Company Overview

Chegg is a digital learning platform that helps college students succeed academically. Known initially for textbook rentals, Chegg now offers 24/7 study support, including practice problems, tutoring, and expert Q&A. Behind the scenes, the support team fields complex student interactions, and now uses that data to shape more than just service outcomes.

The Challenge

Chegg’s support organization was handling massive volumes of student conversations—but the information in those interactions wasn’t being heard by the business. Product teams were outsourcing expensive surveys, executives asked for insights that support couldn’t provide, and metrics like CSAT didn’t reflect the full story.

“My peers kept asking what customers liked or didn’t like about the product. I couldn’t answer that—because our tools weren’t built to share what our students were really telling us.” - Diana Landis, Sr. Manager of Insights and Knowledge

Even as internal scorecards improved, Chegg’s team knew they weren’t unlocking the strategic value of their data.

The Turning Point

Replacing CSAT with conversation-based insight

For years, CSAT had been the default metric guiding experience decisions. But it wasn’t accurate—and Chegg knew it.

• Survey response rates were low and skewed
• Frustration after resolution wasn’t captured
• CSAT trends failed to reflect known product issues
• Student behavior (young, fast-moving, disengaged with surveys) made results noisy and misleading

The Chegg team realized CSAT was holding them back from understanding true sentiment.

They set a bold goal: eliminate CSAT altogether—and replace it with a smarter signal derived directly from conversation data.

The Solution

A new signal: NISPR (Negative Sentiment Post Resolution)

With MaestroQA’s LLM-powered analysis, the team built NISPR: a signal that flags conversations where negative sentiment appears after a resolution is delivered.

This targeted lens eliminated false positives from traditional sentiment tools, which often flagged conversations as “negative” just because students were reaching out about a problem.

What changed:

• Focus shifted from agent behavior to full interaction journey
• Coaching became more precise—targeting tone and delivery
• Strategic discussions with execs shifted to outcomes, not surveys

“It wasn’t about whether a student was upset when they contacted us,” Nate said. “It was about whether they were still upset after we gave them an answer.”

Impact

This wasn’t a top-down transformation. It was a team of operators, analysts, and leaders pushing together to show that conversation data isn’t just useful. It’s essential to building better products, smarter decisions, and a more connected business.

Influencing Product Strategy

Conversation data becomes the voice of the student

Chegg’s insights team used LLM classifiers in MaestroQA to identify trends like pricing confusion, bot hallucinations, and refund friction—issues that had real product implications.

But getting product teams to listen wasn’t easy. Product stakeholders were skeptical of “support data,” viewing it as too anecdotal or too negative.

The insights team changed their approach:

• Sat in sprint planning meetings to hear product priorities
• Surfaced specific examples tied to upcoming launches
• Delivered journey-level context: what students saw, felt, and did

Eventually, product teams began listening—and acting on the findings.

“We brought stories, not just stats. We asked: Do you want any student experiencing this?” said Diana.

Chegg restructured so that support reported directly into product—a testament to how influential these insights had become.

Surviving Layoffs by Proving Strategic Value

During company-wide restructuring, the support team was asked to quantify the impact of potential staffing cuts.

Thanks to the NISPR signal in MaestroQA, they could model exactly how quality would change under different scenarios—and prove the risk to customer satisfaction, brand, and churn.

That data made conversation insights indispensable.

“As painful as it was, I could say, ‘If you reduce this team, here’s how much worse the experience gets,’” said VP of Support Kelly Cutforth. “That’s why we’re still here.”

“You don’t need a bigger team. You need to ask better questions, and use the tools you have to answer them.”
 — Kelly Cutforth, VP of Customer Support, Chegg

“The business wasn’t asking us for these insights. But we brought the data anyway—and proved it mattered.”
 — Diana Landis, Sr. Manager of Insights and Knowledge, Chegg