Subscriber preferences: four ways they can mislead you

Past months have seen a lot of excellent studies done on the way people use email and other online channels.

These studies teach us about the content and offers people prefer, how they like to receive the info, and how these preferences vary between different demographics.

[I'm a big fan of the work done, for example, by Morgan Stewart and colleagues in this field.]

Email marketing campaign results are preference studies in their own right, giving insight on what content or offer works best for your list.

Matching what you send to what people prefer seems like a win-win situation: subscribers are happier and they click more.

But while preference information is incredibly useful, it needs careful interpretation. Here are four traps to watch for…

1. Do they get what they want…or want what they get?

Tom sells oranges. Just oranges. And people buy them from his market stall.

One day he decides to ask his customers why they shop at his stall. The most popular answer?

To buy oranges.

Excellent, thinks Tom, I’m giving my customers exactly what they want.

Well, yeah.

People would actually rather buy apples at the market. But Tom only offers oranges, so “buying oranges” is why they go to his stall.

The public doesn’t get what the public wants. The preference survey reveals that the public wants what the public gets…as The Jam pointed out in 1980.

When people tell you why they sign up for email, the answers are based both on what they truly want from email and simply on what’s available by email.

If nearly all marketing email is promotions, discounts and sales notices, then it’s hardly surprising that people cite these factors in surveys on subscriber motivations (see this eMarketer article).

So if you ever ask subscribers why they signed up for your emails then also ask what else they’d like to get. That way you’ll know whether they get what they want or simply want what they get.

2. The long tail in bar charts

The cult soccer manager, Bill Shankly, once said, “If you are first you are first. If you are second you are nothing.”

This isn’t true in subscriber surveys and campaign reports. But we all have ADD, so we (and the media) focus on the top survey result or the most-clicked link – the “winners” – and ignore the rest.

Even if the difference between the “winners” and “losers” is just a percentage point or two.

But email is not a soccer match or a horse race. There are no winners and losers in the classic sense, just more or less popular results.

Why does that matter?

It matters because we ignore important information just because it’s not at the top of the bar chart or column of figures.

The data revealed, for example, in the eMarketer article show the sign-up motivations cited most often by subscribers are to get discounts, freebies and notices of sales.

That’s important to know if you’re a retailer, but 45% also mention “want updates on products.” Should you ignore that motivation, just because it was further down the list? Definitely not.

You might not want the focus of your emails to be product updates, but there’s a hint there about what you might put in an email sidebar, for example.

3. The cuckoo segment

A related danger is that we see all respondents to a survey, or all the people who clicked on the last email, as somehow representative of everyone else.

We forget that our lists are made up of individuals. We can group these individuals into segments sharing some characteristic, but we certainly can’t regard them as clones of each other.

So when a survey question, or an article/offer draws a large response from one segment, we forget that this segment is NOT speaking for everyone.

A vociferous minority can dominate your attention at the expense of the less-vociferous majority. The loud minority elbows the majority out of your mind, like a baby cuckoo freeing up the nest for itself.

Example 1:

My content newsletter for email marketers gets CTR peaks when it features specialist articles on email design. That’s because there is a minority of designers on the list who really love such articles.

If I assumed CTR was a measure of total subscriber preference, I’d only write about email design. But…

• the majority of people on the list aren’t that excited by design content
• the “email designers” also click on articles about other topics

Focusing on design articles alone would be good for CTR, but a huge missed opportunity. I’d be keeping one segment very happy and inviting everyone else to ignore me.

A better approach is to see measures like CTR as indicative of segment preferences and segment the list and future emails appropriately.

The challenge, of course, is identifying those segments. More on this in the article Beware the Wisdom of the Minority.

Example 2:

What if a national survey showed email responses were highest for messages received at 2pm? Now imagine your emails promote last-minute lunchtime bookings to your restaurant’s local market.

Obviously you’re going to ignore that survey result, because it makes no sense in the context in which you operate.

Yet we forget that lesson all the time when reading about surveys. The survey audience isn’t our audience. We must take the insights and filter them through the context in which we operate first.

4. Raspberries or Strawberries?

My kids like clear answers. “Do you like strawberries or raspberries, Dad?” I have to choose.

Thing is, I like both. Saying I like strawberries is not the same as saying I don’t like raspberries. Yet many survey responses that don’t allow multiple answers are interpreted in exactly this way.

There’s a similar trap when you look at click rates. The likelihood to click is not just a reflection of fundamental interest in what’s behind a link.

Say a web marketing newsletter contains summaries and links to three articles on SEO, email marketing and social marketing.

I might click on the SEO article because it’s the only topic that interests me: that click is a true representation of an absolute content preference. I like SEO, I dislike email and social. In future, you should send me SEO articles only.

But there are other scenarios.

For example, I have little time, so I click on the most interesting article. But the other two topics are also interesting.

My click now represents a relative content preference. I like SEO most, but I also like email and social. Don’t stop sending me articles on email and social!

What if the SEO article is featured first and most prominently, and with the most powerful call-to-action?

Now my click is not just about content preferences at all. I may even like email and social more than SEO, but the links were below the fold!

Tricky!

Solutions

None of the above should be taken as advice to ignore subscriber preferences and what we learn from surveys and campaign results.

The message is the opposite…we need to take more care to find out the real preferences and then account for them as best we can while addressing our own business needs.

All the while avoiding the kind of simplistic interpretations you get from many media articles.

Interpreting surveys and campaign reports is a big topic. And there are others better qualified than me to advise you on it.

My own approach is simply to be wary of taking everything at face value. Be aware that the “easy” interpretation is not always the correct one. Question everything. Test everything. Assume (almost) nothing.

Easier said than done, but worth the effort…

Find related articles:

• Channel: Email analysis

You can follow any comments on this blog post through the RSS 2.0 feed.