Retention causes virality, and vice versa

This sounds straightforward, but completely oversimplifies the problem:

Make sure your product is retaining your users, THEN work on growth. Don’t work on growth until your product is working.

This sounds right, but it’s too blunt of a rule.

For fundamentally social products, it’s hard to separate retention/engagement and virality. Turns out that for fundamentally social products, retention causes virality, and vice versa too.

Engagement to virality
Engagement causes virality because of a simple idea: New users won’t create a viral factor >1 in their first visit. Not even close. So in order to generate any meaningful amount of virality, you generally need multiple visits and multiple opportunities to take them through a viral flow that generates more friends. As a result, it turns out to get growth, you need people to stick around so that they can keep inviting and keep sharing.

Virality to engagement
Growth causes engagement because you need to activate people and keep them engaged. A meaningful amount of retention for any social product comes notifications. It could be from people following you, commenting on your content, or otherwise. If you don’t have a steady dribble of notifications coming into your inbox every day, then you won’t have the opportunity to bring people back into the product. A large % of these notifications will be caused by new users coming into your product, and the small # of actions they do on their first day. So you want them around, and it’ll keep your engaged users happy.

Chicken and egg
So if you have a chicken and egg problem, what’s the right way to solve this?

Well, you don’t need scalable viral growth to get enough users onboarding and generating notifications. You just need a little trickle of growth, and that might be from ads, blogging, PR or something else. You also want to make sure your social product has a low threshold for the minimum social graph required to keep it working- you can either do that if the product would work with just your friends and family, or if you’re going after a densely connected vertical.

All set?
If you have a trickle of new users and there’s enough people in the product to be interesting, then you’re all set. Then you can turn your attention to engagement and retention. Keep the users you get, have them generate more users, and you’re quickly on a good path.

Like this post?
Get new updates via newsletter..

Zach Hobbs

No question that retention is critical for success. My question is, how do you measure it? Month over month, week over week? Retain a user for 6 months? Etc. And what values for these indicate that “your product is working”?
http://andrewchenblog.com Andrew Chen

Every product’s metrics are different, so it’s hard to figure out. SaaS often uses churn rate. Social products often use DAU/MAU or month-to-month retention rates. It just depends- there’s a lot of comps out there, on Quora and Slideshare or just ask around.
http://twitter.com/stephenkeep Stephen Keep

Thanks Andrew great post. One question how big a trickle is required? Are we talking hundreds or thousands of new users a day?
http://www.InkTechnologies.com/ Jillyan Scott

Retention is very hard to measure. In my years of working with contact centers for various brands, I have learned that customers tend to be loyal to that brand that treats them right. I mean most businesses will make a mistake with one customer but what sets it apart from other companies are probably the measures it will take to correct that customers experience and turn it from a bad to good experience.
http://twitter.com/brianpiercy brian piercy

There’s plenty of markets/customers where this unfortunately doesn’t apply.

1) competitive reasons – your product provides a boost in performance or function that constitutes a competitive advantage for them. Virality will be actively discouraged.

2) it’s more than a signup – many products require much more than a “like” to encourage adoption. Integration details and product quality data are just the tip of the iceberg. This can still count as a win, but it only represents the beginning of another sales cycle.

3) the “hassle factor” – many demographics are still disdainful of using social tools to make purchase decisions in enterprise environments. Reminding someone to share a like/signup/purchase can cause unintended consequences.

3) customization leads to “where did it go?” – if you’re selling a single “SKU”, great. What happens when customer #2 requires red instead of blue? Did the customer make the buying decision in a subsequent visit? How are customers #1 & #2 linked?

All I’m saying is … tread carefully.
http://www.guotime.com/2012/03/3-simple-ways-to-stay-inspired-motivated-as-an-entrepreneur/ grant

HI Andrew,

This is a brilliant and much needed post. Short and sweet and to the point right away. When a potential customer tells me that my product is a great retention tool, my question was: how do I get customer first? It turned out, when we serve that one customer well by solving his retention problem, he became of first trickle of customers, words will spread from there. It takes long time to figure that out: serving one client at a time and run our startup as a small business first. Now the question is: how long it usually takes from this first step to steady trickle of customers? Can we engineer and replicate the first one quick and effectively? Is the talk between you and Noel Kegan applicable?(have not watched yet)

Grant
Love to schedule a coffee/tea talk in sf/sv
btpayson

this is a very interesting piece, thanks Andrew. I would be interested in hearing about some of the parameters of retention KPIs. For instance, measuring retention for account holders, accounts with free content, and account with paid content all would present different retention measures and timespans in which they are deemed retained or lost. Also, if your product is content, services, and aggregation focused, then these all have different KPIs. Account holder retention could be less than a couple days without engagement/cultivation efforts (email welcome series, account progressive completion, etc.) whereas an account holder with previous purchases has a vastly different retention window. Some general KPIs could be days since last use (duh, no brainer there), days since last consecutive days of use, use of most current app version, days since last usage of social/sharing/engagement features, average use over time (page views, time on site/app, etc.). What are some other KPIs people are using? Also, curious about what timeframes to consider. For example, content that has expiration of availability or distinct time-based usage patterns, retention would mean shorter relative timespans whereas services could have wider timespans.
http://www.clickdesk.com/ Sudarshan

Hi Andrew, great thoughts. I am skeptical about the word ‘virality’ – I think traction would be a more suitable word. I am in total agreement with your point on retention; if companies spent half the time and resources that they spend on acquiring a new customer on retaining an existing customer; they would real well. Back to virality – a lot of products don’t really have a inherent viral aspect, and mostly virality isn’t planned. Retaining existing customers is definitely a great ploy to generate more users.

@andrewchen