AI that reads hundreds of reviews for you
Feedback about the company comes in from eight different online channels. Previously an employee went round them by hand every week and copied essentially just the star count into Excel - there was no time to read the text. So the company knew its average score but had no idea what lay behind the ratings. We built a mini-app that pulls everything into one place and automatically extracts both the praise and the concrete problems to solve.
online channels merged into one place.
reviews analysed in the initial pass.
positive captured and sorted by theme.
new responses are added automatically and praise and problems categorised.
How it works
The challenge
Feedback about the company comes in from eight different online channels. Before the app existed, an employee had to go round them by hand every week - opening them one by one and copying essentially just the star count into Excel. Noting down the text as well and working out what customers were really writing about was beyond what they could manage. So the company knew its average score and how it was trending, but had no idea what lay behind the ratings - and that turned out to hold plenty of surprises.
The solution
We pulled all the material from the eight channels into one place and built an AI analysis on top of it. It runs automatically: every day new sources are added, responses are categorised, and the concrete praise and problems customers mention are extracted. An initial batch of roughly two thousand reviews showed where the company stood - since then dozens of responses accrue every month on their own.
What the company gets
For the first time the company sees what's really hidden in the feedback - not just stars, but themes. Praise can be attributed to the specific people it concerns, where that can be traced, and they get the recognition. The problems customers flag become a basis for fixing internal processes - the team knows what to improve from them. And because the analysis runs daily, the trend can be tracked continuously.
Instead of "we feel like our customers are happy" we now know exactly what bothers them and how many.- a typical benefit of the solution for the company
What else we've built
AI customer analysis from the web and social media
Watches public sources and social media, extracting signals about customers and opportunities.
Read more →
Invoice processing from e-mail
Extracts an invoice from an incoming e-mail and files it into accounting. No manual retyping.
Read more →
Meeting scheduling for salespeople
Calculates travel between meetings and shows the call centre every salesperson's calendar in one place. No needless driving.
Read more →Want a similar solution at your company?
Start with an audit. We'll find where AI makes the most sense in your processes.