The virality of LinkedIn Endorsements

Musings of a Product Guy

One of the most fun products I have had the privilege to create and launch is LinkedIn Endorsements. As a result, the most common question I get asked by entrepreneurs is: “So how do I get an Endorsements for my product?”

There are actually some great posts out there about how to model virality (I like this one from my friend Rahul Vohra, Founder of Rapportive: How to Model Viral Growth).

However, I thought it would be useful to analyze the virality of Endorsements from a product viewpoint and share the learnings from it.

Endorsements viral loop

Let’s first walk through the Endorsements viral loop below:

1. User A endorses user B, potentially via a suggestion. At this point A is also prompted with suggestions of other connections to endorse
2. B is notified that he/she has been endorsed for a skill
3. B comes to LinkedIn to “accept” the endorsement (B only needs to approve an endorsement for a skill not on his profile)
4. B is prompted with suggestions of other connections to endorse
5. And the cycle continues…

So what worked? How did Endorsements become so viral so quickly?

1. Before even starting the flow, the first key insight came from the fact that there were limited actions you could take when viewing a connection’s profile - Message, Recommendation, Suggest Connection - actions that were not as popular as say Connect on a non-connection’s profile. But we quickly realized that endorse could be that key action.
2. Most of the Endorsements came from flow C in the viral loop above - from a member who was just endorsed. This is a very valuable insight - users who just had something good happen to them (in this case being endorsed) were very likely to continue that action and pay it forward.
3. Similarly, the next most endorsements came from flow B - a user who has just made an Endorsement. If a user has already performed an action, he/she is very likely to continue performing that action.
4. Having endorse be one-click was extremely important to its success. Anytime you can limit the number of steps to perform an action, you will likely see an increase in engagement. We saw a similar bump in invitations when we reduced the number of clicks to invite someone from People You May Know from 2 to 1. And this philosophy is even more true on mobile, where you really want to limit the amount of typing your user has to do.
5. Suggestions were an important factor to having Endorsements be a one-click action. Even just having suggestions was a powerful mechanism, as can be seen by the fact that the vast majority of Endorsements came from suggestions.
6. Getting the product into major flows - in this case viewing a connection’s profile and viewing one’s own profile - was key. Discovery of a new product feature in an existing major contextual flow is vital to its success.
7. As a viral, transactional engagement feature, Endorsements were compounding. The more we could get on Day 1, the more we had on Day 2 because many of the the users who were endorsed were endorsing their connections. We learned the importance of this quickly when we did not send email notifications on a weekend because weekend CTRs were lower. However, the higher CTR benefit from delaying the notifications was far cancelled out by losing the compounding Endorsements from Saturday and Sunday.
8. And finally, Endorsements are fun and feel good! And on a site like LinkedIn that often has a more professional feel, users really liked this fun, one-click way to help their connections.

How to Endors-ify your products

So, if you want to achieve Endorsements-level virality for your product, here are the 3 main lessons that you can apply:

1. Make the number of steps necessary to take an action as few and as lightweight as possible.
2. Any chance you have to have suggestions in your product, use them! This was crucial to the success of viral features like Endorsements and People You May Know.
3. Draw out your viral loop and determine the conversion rate at all part of the flow. Then, when you see which part of the loop has the highest rate, double-down on it and keep that viral engine churning!

Endorsements was a fun product to work on and rewarding to see members get jobs from the feature! I also learned a lot from the product — a subject for a future blog post.

Feel free to reach out with any questions to me on Twitter, LinkedIn, or email.

And good luck Endors-ifying your products!