What Poth is
Most feedback tools are inboxes. They collect, tag and summarize. Summarizing a source tells you what that source said. It cannot tell you why a segment behaves the way it does, because the answer lives in the relationships between sources, not inside any one of them.
The difference in one example#
A support tool can tell you that 312 tickets mention a missing CRM integration. That is a count.
Poth can tell you that those 312 tickets cluster in Enterprise RevOps accounts, that the same accounts show a drop in weekly active seats two weeks after onboarding, that three renewal calls raised the same blocker, and that the connector build is the lever with the largest expected effect. That is a decision.
What it is built on#
A network, not a list
Responses, tickets, calls and behavioural data connect into a graph. Relationships between answers carry information that row-and-column storage throws away. See Knowledge graph.
Hypothesis-driven agents
Agents form competing explanations about your customers and test them continuously, rather than waiting for someone to run an analysis. See Hypothesis engine.
Statistical rigour
Poth applies proper significance testing instead of reporting whatever correlation appears first. A finding that does not clear the bar is shown as rejected, with its confidence.
Adaptive collection
When the data on hand cannot settle a question, Poth writes targeted follow-ups and surveys to collect the missing piece. See Adaptive surveys.
What Poth is not#
- Not a survey tool with an AI summary bolted on. The survey is the last resort, used when existing evidence falls short.
- Not a dashboard you have to interpret. Findings arrive ranked, with the segment and the recommended action attached.
- Not a model trained on your data. See Data and privacy.
Next#
- How Poth works walks the path from raw sources to a ranked answer.
- Your first question gets you to a real result.