Case Study

Scaling self-service at Schwab: Answer Widgets

How a content-led, data-driven approach scaled a pilot program from 50 to 557 Answer Widgets — delivered six weeks ahead of schedule and expanded 605,000 times in its first year.

The problem

Automation stalled. The library needed a human system.

When I joined Schwab in spring of 2021, Answer Widgets, contextual FAQ surfaces embedded in the site's search results page, had been live for less than a year. The pilot had produced 50 widgets. The goal was to scale to 500 by year's end through an automated ingestion process that would pull FAQs from across the ecosystem.

The automation didn't work. The process couldn't reliably identify, filter, or format content at the quality level the experience required. Someone needed to build the library by hand — and build the system that would make it maintainable going forward.

The challenge wasn't just volume. It was creating a scalable taxonomy and keyword architecture that would allow the right answer to surface for the right search query, at a moment when users were trying to act, not browse.

The strategy

Build the master list. Get the keywords right.

I started by pulling a full content report from the Drupal team: every page across both the client and prospect experiences. Filtering by accordion component type (the CMS pattern used for FAQs) gave me a rough inventory. From there, I culled the list to content that could realistically function as an Answer Widget: clear prompt, actionable answer, discrete topic.

The result was a master list organized by experience area — trade, wire, beneficiary, taxes, and more — that gave the team a single source of truth for the library. Keywords were the next critical layer. Using Adobe Analytics search data, I identified the actual terms users were entering for each topic, then mapped those phrases to the corresponding widgets. Getting this right was the difference between a widget that surfaces and one that doesn't.

The process

Batched sprints. Rigorous QA. Consistent velocity.

With the master list complete, I designed a delivery workflow: batches of 50 widgets per biweekly sprint, moving from content build to development handoff to QA. My role shifted from creator to quality control at the back end of each sprint, reviewing each new experience to confirm keywords mapped correctly and that no formatting or grammatical errors had been introduced in development.

The batching approach kept the work manageable, and the QA step ensured every widget that went live met the same standard as the first.

Outcomes

557 widgets. 605K expansions. Six weeks early.

The final batch published in mid-November 2021, six weeks ahead of the end-of-year deadline. Users accessed 557 widgets across 101 distinct topic areas, with steady month-over-month growth in widget views as the library scaled. By February 2022, monthly widget views had reached over 100,000, more than triple the volume from earlier in the year.

What came next

Maintenance, partnerships, and continuous improvement

Following launch, the work shifted to stewardship. The full library was audited twice to address issues introduced by a site-wide design system update. New widgets were developed in partnership with business stakeholders to support seasonal moments (tax season) and product launches (Thematic Investing). The longer-term roadmap included re-initiating automation, integrating the Schwab Assistant into search, and adding an in-experience feedback mechanism to surface content gaps from user behavior.