Case Study Deskcasestudydesk.com
Product ResearchSourced

Case Study: How Dovetail improved WER by 36% with AssemblyAI to power continuous customer intelligence

Dovetail Case StudySourced & dated by Case Study Desk
Key facts · TL;DR
Company
Dovetail
Industry
Product Research
Challenge
Messy, multi-accent customer conversations demand precise, diarized transcription.
Headline result
Dovetail selected AssemblyAI as its preferred transcription provider after an accuracy comparison

Key results

36%
Improvement in word error rate (WER)
vs. prior provider

The challenge

Dovetail's platform ingests feedback from interviews, support calls, and product reviews and turns it into shareable insights. That source audio is often full of overlapping dialogue, informal phrasing, and unique accents, so transcription accuracy is critical to the value Dovetail delivers to its customers.

The solution

Dovetail ran an in-depth experiment comparing transcription providers and adopted AssemblyAI as its preferred provider, valuing precise, diarized, context-aware transcription and the ability to fold usage into its rolled-up cloud costs.

The accuracy of transcription is critical to the value we deliver to our customers. Humans communicate in a very messy fashion.

DR
Doug Rathbone
VP of Engineering, Dovetail

The results, in context

The partnership delivered a 36% improvement in word error rate (WER), improving the quality of the insights Dovetail surfaces for continuous customer intelligence.

Products used

AssemblyAI Speech-to-Text APIAssemblyAI Speaker Diarization