• by Mark Klein , January 29, 2014

I recently rewatched Nora Ephron’s 1998 romantic comedy, “You’ve Got Mail,” and I was struck by how much has changed since the movie’s release. Back then, receiving an email was a big deal, and the AOL chime notice was a recognized part of pop culture.

Today’s reality is that you’ve got mail, and mail and mail: so much of it that managing email can steal hours from your workday. Email is a victim of its own success. It’s so inexpensive and effective that many companies send their potential customers a daily dose, often more. Response rates and click-through percentages can be less than one percent and still be profitable. As long as it’s profitable, companies keep cranking it out.

We do have laws that enable recipients to opt out of the torrent, but those who do are a small subset of those who continue to allow it. For most companies, the dropouts are merely collateral damage that can be tolerated as long as response rates are high enough to maintain some profitability for the campaign.

Companies that send excessive email don’t realize the financial penalties associated with opt-outs. When a customer opts out, the blocked company is deprived of a powerful channel to reach that customer and consequently loses significant revenue opportunities. We know of companies where emailable customers spend almost 50% more per year than customers who will only accept postal communications.

The way through the dilemma is to make email more relevant and thus more attractive and useful for the recipient. It’s now clear that relevance raises response rates (and thus revenue). The increase is dramatic. When a company sends out a million emails and each email is unique because it has specific product offers tailored to the recipient, we know it will be attractive to the recipient. And it’s guaranteed to raise response rates.

True one-to-one marketing has been the holy grail of marketers for several years. “Right offer to the right person at the right time” has become a way of covering up segmentation campaigns based on demographic data or logistic regression. Perhaps five or six different emails are sent, not five or six million. Personalization with “Dear Bill” or “Dear Mary” is not true one-to-one, not true individualized marketing. Customers have different needs, so what might have been the right offer or time for one customer was likely the wrong offer or wrong time for another.

The fundamental problem is that almost all customer analytics to date have been done at the segment level. That is, groups of customers are analyzed to predict what all members of the group might buy. The flawed assumption is that the group is homogeneous. But just because you have the same zip code as your neighbor doesn’t mean your interests are the same. Just because you are in a group that previously purchased a TV doesn’t mean your next purchase will be the same as others in your group who purchased a TV. To predict what an individual customer is likely to buy, you need to do the customer analytics at the individual customer level. You need to predict the likelihood of each customer buying each product or product category, even when there are millions of customers and tens of thousands of SKUs.

Three changes are now making relevant, individualized marketing a realistic option. First, companies are capturing the necessary transaction data. Second, new methodologies are doing customer analytics at the customer level, not the segment level. Third, more printers and ESPs can now handle variable data, so everyone can get that unique piece of marketing communications.