top of page


What are the consequences of taking retention lightly?

It’s a common assumption among early stage founders that growth = acquisition. “I get more customers, my startup grows”. That’s like saying to fill a bucket you just need a large tap. In an early stage founder’s mind, that tap looks like a firehose and the water is free, and never ending. In reality, it’s not a tap, but a hand pump, and sometimes it’s expensive to pump! Even if it is a large firehose, if you don’t plug the holes in the bucket, you can run out of water - this is what’s happening to Twitter - it’s running out of people to acquire!

Retention is about ensuring you plug all the holes in the bucket. Retaining existing customers is (1) way cheaper than acquiring new ones (2) proves your product is still the preferred solution among all the options available  (3) bridge to upselling, expand to new product lines, acquiring new customers through referrals, word of mouth etc.


What & How

It costs considerably more to acquire a new customer than to retain or keep a current customer (4 - 7 times more, according to various popular studies). For early stage startups Retention is a measure of Product Market Fit


For those Post PMF, Retention is about minimizing acquisition costs & maximizing LTV (and hence Revenue and $$$ to spend). Beyond Iterating on Product to create habits, post PMF startups should also pay attention to Retention Marketing. Product personalization, habit building features, push or email notifications to re-engage with product (observe emails that Medium, Twitter, Linkedin, Facebook etc. send you), Retargeting, Abandon cart features/emails, upsell tools, loyalty and rewards programs are all Retention tools that are relevant at different points of customer lifecycle.


The period used to calculate retention is dependent on the natural behaviour of the user in the context of your problem. For example, a high frequency usage app like a messenger app is expected to be used every day, (in fact, several times a day), so retention is measured in Day 1, Day 7, Day 14, Day 30 etc. intervals. A clothes shopping may consider someone who hasn’t bought in 3 or 6 months a churned customer.


Common Mistakes Founders Make

  • Not talking to lost customers enough.  If you have enough volume, your data can tell you “what” happened (Ex: 5% churn rate means you lost 5% of your customers this past period). Talking to customers helps you understand “Why” they left.  This data is a lagging indicator and you attempt repairing the relationship/resurrecting the customer.

  • Not talking to current customers enough. Since talking to them after losing them only gives you a chance at repair, but not prevention, talking to current customers can be leading indicator of what might be coming up. People like knowing their opinions are important, ask nicely. You can ask for feedback every time you’ve delivered value (via email, push notification, survey, phone call, in-person meeting…). They are bursting to share every time they love or hate the experience you delivered them - so look for places in your product when they expect something of you and you think you delivered. Ex: Thank you page, when product is shipped/delivered, when they’ve used your app/service, experienced core value for the first time/10th time... Make sure your asks are contextual and relevant.

  • Not understanding effects of good/bad retention rates are compounded over time.


Fictitious Example

Best way to measure retention is tracking the engagement or core activity of a set of customers over a period of time - also called cohort analysis. In this example below, while Startup A initially looks like it has better retention (more of the initial set of customers are returning on day 2 (or week 2 or month 2), eventually, after a few periods, no one returns - this curve go towards zero. As opposed to Startup B, where the engagement curve flattens out - indicating a non-significant percent of customers continue to use the product after a long period of time - there is at least a portion of the initial customer base that finds the product useful enough to continue to use it.

Real World Example

Andrew Chen talks about retention for mobile apps

Quettra blog post about retention of Twitter app:



Related Keywords

Cohort Analysis, Churn, LTV, Resurrected customers, Net Negative Churn

Related Links

Growth is good, retention is foreverBrian Balfour - VP of Growth at Hubspot

Retention and Product Market Fit - Casey Winters, Growth at Pinterest (slides)

Lessons in retaining users - Casey Winters, Growth at Pinterest (slides)

Retention Marketing - Jeannette Sacks (Slides)

Retention and Cohort Analysis (slides)  Bei Lu, Director of Analytics at Smule

bottom of page