Automatically segment communities to retain influencers

About Tribe Dynamics

The most comprehensive influencer marketing analytics platform that helps brand’s understand their earned media program’s impact and improve their digital marketing strategy.

The mission, is to help iconic brands harness the power of influencer marketing with best-in-class software and service.

The customers, are companies in passion-driven verticals with large online communities of ambassadors i.e. social media influencers.

Jobs our Software is hired to solve

Discover & Evaluate your influencer community

Capture & Organize your social media content

Measure & Analyze your influencer program

Project background

Business objectives

Increase Customer Retention

Increase Annual Revenue

Product objective

Similar to a customer marketing funnel, successful brands retain ambassadors and prevent attrition. Novice brands continuously acquire new ambassadors, which requires a significant upfront investment but may yield limited long-term loyalty.

I hypothesized that if our software helped customers create a strong ambassador retention flywheel, it would improve the effectiveness of the brand's influencer program and ultimately lead to higher customer retention for Tribe Dynamics.

Research

In order to test my hypothesis of that helping brands retain Ambassadors would lead to more successful outcomes, I conduced user research and leveraged product analytics.

Methodology

User Interviews

Target research segment

• Experts - Daily users who were tracking at least 800 ambassadors profiles with high MoM retention
• Novices - Daily users who added 80+ new ambassadors profiles with low MoM ambassador retention

Continuous Synthesis - Completed 8 live discovery sessions, with summary snapshots, and experience maps.

When I asked users about how they select their influencer campaign participants…

“If I have a campaign with 100 participants, I just add the top 100 by engagement so it always includes these influencers who we had a paid partnership with.” - Novice community manager

“I export this list so I can sort on last post date, to review who has been posting a lot in the last 3 months as well as who has dropped off in the past 6-months so that we need to re-enage.” - Expert community manager

Research Synthesis

Pattern Recognition - Created Journey Map to understand the distinct moments and identified key Opportunities (pain points, wants, needs) from user interviews

Mapping relationships - Mapped pain points on to Opportunity Solution Tree to visualize parent-child relationships of user pain points

Key Research Findings

I conducted a post-launch feature analysis for the “6-month posting heat map”, with the following key findings:

1. Quantitative research: User clicks on “6-month posting heat map” header to sort table, which was not supported.
2. Qualitative research: Default sorting of the Ambassadors table, led community managers to always select the top EMV earners, without considering recency of posts.

Define

Prioritized Problem

How we prioritized

Prioritization Framework

Opportunity Sizing

Market Factors

Company Factors

Customer Factors

Personas

With personas already defined, we decided to focus our solution ideation on two specific ones in order to specify the product requirements.

Must be…

• Elevated on page hierarchy
• Simple to use
• Easy to get additional help resources

Must be…

• Efficient to use
• Performant in the app

Solution

How might we…

…help community managers quickly identify which ambassadors to engage with based of the influencers’ posting behavior.

Considerations

• Must be simple, intuitive, and efficient
• Make use of Material Design components
• Ambassadors can be identified and added to a campaign

Ideation

Which Ambassadors → Community Segments

Based on our user research, the relevant posting window as 6 months and decided to compare the most recent 3 months of posting behavior to the 3 months prior.

This resulted in 4 simple community segments.

Testing Assumptions

To de-risk this decision, we worked with engineering to produce these 4 groups based on real customer data. We delivered the information to 6 customers via excel spreadsheets to get feedback on the methodology.

We received the following feedback:

• Names of groups did not match language used by most customers
• 5/6 customers agreed that the methodology was appropriate and matched their mental models

Rapid Iterations through Alignmnet

As a result of the feedback, we updated:

• The group names and descriptions:
• Loyal Fans → Retained Fans
• Recent Fans → New Fans
• Past Fans → Lost Fans
• Potential Fans → Potential Fans
• Aligned on segment descriptions to spell out the exact time ranges we use to automatically-segment ambassadors

Quickly Identify → Community Groups on Ambassadors Page

Based on our research, users were building their activation lists on the Ambassadors page by applying various filters and sorting the table.

Therefore, we decided to add component between the filters and table to display the community segments.

Testing Assumptions

Initially our idea was to add saved filters for the 4 segments that could be clicked to narrow down the ambassador table.

We received the following user feedback on initial sketches:

• Users were confused what the exact segment definitions were
• Users wanted data points to measure progress
• Users wanted to clearly see which groups were on versus off

Rapid Iteration through Sketching

After receiving feedback, we decided to emphasize the component by adding descriptive information to guide the user. We also included a clear toggle to give the user control and indicate when a segment was included in the Ambassador list below.

High-Fidelity Design

UX and UI decisions

I partnered with our only-designer on the team to turn wireframes and pre-selected Material Design components into a high fidelity design.

In order to help users, especially those that fit the “no natural knack” persona, I wrote copy for each of the 4 segments.

Prototype

Tribe Dynamics - Ambassador Community Groups

Last modified by valerie pitsch ⋅ a year ago

Usability tests to de-risk solution

In order to de-risk the solution we tested the methodology as well as the toggle interaction for filtering in or out a respective community segment.

Findings

• 6 out of 6 customers successfully narrowed down the Ambassador list to only the “Retained” segment
• 6 out of 6 customers strongly agreed that it was easy to find influencers to re-engage

Hand-off and Development

I prepared a Product Pitch with detailed information about the problem and solution requirements. I presented this pitch to Product and Engineering leadership during our betting meeting. After the project was green-lit, I hosted a kick-off meeting to hand-off the detailed specifications.

Once development was underway, we conducted several rounds of QA to ensure the intended solution was implemented as expected to reduce potential bugs.

Outcome

Measuring Success

Next Steps

• Reusable design components - Significantly reduce time for design and QA
• Testing a solution before building it - Developing the methodology and using excel to validate the direction proved to be efficient and cost effective
• Continuous customer feedback - We kept involving a group of customers to get their feedback, which created a strong bond between the product team and the customers.

Future Iterations

• Flexible date ranges - Some users wanted alternate time ranges to the opinionated 6-months.
• Customizing segments - More advanced customers wanted to build additional segments, which was out of scope for our design and development appetite.