I like how Permut's conversational‑style feedback collection completely shifts the quality of responses we get from retail shoppers compared to rigid, traditional survey tools like QuestionPro. Customers tend to mindlessly click through checkbox‑heavy surveys and leave shallow, one‑word answers, but Permut’s autonomous agents carry out natural back‑and‑forth exchanges, encouraging shoppers to share unfiltered, open‑ended thoughts. When we rolled out our high‑demand linen bedding collection earlier this year, we ran a feedback campaign targeting shoppers who visited our Berlin online shop but did not complete checkout. In our old static surveys, most abandoning visitors would just select “price too high” with zero further explanation. With Permut, many shoppers voluntarily elaborated; one large group explained they hesitated because our third‑party logistics partner, Spedition Huber, had bad local reviews for delayed furniture drop‑offs. That granular context would have been completely invisible to us otherwise. I appreciate that Permut scales really well for our 120‑person retail team, cutting out all the busywork of coordinating, scheduling, and transcribing dozens of manual customer interviews. The platform auto‑tags recurring sentiment and key themes across thousands of conversations, so I do not need to read every single reply word‑for‑word. It works equally well for our website traffic and post‑visit in‑store shoppers, letting me gather consistent qualitative input across our omnichannel business. Even with those annoying dashboard performance hiccups we regularly run into, Permut delivers richer, more authentic customer commentary that directly guides our product tweaks and website copy adjustments. Review collected by and hosted on G2.com.

The biggest pain point for our retail research work is Permut’s dashboard stability and data‑filtering limitations. When we pull datasets from large‑scale campaigns, such as our sofa launch with over 2,700 total customer conversations, the dashboard frequently lags, and partial data segments fail to load entirely. Last quarter I needed to isolate feedback exclusively from customers who visited our Charlottenburg physical store location. Every time I applied that channel filter, the dashboard would time‑out and kick me back to the main campaign overview. I lost around two hours that afternoon, and ultimately had to export the full raw dataset and perform all my segment filtering manually inside Excel, eating into time I had allocated to deliver insights for our buying team’s product meeting. Another major frustration is inconsistent support for German‑language character sets during CSV export. Umlaut characters like ä, ö, ü get garbled inside downloaded CSV files. This actually caused a small misstep a few months ago: comments mentioning our “Eiche‑Eckschrank” oak corner cabinet turned into broken text after export. I misread the distorted feedback and initially attributed complaints to a different product line. I only caught the mistake after cross‑checking against original conversation records. This repetitive manual cleanup work is error‑prone and constantly slows down our end‑to‑end research workflow. Review collected by and hosted on G2.com.