .