How Chamath Palihapitiya Dramatically Increased Facebook User Retention

"7 friends in 10 days", The most well known "AHA" moment on the face of this earth. This is the story of how they got to this insight. The "AHA" moments you come up with your gut feeling are not real "AHA" moments. We know people cannot predict correctly, but data certainly can.

If not for the gut feeling, how did he do it? and why? Let's dive into data for some insights!

Let's start with the why. Why is retention important?

If you have poor retention nothing else matters! This certainly is something Chamath and his team truly believed from early on. For a platform that relied on user generated content for engagement, retention meant everything.

Higher Retention means

• Increase Customer Life Time Value --> Ability to allocate higher budgets for Acquisition
• Increase Promoters(Virality) --> Decrease the cost per acquisition (Your customer will do your marketing for you. Yay!)
• Increase up selling opportunities --> Grow revenue, cut down pay back period

It's not just one of these that will come to you with higher retention, but all of them.

How did he get to 7 friends in 10 days insight?

Let's take a look at the retention curve. I do not have access to facebook data, however the below diagram should be sufficient to get an idea of what a retention curve is and how it can be used to drive insights.

The X-axis represents the number of days since the date of joining. For example 10 on the X axis means that the customer has been using the platform for 10 days. The Y axis represents the number of active users. For example, if you look at the Red curve, it has about 30 new users and 20 users have dropped out within the first 10 days.

This is a consolidated curve. There can be people who join in January, February, march or any given month. These curves can be mapped out individually on this graph. However to easily analyze we consolidate all these months and create a single curve by adding these curves together. For example 100 new customers in January, and 200 customers in February. When consolidated this comes to 300 new customers on the 0th day.

Identifying all possibilities

It is common to look at user source, entry path, technology, content type and persona for segmentation. For the sake of simplicity, let's assume that the only variables/possibilities are the ones given below.

• Channel - Search, Direct, Mailer, Referral
• Actions - Updated Photo, Updated Bio, Added 1 friend, Added 2 friends & so on
• Time - within an hour, within a day, within 5 days, within 10 days & so on

We then create a combination and segment them together. These are called cohorts.

Let's consider the cohorts with the following combination of steps.

Cohort 1

Within a day, persona X

1. Joined through search
2. updated photo
3. updated bio
4. added 1 friend

Cohort 2

Within 10 days, Persona Y

1. Joined through referral
2. added 2 friends

Cohort 3

Within 10 days, Persona Z

1. Joined through mailer
2. updated photo
3. updated bio
4. added 7 friends

It is essential that we look at every possible combination available with all the possibilities. Once you have created cohorts for each possibility you can now map the retention curve for each of these cohorts.

Which will look something like the diagram given below

Each color represents a different cohort. Each of these cohorts has a different retention curve.

Understanding the Retention Curve

Let's start with the green curve.As you can see the curve declines continuously to zero. Which means that all of your hard earned users will be lost. You might probably have a product market fit problem. (This is entirely another area that I plan to cover in one of my future blog posts, stay tuned )

The blue curve, on the other hand, looks more promising. Certainly, you have lost a few users over time but you have flattened the curve after a certain number of days ensuring that they remain for a longer period of time.

FLAT CURVES ARE THE BEST RETENTION CURVES

Chamath and his team understood this and the curve that had the best retention was the curve of the cohort group that has made 7 friends within 10 days. They actioned this insight by recommending new users relevant friends with all the possible sources. This insight and the actions they took put them on the road to become the most successful social media platform the world has ever seen to this day

3 Simple things. Measure, Test, Try ~ Chamath Palihapitiya

----------------------------------------------------------------------------------------------------------

ReapDigital Learnings

Insight

A leading Apparel brand was struggling to retain their online customers, repeat customer rate was significantly low. We created cohorts by product type (Shirts, Pants, etc), and product size (S, M, L, XL, XXL). We learned that those who buy XL and XXL tend to have higher retention, and does repeat purchases more than all other cohorts.

Action

Special discount coupons were given to 1st-time buyers who purchased any XL or XXL products making sure that they repurchased and got into the habit of purchasing. We also saw a tremendous improvement in retention due to coupon codes and rolled out coupon codes to all 1st-time buyers. This resulted in 12% upwards shift in the retention curve.

----------------------------------------------------------------------------------------------------------

Insights are great but clearly useless without knowing how to action them. Let's take a look at how we can use insights we learned to drive growth

How to action your insights to drive higher retention

What part of retention do we need to optimize?

This depends on who your customer is and what your end objective is. The diagram below shows the four main parts of retention and different tactics that can be used to optimize different parts.

New User Experience

New user experience is the natural behavior of the product.It is important to understand how do people use your product. Chamath and his team realized that if you make 7 friends within the first 10 days then the curve will stop declining and will become flat. They pulled together all their resources to make sure users get to experience the core value of Facebook as quickly as possible.

"Tactics to impact this part of retention will help you shift your retention curve up making sure more users will retain in the long run."

• Growth Hacking Tactic: If you remember when dropbox launched, they had two simple options on the home page. Download the product or watch the video. If you already downloaded the product it gave you 250mb of free storage when you watched the "What is dropbox" the demo video. Making sure everyone watched the video and understood what Dropbox exactly was. That is their growth team getting you to experience the core value as quickly as possible. Using a simple video explaining what your core value addition to the user can help reduce the drop-off and reach a flatter curve.

Mid-Term Retention

Once you have allowed your new users to experience the core value it is important to create habits around the core value. By having a multiple number of habits that triggers our emotions will likely drive you to be very engaged and involved with the platform/product.

• Growth Hacking Tactic:"You've just been tagged in a photo on Facebook" very familiar words? Did it excite you? Do you want to know WHO tagged you, and when did you take it? where exactly? you do have a million questions and our inquisitive selves cannot wait to check this out. We log in to check it, like it, comment on it, and if its a great picture might even share it. Before you know it, you cannot help yourself browsing through your distant cousin's tagged photos and tagging your family. In a nutshell, stimulate user emotions to build habits

Long Term Retention

Once experiencing the core value has been turned into a habit, it is important to make sure the users experience the core value as often as possible. Making the curve flat and making sure it stays flat without declining can be ensured through delivering an excellent experience.

• Growth Hacking Tactic: How many exciting projects have you got involved in through GitHub? Quite a few I suppose? How often have you had help from the community, or discussed an idea or contributed in the git community? Creating a community that is vibrant and relevant to the product itself can guarantee long-term engagement and retention.

Reactivation

There must have been some reason a user joined your service. If they didn't understand the core value right away the chances are that they have become dormant. Remind them and reform their opinion. Resurrecting existing users are cheaper than acquiring new ones. Create triggers for them to start experiencing your product again.

• Growth Hacking Tactic: " We Missed You" emails are becoming largely popular and no doubt you have received a few yourself. Dropbox sends a very simple email with a before and after illustration, which also speaks about the core value of dropbox. Remind your users about the core value of the product and incentivise to get started with the product again

Understanding what part of retention you want to impact will be clear once you have defined your objectives. Creating a measurement framework will help you ensure that you have reached your set objectives.

“Your diamonds are not in far-away mountains or in distant seas; they are in your own back yard if you will but dig for them. – Russell Conwell 1862”

Understanding retention is pure joy, triggers all emotions of the data geek inside you. Let me know how this turns out, stay tuned for more on Malinda.xyz