Don't send me the same message as everyone else!!!



DON’T SEND ME THE SAME MESSAGE AS EVERYONE ELSE!!!



Everybody is different.  When you’re talking to someone you may repeat elements of what you have told other people, but the conversation will be unique in that moment.  It will be based on what you already know about them and the interactions you have during that conversation.  They would think it was weird if you said the exact same thing to them that you said to everyone else without responding to their questions, intonation, and body language.

So, interaction on a personal level usually works best.  But let’s say I have 10,000 customer records on my webstore and I want to tell them about my new range of products. It could take me about 100 days to have those personal conversations (assuming each interaction lasts 3-4 minutes and I take 8 hours a day for sleeping and eating).  That’s not going to work very well is it!

Segmentation vs Stereotyping

To get around this we use segmentation.  We stereotype or characterize our customers based on common characteristics.  This lets us build campaigns that will be more engaging to that group of people than a generic email blast. I know stereotyping gets some bad press, but it is a very efficient way of starting to deal with people when we have limited information on them (we do this subconsciously and our subconscious mind is very effective).  As we get to know them better, we can get more tailored in our messaging – we can ultimately create personalized engagement with them so they become a valued customer.

Personas

A useful way of developing campaigns is to create personas for those segments, where it makes sense (i.e. for a valued segment and we know enough about their traits to build a reasonably common story around them), and then build out their likely customer journey.  This allows us to understand what their pain points are and when/how we should be engaging with them.  Artifacia[i] have put together some great sample personas for online retail.  You could select a few to use as a template in order to create your own.



How do I start putting together my customer segments? 

Well, your store is unique, so your segments will be different from other stores.  Sometimes the difference can be small but still significant.  The good news is that you already have the data in your store to allow you to do this.  Your store holds details of customers who have purchased from you.  You can start by segmenting based on their characteristics.  An objective for how you're going to categorize the segments is required - customer Lifetime Value (LTV) is a good one.

Segmentation Characteristics

Characteristics for segmenting can get very detailed and the more detailed you can be the more effective your campaigns will be.  Examples of characteristics are:
  • Demographics (age, salary, marital status, education, occupation, etc.,)
  • Geographic (zip code, country, city/countryside, etc.)
  • Behavioral (benefits sought, usage, loyalty, etc.)
  • Psychographics/lifestyle (opinions, attitudes, personal values, channel, etc.)
This can get very complicated very quickly and you might start thinking you will need a data scientist to help you!  However, a good place to start is with any of these three objectives below that are related to LTV.  You will have them in your current data:
  • Recency – when was the last time the shopped with you
  • Frequency – how many orders have the made over the last period
  • Monetary value – how much have they spent with you
Shopify[ii] have a great article explaining these elements and how they can be used for creating a combined LTV score.

How Many Segments Should I Have?

Fundamentally the segments that you setup should satisfy these requirements[iii] to some degree:
  1. Identifiable
  2. Substantial
  3. Accessible
  4. Stable
  5. Differentiable
  6. Actionable
Initially its best to start with a small number like 3.  That will make sure that they are substantial enough but not so overwhelming that it is not actionable.  Run the whole process for these 3 segments starting with the highest LTV and then increase the number of segments as you are; gaining experience, gaining customers, have time, and ad money to spend.  You can start performing A/B testing to understand what campaigns work best with each segment.  Overtime start to build up a set of personas that help you humanize and build empathy with that segment.  Ultimately you will build a better understanding of the segments (but be aware the segments themselves will also change overtime) and know when you need to split, recombine and change their characteristics.

Custom Audiences

Now once we have understood the segmentation (we are now customer centric!) we can expand our campaigns to potentially new customers that are similar to our current customers.  We know what this type of customer likes.  Facebook and Google ads will create these new audiences as long as you can provide them with the details of the segment you want to base them on.  For a deep dive into custom audiences check out this post[iv] by Neil Patel.  These really work, SoFi experimented[v] with this on Pandora’s platform and achieved 18% higher leads and 93% higher conversion rate using a lookalike audience instead of a generic audience of college graduates (SoFi’s usual customer)


Conclusion

In summary, if we want to stand out from the crowd then we need to be able to understand our customers in more detail.  That allows us to engage with them on a more personal level.  It’s usually not practical to do this for every individual, so we need to use manageable segments.  We can then tailor our campaigns to be the more effective. Platforms like Facebook and Google let us add more people to our segment based on the characteristics of our base segments. 


Do you use segmentation or personas today and how many different segments do you use for your campaigns?

[i] Verma S., Nov 2016, 12 Buyer Personas for Online Retail Shoppers, http://blog.artifacia.com/12-buyer-personas-for-online-retail-shoppers/
[ii] Donnelly K., Jul 2017, Why Customer Lifetime Value Matters (and How to Calculate It for Your Business), https://www.shopify.com/blog/customer-lifetime-value
[iii] Gavett G., Jul 2014, What You Need to Know About Segmentation, https://hbr.org/2014/07/what-you-need-to-know-about-segmentation
[v] Pandora, Sep 2018, SoFi Drives New Leads with Lookalike Modeling, https://www.pandoraforbrands.com/success-story/sofi-drives-new-leads-with-lookalike-modeling#_ftn3

Comments

Post a Comment

Popular posts from this blog

I’m doing Emotional Marketing, why do I still need data?

Image
 I'M DOING EMOTIONAL MARKETING, WHY DO I STILL NEED DATA? Emotional marketing is very powerful but also very difficult.   You need to be able to create an engaging story with empathy.   If it is not true to your brand and doesn’t resonate with the customer, it could easily backfire.   Even when it is successful, it may alienate some other people.   It takes real understanding and commitment. The Ultimate Guide to Emotional Marketing Hubspot have a great blog that explains Emotional Marketing and how to do it [i] .   The steps it describes for an Emotional Marketing Strategy are: First know your audience Lead with color Tell a story Create a movement or community Inspire the impossible Project an ideal image Actions are driven by emotions – data can track actions The first step involves using your data and engaging directly with customers.   It should be a thorough understanding of your data in order to segment, create p...
Read more

Where am I going to find more customers?

Image
WHERE AM I GOING TO FIND MORE CUSTOMERS? We have some great products and services to sell!    Our little community loves what we do!    How do I find more customers like the people in our community? Easy!!!   Just setup a custom audience of your best customers, then create a lookalike audience of new potential customers that share the key characteristics of your best customers. Who knows them? Organizations like Facebook, Google, and LinkedIn have a vast amount of data on individuals and potential customers.   Not only do they have fundamentals such as demographics and geography, but they also their interests and behaviors.   How do I define my type of customer? Start with a group (segment) of people (customers) that you have a unique identifier for (an email).   By applying some machine learning techniques (ML) you can identify common characteristics for this segment.   Then using some more ML you can identify other peopl...
Read more