Skip to Main Content

Audience & Customer Cohorts Everyone (Even You) Can Use

Posted on 6.27.2017
Grouping people together based on shared characteristics at a certain point of time is the basic idea of cohort analysis (covered more in-depth in Website Magazine's July issue available next week). 

By understanding how these groups (email sign-ups in May, new customers in June, first-time visitors in July) "perform" can provide marketers with insights into patterns of churn, time to purchase, propensity to refer, course completions, customer lifetime value based on source, retention after sign-up and all kinds of valuable information that can be leveraged to improve campaigns in the future particularly when those cohorts are compared to others.

Creating, monitoring and optimizing for cohorts can be time consuming, of course, but once the reports are configured (check out this step-by-step guide for doing so within Google Analytics), many analytics systems help "tell the story" with some making predictions and suggestions based on that narrative.

Even before a data analytics platform is chosen (Looker, Tableau, Kissmetrics, RJMetrics, Custora, Adobe and others are popular choices), enterprises should consider what cohorts would be worth tracking. As part of our look at cohort analysis (again in the July issue of Website Magazine), below we explore cohorts for e-commerce retailers, information publishers and service providers as most Web-based businesses fall within these categories.

E-Commerce Retailers

Whether it's a tangible product like a chair or a T-shirt or a digital product like software or music, those providing a product to end-users have a wealth of data at their disposal. By grouping the following people together at a certain point of time, e-commerce retailers can start to see patterns emerge that can be helpful for future initiatives and forward-thinking (e.g., product improvements, ordering, etc.). For example:
  • Holiday shoppers: group this segment to understand their buying habits after the busy shopping season to help determine, for example, whether marketing dollars should be spent on acquiring holiday shoppers or on retaining existing subscribers based on their performance over the course of the year.
  • Deal seekers: group those who purchased after a major sale or promotion to identify whether they buy non-sale items as well, to determine lifetime value and other identifiers that will help marketers remove or add this cohort to campaigns in the future.
  • Customer Lifetime Value (CLV) by Source: Over time, monitor purchases made by cohorts who buy based on different sources (email newsletters, organic search, Facebook ads) around the same time.
  • Referral Revenue: Monitor customers who bought based on a referral and track how much they spend over time (this can help determine the cost a retailer pays for a referral like the bonus given to the original referral). 
  • First-Time Buyers: Group these by month, for instance, and monitor whether they will buy again in one, two, three or 12 months or never again.
  • Warranty Purchasers: Track those who bought an additional warranty during a specific period and how often they use it. 
  • First-Time Visitors from ___: How long does it take first-time visitors from Southern California, for instance, to purchase compared to another location?
  • Purchase Date & Issues: If a company providing a large-ticket item (e.g., home appliances, electronics) notices that people with X purchase date tend to start reporting issues around X months or years, they can proactively set up service to avoid further frustrations and improve the chances of that customer buying from them again.

Cohorts are only limited by one's imagination, but they should be created for reasons other than curiosity. A business should understand the true value understanding cohort behavior will bring to their organizations as the analysis is meaningless if it cannot be put to good use.

Information Publishers 

With many companies publishing content these days, it can be difficult for traditional information publishers to earn the clicks and conversions they desire (whether it's for their business or their advertisers). Like any business, however, there is data that can be mined to make better decisions (whether it's the content itself, where it's being delivered or how it's being promoted) and do more with what is already readily available. Here are some sample cohorts for information publishers to set for specific periods:
  • Acquisition Date & Activity: When did a person first subscribe to the newsletter, visit the blog or take any other conversion action and how did they perform after that? Did they view the blog daily, weekly, never again?
  • Device-Type Performance: Do iPhone users download whitepapers or other gated assets at a higher rate than desktop users during a specific period?
  • Frequency of Asset Downloads: Does a group of people who download gated material on Day 1 download any other material seven days later, 30 days later?
  • Page Views by Source: How many pages do people acquired from a paid search ad, for instance, look at in a period of time following their acquisition?

Like with retail, information publishers can create cohorts for any number of groups. The key is to ensure they are being grouped and monitored within a set period and that information can be gleaned that makes an organizational difference.

Service Providers 

Whether it's delivering a subscription or replacing a garage door, service providers have as much available data as their retailer and publisher counterparts. Let's take a look at some sample cohorts to create and track:

  • Referral Source & Return Rate: Say a plumber acquired a customer from a home warranty company but has since done work for the customer without the home warranty company's involvement. They can track these types of customers (even if it's by year) and monitor how often they return to them for service. If that answer is frequent, they may want to consider setting up their own monthly or yearly service fee to cover the types of services they tend to need.
  • Referral Performance: Create a cohort of people who have been referred by current customers and track their retention, upgrades, their own referral rates and more over time. 
  • Acquisition Date: Like every company type mentioned here, create cohorts by acquisition date and segment with time to purchase, time to upgrade, time to churn and more.
Whatever groups chosen to create as a cohort and monitor over time, be sure there are fixed elements (like acquisition date) and fixed periods to check in on them (it will depend on the cohort and business type whether that is daily, monthly, yearly, etc.).
computer-analytics
Today's Top Picks for Our Readers:
Recommended by Recommended by NetLine

Leave Your Comment

Login to Comment

Become a Member

Not already a part of our community?
Sign up to participate in the discussion. It's free and quick.

Sign Up

 

Leave a comment
    Load more comments
    New code
  •