Mobile App Codeless Tagging for Analytics

Suggesting Features to Tag (Track in Analytics)
B2B
Mobile App Analytics

Team Members

Product Designer (Me)
Product Manager
2 Backend Developers
2 Front End Developers
Scrum Master
QA

Tools I Used

Pen & Paper
Whimsical
Figma
Dovetail

Scope of Work

Research, User Interviews, Wire framing,
Prototyping, Usability testing, UI Design.

Background

A code-less digital adoption and analytics platform that helps product managers understand user behavior across apps and provide the best experience by data and creating personalized in-app guides/tooltips.
Code-less analytics means collecting all data without having to rely on developers.
Tagging is the process of selecting features to track in order to collect analytics regarding those features.

The Challenge

The customers (not the end users) are spending too much time in tagging. The company's support team is troubled by dealing with too many calls. As a result, users don’t see value from analytics until they go through the tagging learning curve. Feature tagging on mobile is time-consuming and requires a lot of effort. We want to reduce the time and effort needed to complete tagging and move towards gaining value from analytics and guides.

Users are PMs, CS & Engineers

The users are the product managers, customer success, engineers and sometimes designers that want to learn how their users are using their apps.
They are pulling data from multiple sources to answer a complex question or uncover previously unknown information to inform business decisions.
In general they are not technical users.

Moderated Comparative Usability Tests

7 customers. 2 prototypes in 7 sessions. 320 minutes.

Research Plan

Hypothesis
- Basic Users get a benefit from selecting suggested tags.
- Advanced users get less benefit from selecting suggested tags
Goal
Users understand how to quickly select suggested tags.
User Segments
- Basic Users: Tagged at least five pages but no more than 30 pages in the last 30 days
- Advanced users: Tagged more than 30 pages in the last 30 days

These are the key research conclusions:
1. Suggested features saves customer's time as they tag multiple features on a page.
2. Customers tend to capture a page and tag its features before moving to the next page to make sure they haven't missed a feature.
3. Customers need an easy way to verify that suggestions are accurate before fully trusting them.
4. Overall trust in the tagging process was raised.
5. SUS was measured and proved a high usability score: 92.2

Pain Points

High Friction in The Tagging Process

The tagging process is too complex for non-technical users.

SDK and Mobile Frameworks issues

SDK and current mobile frameworks cause too many inconsistencies in customers apps resulting in problematic tagging.

The Solution

When the user tags a new mobile page, We will automatically suggest a list of features to auto-tag for the user. These features will be all the clickable elements on the screen with more than 0 click events in the past 24 hours.
At a later stage click tests will be done to establish value for high engagement vs high accuracy so that valuable tags will be suggested.

Mid-fidelity Designs

Started with pen and paper as always and moved to whimsical to not waste too much time on specific design.

Hi-Fidelity Design Flow

Refined designs

1. Suggested features that were tagged will appear in the 'Tagged Feature's list (under the 'Suggested Features to Tag'). Users will be able to edit feature rules to make sure the feature is unique.
2. Testing showed that there is a need for users to delete suggestions that are not valuable. Users can delete suggested features or collapse the suggestions list.
3. Once a suggestion is tagged, a quick way to name the feature and select a product area is presented, This raises a sense of control and thus confidence for users.

Next Step: Testing Valuable Tagging

Using click tests to establish value for high engagement vs high accuracy so that valuable tags will be suggested.