Koinz viral loop

2021 — RESEARCH, EXPERIMENT, ITERATIONS

Koinz viral loop

Saudi Arabia , Egypt

OVERVIEW

Koinz is a food aggregator mobile app that operates in Saudi Arabia & Egypt. It serves +1M users. As part of Koinz growth we needed build a continues growth tool.

DURATION

Oct 2021 - Jun 2022

MY ROLE

Product Designer

TEAM MEMBERS

Your device is too small to view this page

Case studies are better be viewed on a Desktop or a Tablet. If you think you shouldn't see this message please let me know to fix this issue

Report it to Omar

Your device is too small to view this page

Case studies are better be viewed on a Desktop or a Tablet. If you think you shouldn't see this message please let me know to fix this issue

Report it to Omar

Problem

Koinz growth team have multiple tools but non of them were depending on user directly. They wanted a tool that would generate a continues growth for the number of app users.

Product Managers conducted there research that later found that viral loops would best suit the case

Understanding

What are viral loops? How do we measure it?

DESK RESEARCH

Viral loop is a mechanism that being used to continuously generate users through referrals.

They tend to work through 4 stages:

Measuring a viral loop success happens through a metric known as Viral Coefficient:

Other metric is Viral cycle time. Which basically the time it takes referral receiver to See, Install, and develop a Desire to share then Share.

Applying to Koinz

What we have? What to build? What to enhance?

To understand our current resources we assessed them based on viral loop stages:

= Missing

= Need enhancements

= Working

Problems breakdown:

A) Implementing a referral program

B) App listing and Onboarding enhancements for receivers

C) Enhance Activation

Work setup:

As Koinz is working through experimentation. Each experiment get % of users to apply to it.

Product managers managed getting 10% of Koinz app users ready for the experiment.

Experiments method in Koinz depend on setting number of metrics then release, learn, and re-iterate until it’s decided to fully release the experiment as a feature or kill it for good.

Referral program

Starting the first iteration of referral program as the missing link in the chain. We collaborated with Product managers to agree on main points that would help to release and test it.

Where to implement the ref. program in the flow?

We conducted a market research that showed results about other tools/apps implementing their referral program after a happy case closing ( Happy rating, Happy review... etc )

As we already have a restaurant review after user receives their order. We thought it’d better lay their after the good review.

What would be the value to share? How we can avoid high acquisition cost?

Koinz already have two valuable rewards to share: Gifts or Points.

Points are redeemable as gifts too

We decided to go with Points as they have more potential in spending them.

To avoid high acquisition cost product managers decided to run this experiment through specific restaurants that agree to fund these points from their loyalty program with Koinz

Scarcity

To increase the chance of installing we agreed to implement a time limit for redeeming points before they expire. Starting in 1st iteration with 1 week.

1st Iteration flow and mockups and results:

Referral announcement

Choose contact

63%

Customize message

82%

Send SMS

41%

Social sharing

There’s an acceptable conversion happens from Referral announcement to Choose contact
Significant drop on “Send” button tapping
Very low usage of social sharing
Low conversion in receivers

Learnings

After contacting users through phone calls to understand more about how they deal with the feature we ran an internal usability test and through these two steps we detected some problems:

Users didn’t realize the value they’d share
There was a bug in sending message textbox
Send button wasn’t clear to the users
Users didn’t get that Koinz send an sms to the receiver then we encourage them to share on social app with the very same receiver. In fact they didn’t get the whole idea of multiple sharing ways

2nd Iteration

Through the learnings we got from 1st iteration we were able to agree and conduct some changes:

Keep referral program in the same place in the flow
Change visual hierarchy & Presentation to make value realizable

Fix message textbox bug and try a more popular send button pattern ( Made it open by default )

Depend on intended latency to help absorbing the info of SMS and Social sharing

Initiate visual linking through one illustration and Add situational entry point

2nd Iteration results

Iteration came with good results as there was an increase in overall conversion of the experience. Yet, increasing in receivers conversion to signup wasn’t as needed
Calls with users showed clearer understanding for social sharing

The viral loop team decided to move to enhance next stages in the funnel. Our next should touch all other points: