Essential Guide

Browse Sections


This content is part of the Essential Guide: Trend watch: Data management and business intelligence technologies
Get started Bring yourself up to speed with our introductory content.

Don't misfire on your targeted-marketing campaign

Targeted-marketing campaigns don't always find their mark. When they go astray, the results can be ugly.

The flurry of catalogs started falling in late October. It hasn't let up. They've been accumulating at a pace of about three or four per week. Their products cover the spectrum from religious items, like T-shirts with Bible verses written across the chest, to bizarre kitsch, including an entire page of squirrel-themed paraphernalia.

Somehow a company decided that I am the type of person who buys holiday gifts from catalogs and has responded by sending me an endless hail of them featuring items I would never buy for myself or anyone else. Never mind the fact that I can only recall ever buying from a catalog once in my life, the deluge has been relentless. This made me wonder what data this catalog company has on me and what kind of analytics, if any, they are running on that data to do targeted marketing.

I've been to dozens of conferences at which some retailers tout the power of their targeted-marketing campaigns. With just a few data points, they say, they can accurately determine who is likely to buy, how much they'll spend and whether it's worth it to try to engage individual consumers. So why then have the efforts of this holiday-themed catalog company misfired so badly?

One scenario is that they aren't doing any analytics. The company may have a set of names and addresses and is simply carpet bombing my area, hoping to hit at least a few targets. But this approach has a very 1990s ring to it, and I thought the industry was past this point. Also, blindly sending hundreds of pages of what amounts to waste paper to the mailbox of someone who happens to have an Appalachian Mountain Club sticker on their door likely is not the best way to engender positive feelings toward your products.

Alternatively, it could be that my past purchase history puts me into a profile that suggests I am a likely buyer, but that's hard to imagine. My credit card swipes are typically confined to gas stations, craft-beer pubs and Target. Does that seem like the profile of a shopper who would purchase a T-shirt with a picture of an eagle on the chest above the slogan "Tough Old Bird"? Nothing says 30-year-old city-dweller quite like a die-cast model of a '67 Chevy, right? If your algorithm answered yes to either question, it's time to go back to data-science school.

So it would seem that the catalog company either does no analytics or does bad analytics. I don't know which is worse. The National Retail Federation expects consumers to spend $617 billion this holiday-shopping season. To think that a retailer, particularly one that attempts to engage customers across the impersonal distance of theUnited States Postal Service, would take such a shoddy approach to attracting a portion of that spending is absurd.

This analytics fail also makes me wonder how many of those targeted marketing "success stories" I hear about are actually effective. It's one thing to talk about all the data you collect and algorithms you run in order to optimize your marketing campaigns. But unless you can show a meaningful lift in response rates or sales, you may be doing the equivalent of sending catalogs that feature Vincent van Gogh-themed neck scarves to a male whose most recent outerwear purchase was a pair of hiking boots.

Ed Burns is site editor of SearchBusinessAnalytics. Email him at and follow him on Twitter: @EdBurnsTT.

Next Steps

Marketing and advertising are top users for analytics

Targeted marketing campaigns raise privacy concerns

Big data and predictive modelingare perfect pair in marketing

Dig Deeper on Customer analytics

Join the conversation

1 comment

Send me notifications when other members comment.

Please create a username to comment.

Great article. Unfortunately most retailers are executing poor analytic strategies. Carpet bombing is common. Simple deciling on RFM metrics is common. True lift measurements are rare.