6 reasons over-segmentation is hurting your marketing

Reading Time: 4 minutes

You have a nice plaque on the wall showing your mission, vision and values.

Do your employees think that’s what you really stand for?

You also have a nice list of value proposition points in your marketing department.

But, is that what your customers really think of you? How do you know?

Where did those value proposition statements come from? Possibly from awesome customer research methods that actually deliver spurious results.

The most important thing you can optimize is your value proposition.

Your value proposition is not what you write in your marketing planning documents. Your real value proposition is what your customers perceive. It’s what the world actually thinks of you.

To add a wrinkle to this, each person is different. Yes, the snowflake analogy works here.

Each person looks at the world through different eyes. Your visitors have unique frames of reference that are colored by their perceptual filters. It’s as if they each wear different glasses that change what they’re looking at before it reaches them.


That’s true. Everyone is unique and perceives uniquely. That’s one of the reasons your small member qualitative research doesn’t apply to your overall market.

Many today are striving for a technological utopia where you can divine individuals’ uniquenesses, place each prospect into a personalized box and target him and her with self-actualizing, psychologically-triggered ultra-messages.

But, as much as technologist would have you believe that you can speak uniquely to each individual and magically hit their individual needs and wants, life ain’t like that. It’s not so simple, and there’s a better way.

The key is in finding relevant prospect groups. Your goal should be to find the largest possible relevant segments that can be aggregated. Yes, I said the “largest possible.” Gasp!

But, hang on for a moment, Chris. Don’t you buy into the drive to solo-segmentation?

In a word: “No.”

But, segmentation *is* important.

Segmentation is important for relevance. But too much segmentation is harmful to your results.

It’s just like carrots. You know they’re good for you, but too many could give you adverse side effects.

The most important thing you can optimize is your value proposition. And the drive to infinitesimal target segments hurts your value proposition discovery.

Customer segments allow you to group people with similar needs and information filters. As a marketer, you need to look for more commonalities among customer needs rather than just differences.

6 reasons you should create the largest possible target segments

1. Most importantly, it allows you to optimize. Testing and optimization requires traffic. Traffic gets much less useful when it’s split up into dozens or hundreds of tiny segments.

2. It reduces maintenance costs. Imagine you have a product or offer update and you need to change your landing pages to reflect that. Would you rather update and QA (you do a thorough QA process, right?) four pages, or four hundred?

3. Clarity trumps personalization. Clarity of your most important value proposition is more valuable than individually unique customization. You’ll make greater improvements with finding the right message for your product than individualizing your message.

4. Macro insights beat micro improvements. The faster you can find business-wide insights in messaging, design, layout, content, eyeflow, copywriting style, etc., the faster you can make improvements that generate huge revenue lift company-wide.

5. Small data lies. Too much customization is based on increasingly spurious data. You can’t really know anything about a person based on a handful of actions. Over-customizing messaging often increases Distraction and reduces Clarity. Mass personalization technology isn’t new. It’s just experiencing a revived sexiness online.

6. Over-personalization is creepy. Unless you’re targeting digital marketers like yourself, who are excited to see examples of personalization, your customers are creeped out by how much you know about them. Just ask the NSA how we feel about it.

Prior to founding WiderFunnel, I won awards for planning a complex, mass-personalization direct marketing campaign. And I can tell you that, as mind-bendingly complex and technically advanced as it was, the marketer got more benefit from testing the overarching messaging.

Yes, technology exists to target infinite segments. But, just as with any shiny, new marketing toy, the benefits must outweigh the costs.

Emphasize value proposition testing

Start by testing your value proposition to find out which points are most important to emphasize. Then, segment your messages with discipline, only creating new target segments that are proven to respond uniquely.

If you need to brainstorm value proposition options, here’s a framework I developed to answer the question “how do I create an awesome value proposition?”

What do you think? How do you optimize your segmentation? What segmentation methods have you found most useful? Add your comment below.

Update: There’s now a follow-up counter-point to this article showing how to get great results from website segmentation.

Enjoy this post? Share with your friends and colleagues:

  • Blair Lyon

    Chris, well said. It's natural to focus on micro-segments or try and go for the 1-to-1 experience, but this is often a fools errand. You need to go after the largest most valuable segments first, then work down from there.

  • I take it you're not a particular fan of the likes of Monetate then Chris!!? Great article, thanks for sharing your experiences and recommendations

  • Hi Chris,

    You say: Clarity trumps personalization.
    Have you tested this? Just wondering because I think it could go both ways.

    • Yes, we regularly test many types of segmentation, personalization and targeting. As I said, segmentation can help with Relevance.

      However, these tactics are often given too much emphasis, especially by technology vendors. Don't let them distract you from the bigger potential wins.

  • You’re so right.

    But there ar exceptions, also. One of our clients is selling winter clothing. So, his segments have 4 layers:

    Traffic source (different purchase intention & reasons to abandon cart)

    Location (different delivery options)

    Behaviour (different expectations)

    Weather condition.

    Through interactions, he is pushing sales only for those that are in locations where it’s snowing.

    That means he is using up to 16 mixed segments.

    • chrisgoward

      Yes, those seem like reasonable segments to test treating differently, Valentin. Thanks for sharing!

      The question you should ask through your tests may include:
      – Which traffic sources respond to messages differently and which can be aggregated?
      – Which behaviours lead to measurable response differences?
      – Which messages are most relevant in different weather conditions and do the assumed weather conditions really affect sales patters? Can we test offers across weather patterns to test our assumptions?
      – Are these the most important segmentation methods?
      – Which elements of the website can we keep consistent across segments to minimize maintenance?

  • Very good Chris. I think this is particularly relevant to B2B marketers (like myself). We tend to (over?) use our Marketing Automation systems to 'slice and dice' into segments. Sometimes getting insight (and good valid data) get's lost in the mix.