Overview
Cohorts are just groups of customers who share some kind of action, trait, or behavior for a given business. The strategy team has developed a Purchase Frequency Cohort Analysis that buckets customers into cohorts based on how many times they have bought from the brand.
This type of analysis is helpful because it allows us to make strategic business and marketing decisions, identify areas that are underperforming, and guide us to determine the right amount to spend to acquire a new customer. Understanding what percentage of customers actually buy more than once, what their typical lifetime value is, and how soon after their first purchase do they buy again all helps in making a business more successful.
Below are common findings and recommendations you can give to your client:
Many Customers Only Buy Once
Large majority of all customers are in the “one and done” cohort
This implies a lack of need or want to purchase again. Attempt to investigate and find the root cause issue (reviews, discussing with the client, data, etc). Examples:
Product quality or value
The customer is disappointed with the product that they did purchase, and therefore are less likely to buy again or recommend it to a friend. When new products become available, they disregard them. Marketing to these individuals is wasteful and unlikely to “re-convince” them.
Recommendations:
Reduce retention spend and efforts as much as possible
Work to improve the quality of the actual product itself.
Service quality or process
The customer is disappointed or frustrated with the actual process of purchasing or acquiring the product. Slow shipping, poor customer service, difficult experiences, or similar problems make a customer unlikely to buy directly from a store again even if they like the product. This often shows up in brands that sell both on Amazon and direct or through retailers, and get a majority of sales through other venues than direct ecommerce.
Recommendations:
Shift budgets and resources toward "better" platforms (such as Amazon)
Simultaneously improve the UX, customer service, and logistics for D2C
Low variety or repeat need
The customer is satisfied but there’s not much in the way of new variety, need, or use-cases for further products. A common example of this is a utilitarian tool that has niche applicability; the customer buys it once and it will work for them for the next 30 years. There’s little to be done from a marketing perspective to solve this, other than to reach more customers or encourage the business to expand into new product lines.
Recommendations:
Develop new products, bundles, or accessories. Until complete:
Expand into new markets (such as new geos)
Expand into new customer segments (such as a new persona)
Deploy a "broad base" or "mass reach" campaign
Loss leader or hook product
Many customers purchased a product when it was heavily discounted, or when it was made for a specific event. This is most common around Black Friday, or if a one-off sampler or similar product is released. Many new customers came in off of this great deal, but aren’t actually within the businesses’ ideal customer profile (ICP) and are unlikely to buy again.
Recommendations:
Actively attempt to upsell to these customers
Potentially with similar promotional messages (through affiliate or email, etc)
Paid marketing should be minimized against this audience
Especially if they have characteristics (demo, HHI, etc) outside of the ICP
Poor cross-sell or upsell
Customers simply aren’t aware of or encouraged to buy other supplementary products that are likely to be valuable to them. Perhaps they love the initial product that they got, but don’t know that the brand also sells a complimentary add-on, or a similar quality product in a related field. This is common in tech or hardware brands (e.g. the customer buys a high quality PC gaming mouse, but doesn’t buy their keyboard or monitor).
Recommendations:
Build a more intentional lifecycle program (better email, SMS, etc)
Analyze and likely shorten the window that upsell emails are sent
Many customers buy within the first 30-60 days of their first purchase
Segment and personalize email strategy by cohort or product set
Take advantage of SMS and easy conversions
Consider paid retention or win-back campaigns against lapsed audiences
Add post purchase up-sell
Complete a basket analysis to determine products that naturally bundle
A basket analysis can be completed by the Data Intelligence team
Pareto Cohort Buoying the Business
The vast majority of revenue comes from a small segment of the customer base
This is sometimes called the Pareto Effect or the 80/20 rule, and is common in brands that are more established that have recurring revenue (e.g. fashion, subscriptions, etc)
This implies that if someone can be converted into a “whale” or “loyalist” they will exponentially increase their lifetime value.
Recommendations:
Analyze the loyalist’s characteristics: do the high value loyalists have a demographic or behavioral trait that makes them more likely to buy? Income, location, demographic, education, affinity, etc.
This can be difficult to do without third party data enrichment, but look at Google Analytics and DataQ for guidance
Once these customers are well understood, build a new program around converting more of these ICPs
Product and basket analysis: do the high value loyalists buy a certain product or sequence of products that is different from the brand’s core customer? Is their entry product different from most?
If so, shift marketing’s efforts to focus on the high-value product chain and de-emphasize the others.
Convert to loyalists: build specific upsell and cross-sell programs that target ‘high potential’ but low value cohorts to become high value loyalists.
This can be through specific lifecycle marketing initiatives, but also some very targeted paid media targeting those segments to attempt to incentivize them to start more recurring based revenue or high value products.
Geographic analysis: a secondary analysis should be completed to determine if the high-value cohorts are from a particular geography, such as a country. Many businesses find a global audience, but identify that one particular country might be worth 10-20x per customer than others (e.g. the US customers spend $1k whereas customers from India spend $100).
If this is the case, efforts should be made to at least set up tailored CACs by country; or completely remove paid marketing from low-value geographies.
Second Purchase Delay Is Different
The cohort analysis shows that customers make a second purchase quicker or slower than anticipated or seen historically
There’s no benchmark for “typical repeat purchase time” as it varies so significantly by product, brand, vertical, and customer location
However, most ecommerce brands are most likely to buy again in the first 30 days after first purchase, or right when they get it shipped to them
This implies that the business made blanket assumptions about behavior.
Recommendations:
Repeat purchases are quicker than anticipated: send lifecycle marketing sooner than what you currently do.
If the brand assumes that customers waited 6 months before considering buying a second product, but the cohort analysis actually shows that the typical second purchase occurs only at 3 months, then the lifecycle marketing efforts deployed at the five month mark are probably too late.
As such, begin retention efforts two months after purchase to encourage that second purchase and increase the overall repeat purchase rate.
Repeat purchases are slower than anticipated: send lifecycle marketing slower than you currently do.
Sending emails and similar upsell or lifecycle messaging too early can come across as a nuisance and actually damage the likelihood that a customer will repeat a purchase. As such, the brand should delay these efforts to approximately one two to four weeks prior to the typical repeat purchase rate.
Normal Distribution of Cohort Value
The number of customers and their value follows a bell curve, in which the majority have more than one but fewer than many purchases
This indicates that with careful optimizations the entire lifetime value of the audience can likely be increased, with a greater number of customers purchasing at a higher cohort range. However, many brands are happy to achieve this type of curve and don’t wish to upset the delicate balance.
Very Few Low Value Customers
The majority of customers repeat purchase and are very receptive to upsell
This implies that the product and relationship to the customer is very good, but perhaps pricing is too conservative or marketing not aggressive enough
In short, the brand will want to take advantage of this rare opportunity:
Increase frequency of lifecycle messaging: continue to moderately test and experiment with more frequent messages across email, SMS, and winback campaigns in paid marketing. Some brands find their audience is surprisingly resilient to these messages and can tolerate as many as daily communications.
Product expansion: these customers love your brand and the products you create, so continue to solve their problems and they will continue to buy from you. When new products are launched, build a preferred pre-order program to reward the loyalists.
Recommendations:
Invest more heavily into acquisition: if customers are likely to remain loyal and provide a high value, then the cost to acquire a customer (CAC) should also increase. Marketing can get more aggressive, new first-time purchase incentives can be launched (e.g. affiliate or promotions), and heavier brand building efforts can be rolled out (CTV, billboards, broad-reaching video, etc).