Recommend templates based on customer behavior

Recommend templates based on customer behavior

Recommend templates based on customer behavior

Background

Atlassian had recently developed a Confluence Growth Program dedicated to improving the end-to-end experience for new customers and increasing paid activations.

One opportunity our team identified early on was a positive correlation between customers engaging with Templates and down-stream conversion. Previously our group had experimented with helping customers explore Templates on Atlassian's Web Platform - but saw an opportunity to connect customers with Templates as part of onboarding.

Role

Design Lead

Category

Onboarding

Timeline

3 weeks

Background

Atlassian had recently developed a Confluence Growth Program dedicated to improving the end-to-end experience for new customers and increasing paid activations.

One opportunity our team identified early on was a positive correlation between customers engaging with Templates and down-stream conversion. Previously our group had experimented with helping customers explore Templates on Atlassian's Web Platform - but saw an opportunity to connect customers with Templates as part of onboarding.

Role

Design Lead

Category

Onboarding

Timeline

3 weeks

Background

Atlassian had recently developed a Confluence Growth Program dedicated to improving the end-to-end experience for new customers and increasing paid activations.

One opportunity our team identified early on was a positive correlation between customers engaging with Templates and down-stream conversion. Previously our group had experimented with helping customers explore Templates on Atlassian's Web Platform - but saw an opportunity to connect customers with Templates as part of onboarding.

Role

Design Lead

Category

Onboarding

Timeline

3 weeks

challenge

first time Confluence customers struggled recognizing value while onboarding and staying active in the product after day-0

Why this mattered

Behaviors to support

Behaviors to support

Behaviors to support

  • completion of setup and onboarding flows

  • post-onboard template exploration and utilization

  • creation of Confluence pages with Templates

Limiting factors

Limiting factors

Limiting factors

  • the onboarding and setup flows ask intent-based questions, but don't offer customized experiences based on prompts

  • three different business groups, codebases, and libraries had domain or influence over the onboarding and setup flow

  • when customers complete set, the initially land on a distracting and unexpected screen, 'Confluence Home'

North star metric

North star metric

Track customer engagement through the onboard/setup flow and monitor if customers used the recommended templates, and had stronger down-funnel activations as a result.

our hypothesis:

IF customers are able to select a template to start with during Set Up based on their JTBD and role type and land in that template in Editor



THEN more users will better understand the value of Confluence faster and drive d1-d6



BECAUSE users will get into Confluence faster, engage with a contextually relevant template that will help them realize value faster and convert to d1-d6 at a higher rate

IF customers are able to select a template to start with during Set Up based on their JTBD and role type and land in that template in Editor



THEN more users will better understand the value of Confluence faster and drive d1-d6



BECAUSE users will get into Confluence faster, engage with a contextually relevant template that will help them realize value faster and convert to d1-d6 at a higher rate

my impact

Responsible for:

  • End-to-end experiment design

  • Sparring work and connecting outputs with leadership and stakeholder teams

  • Partnering with engineering to map logic scenarios

team + stakeholders

To complete this work, I partnered and worked within the FIND Pod. Due to this experiment touching parts of the Signup and Onboarding flows, we developed a close partnership with the Confluence in-product team that 'owned' this part of the flow. Through the course of this partnership, we sparred and made numerous tradeoffs.

Core team composition:

  • Product Management

  • Engineering

  • Content Design

  • Product Design 👋🏼

Our Audience

First-time Confluence users

First-time Confluence users

First-time users struggled with getting started with Confluence.

After signing up, customers are dropped into a blank home screen populated with placeholder documents and pages. Often these placeholder materials were non-industry or use-case specific. Coupling this with a confusing internal architecture resulted in many customers abandoning Confluence Home and not returning

First-time users struggled with getting started with Confluence.

After signing up, customers are dropped into a blank home screen populated with placeholder documents and pages. Often these placeholder materials were non-industry or use-case specific. Coupling this with a confusing internal architecture resulted in many customers abandoning Confluence Home and not returning

This is great! I can download this template from Confluence and import this into Microsoft Word so that my team can leverage this.

Customer Interview

Highlighting Confluence's perception challenges

delivery

a summarization of the process, effort, and output

delivery

a summarization of the process, effort, and output

We made changes to the Confluence onboarding flow in two critical areas: the flow and destination.

The Flow

After selecting JTBD and Role within SETUP, our team created a new dynamic step focused on the Template experience. Based on user input, this new screen would recommend a template for customers to get started with.


The Destination

After SETUP, customers traditionally landed in Confluence Home. To help customers better connect with the tooling, we decided to route customers directly to the Page Editor, with their selected template pre-loaded.


View in Figma →

impact

+0%

increase to setup completion rate

-0%

reduction in time-to-provision

0%

uptick in customer activations

0

experiment variants

0

personalized templates per onboarding variant

0

permuations of role vs industry

impact

+0%

increase to setup completion rate

-0%

reduction in time-to-provision

0%

uptick in customer activations

0

experiment variants

0

personalized templates per onboarding variant

0

permuations of role vs industry

impact

+0%

increase to setup completion rate

-0%

reduction in time-to-provision

0%

uptick in customer activations

0

experiment variants

0

personalized templates per onboarding variant

0

permuations of role vs industry