Architected Voyager, a company-level product-intelligence framework that correlates customer voice, usage, and releases on demand to trace customer pain points and better understand the business impact behind them.

Overview

The Voyager Framework is a product intelligence framework that correlates customer voice, usage metrics, product releases, and business context on demand to surface the impact we have on our customers and business, so we can answer questions like:

• "Person X hits problem Y, here is the evidence, therefore Z suffers."
• "Customers are loud about X, usage confirms it by Y, and the accounts asking are worth Z."
• "We shipped X, it shifted usage by Y, which moved the business by Z."

The problem

LaunchDarkly has a wealth of data to help product teams make informed decisions and grow the business. The problem is that querying and correlating it to identify pain points and uncover themes is difficult and time-consuming. Customer voice, business metrics, feature releases, and user behavior each live in a separate platform with its own data model.

The cost of that friction:

• Product teams query only in the beginning phases of a new project
• Metrics get used infrequently, and decisions get made on assumptions
• Improving the UX becomes reactive instead of proactive
• Data accuracy and trustworthiness stay unknown because instrumentation is unverified
• Designers build their own custom, ad-hoc metric-collection tools to fill the gap

The approach

I built a small proof of concept to validate the product team's need to stay current on usage behavior. This was an Experimentation Weekly Digest that pulled usage and voice into one cited Slack report, and the pilot returned two insights I built the framework on. The short executive summary posted to Slack got read and valued by the team. The longer weekly report ran too detailed to hold attention, and it correlated two of four signals, so it could not yet connect customer voice to what we shipped. Both findings became design inputs: the executive summary proved the format, the report's depth marked the opportunity to tighten.

After reviewing the pilot's results with product teams and leadership, the VP of Product Operations and the Head of UX sponsored the project. The Voyager Product Intelligence Framework had an official kickoff, and as the lead my next steps were clear:

• Wrote a product requirements document to set the project's scope and expectations
• Made the data-modeling calls and built a reusable lookup table to query multiple sources efficiently
• Interviewed product teams and leadership to learn their use cases and expectations
• Worked with designers to consolidate their research tooling into Voyager
• Expanded the set of data sources beyond the initial scope
• Formatted the response template around executive-level summaries with cited sources

The story continues

This is an active project, improving as it runs, and product teams are already finding value in it. Voyager's entry points are expanding beyond a single chat agent. We are building success-metric automations for senior executive review meetings, product-team dashboards for on-demand insights, and in-context insights through a Chrome plugin.