How ETS turned a team of one into an early warning system for five brands
Jessica Dillon (
Using AI-powered sentiment analysis and Tagging in the Smart Inbox, Jessica turned a fragmented, reactive care process into a system that uses Sprout Social to surface patterns early enough for other departments—not just social to act on them.
Turning individual conversations into company-wide signal
As customer inquiries and engagement volume grew, recurring patterns around at-home testing, score reporting, and registration provided valuable insights into the customer experience. By identifying and tracking these trends, ETS was able to leverage social data as an early signal of emerging needs and opportunities, ensuring that key insights were shared across teams and incorporated into broader conversations.
"The questions test takers asked most often helped shape new social community content and FAQ resources, allowing us (at ETS) to provide answers earlier, reduce uncertainty, and better support learners when they needed it most." — Jessica Dillon, Social Media Community Specialist, ETS
The how
Jessica built a system that turns high volume, unstructured conversations into decisions other teams could act on:
- Prioritizing risk with sentiment tagging: When a test taker continued to seek guidance regarding disability-related accommodations. AI-powered sentiment tagging surfaced the issue in real time, allowing Jessica to proactively engage the disability services team, validate next steps, and respond publicly with accuracy, empathy, and confidence.
- Spotting demand shifts before they became complaints: Real-time trend monitoring showed a 126.5% jump in "About the Test" inquiries and a 150% jump in testing location questions for TOEFL within a single fiscal quarter. Instead of waiting for repeat questions to pile up, the team built proactive FAQs and social guidance around what the data was already showing before the next quarter’s testing began.
- Maintaining response time as volume grew: Even as GRE message volume rose 64.1%, response times stayed within a 3-hour SLA – a consistent performance for ETS across every brand. Corporate response time improved from 2 hours to 1 hour 20 minutes, and TOEFL response times improved from 1 hour 26 minutes to 1 hour 7 minutes.
- Reading friction reduction as a success metric, not just a volume metric: As the team expanded proactive messaging and FAQ content, Praxis experienced a 93.3% decline in social sentiment categorized as unfavorable, while message volume fell 51.8%. TOEFL message volume also decreased 58.4%, reflecting the value of providing answers and resources early.
The business impact
Operating as a team of one, Jessica's approach scaled care across five brands without sacrificing speed or brand voice. GRE saw a 26% increase in positive sentiment alongside a 13.5% rise in support interactions. In parallel, LinkedIn direct messages climbed 52.9% for GRE and 11.8% for TOEFL, signs that customers increasingly see social as a primary channel for support. Declining volume and negative sentiment across other channels point to a system that's catching problems earlier in the customer journey, before they become recurring ones.
By monitoring conversations across five brands, Jessica transformed customer engagement into an early warning system, surfacing trends and questions in real time and helping ETS proactively address audience needs through social platforms.
What could this look like for your team?
If your care team is managing high volume with limited headcount, what would it take to turn every conversation into a signal the rest of your business can act on? We'd love to hear what's still getting handled one ticket at a time that could be surfaced as a pattern instead.