There is a fierce competition for attention in your subscribers’ inbox. Now! is the perfect time to make your emails effectively deliver more value.

Segmentation is a seemingly simple but extremely powerful mechanism. Many are already segmenting to some extent, but how about taking advantage of more advanced email segmentation?

The art of thinking in groups: segmentation

Email marketing segmentation is the art of splitting your list into different groups. These groups – or segments – consist of people with similar characteristics. Segments can then be emailed reflecting the content and timing appropriate to them. So far, so good; emailing to groups.

Look beyond segmentation for increased results

There is an overwhelming number of ways you can segment your list. Still it is not only the segmentation that is going make the impact. If your marketing messages are highly segemented but badly executed, they will end up being highly segmented highly targeted lousy emails.

Segmentation Model x Execution = Combined Impact

Making your email segmentation more sophisticated is one of the ways to make a better match between your email campaigns and your subscribers. Crafting a sharper email or better use of the segmentation data (for instance more creatively) can be another. The end goal is to bring a plus to the bottom line, while retaining ROI. So the segmentation model needs to have enough impact and represent viable groups of sufficient size.

Execution smarts in segmentation can mean that you look beyond what you think is obvious. For example, instead of making the emails fit with the current place of your subscriber in the customer lifecycle and what they are currently buying, it can also be made to appeal to their aspirational self; the person they want to be seen as. Or to aim (or nurture) towards the future self; thinking ahead incorporating future needs and wants.

Layered email marketing Segmentation Models

Email marketing segmentation doesn’t stop at selecting group A or B. Once you start layering segmentation criteria, the possibilities are almost endless, it is only limited by the available data. That is why we create segmentation models. All the data in your database – or even combined, derived or external data – can be used to create these segmentation models.

Segmentations can be based on all types of data, including:

Preferences (the likes and unlikes of a user)
Demographics and profile (age, location, gender)
Psychographics (What will they be prone to do and react to?)
Behaviour (purchases, opens, clicks, website browsing, etc)

Attention all email marketers: If you have behavioural information available, it is extremely important not to let this go to waste. It might be a bit harder to get that data prepped for marketing purposes, but once it is set-up it can be a great indicator for future interests and behaviour. By sorting past order and purchases by category, price range, etc  a whole new plethora of segmentation options opens up.

7000% increase in email marketing revenues

An example of huge increases in revenue by segmentation comes from Totes Isotoner Corp.. They noticed that a lot of online shoppers repeatedly visited just a single product category: umbrellas, gloves, or boots. They decided to pitch these consumers with segmented email marketing campaigns focussed on the categories that were visited. The email offers brought many of the “lookers” actively clicking and buying.

Data can be combined or layered to achieve the level of targeting needed. For instance sending an invitation for a popconcert, we select the people that live close to the location and have indicated that they like that type of music. We add a focus on one of the artist that is performing at the concert if they bought a song or browsed the artist’s page recently. Of course we send a very different promotion to the group that is more interested in very different type of music.

Advanced email segmentation

More advanced email segmentation campaigns make use of combined or derived  data to go a step further. For instance segmentation on Customer Lifetime Value, crafting Buyer Personas or propensity modeling.

Segmenting on Customer lifetime value

For instance the customer lifetime value can be a segmentation criteria. It takes multiple pieces of purchase, costs and profile data to make such a calculation. But after you have calculated customer lifetime value it is possible to target your best customers, exclude your worst customers or craft specific campaigns for the part of the email list with the highest potential for growth.

Buyer Persona’s and customer behavior
There are different types of clients. Sometimes these buyer types are typecast as personas, but it can on a different (more practical) level too. By selecting the people that were high spenders and have seen a decline in spending the last months, you are looking to spending patterns. Looking back at previous behavior it is possible to identify and craft your messages to fit this group.

Predicting future behavior with propensity models

A group of statistical methods and scorecards that is used to predict future behaviour of customers. This behaviour includes the likeliness to respond to different types of offers, risk of churn, cross and upsell possibilities, etc. Now you might be thinking: how advanced and sophisticated should I make my segmentation? This differs from case to case. Let me give you an example.

An example of email marketing segmentation

An online fashion retailer for instance, can get great results segmenting on gender. Showing the men’s items to men and women’s’ items to women. But if they combine it with size and style preferences they can level up to even better results. Offering the right sizes on the landing pages and excluding sizes not in stock. Offering casual, business, and other styles the recipient is most likely to buy.

Adding the amount someone is likely to spend can furthermore increase results. Send sales emails to coupon clippers and VIP invites to exclusive high end shoppers.

Excluding the ones that just recently purchased and offering them an alternative makes those email more effective for that group.

Choose your segmentation model based on impact

Now which of those (or combination) has the highest impact per recipient effected? And which has the biggest impact on bottom line total revenue? Yes, some segmentations can be very costly to realise.  If you are playing around with segmentation models, try creating grid of segmentation options, costs versus their impact to find the “Quick win” and “build upon” segmentation models. Then test and improve upon it before making it your base and your running further segmented campaigns.

Segmentation as a part of the email marketing plan

Every time you select a group from your email database, it’s email marketing segmentation. This is a seemingly simple but extremely powerful mechanism. An email marketer can slice and dice his list and craft messages that fit the segments, ultimately lifting results. Segmentation should therefore be a part of every healthy email marketing strategy.

Image by Alice Bartlett (cc)