Pendo

Pendo combines product analytics with in-app guidance, positioning itself around surfacing where users and agents struggle and then acting on it, with the vendor reporting more than one billion user experiences and framing itself against software that is shipped without regard for whether people can use it.

Combining measurement with intervention in one product is the structural idea, and it addresses the gap that makes most product analytics disappointing.

Analytics alone identifies problems and cannot fix them. A team learns that sixty percent of users abandon a particular workflow, which is genuinely useful, and then the fix requires design work, engineering time and a release, which means it enters a backlog and competes with everything else. The insight is real and the response takes a quarter.

In-app guidance changes the response time for a specific and substantial class of problem: the ones caused by users not understanding an interface rather than by the interface being wrong. A tooltip, walkthrough or contextual prompt can be created and published by a product manager in an afternoon without a release, and for confusion problems that is often the correct fix rather than a workaround.

The limitation should be stated. Guidance layered over a genuinely bad workflow is a patch, and organisations can use it to avoid fixing structural problems indefinitely. The discipline is distinguishing what people cannot find from what is actually broken.

Including AI agents alongside users is the newer and more interesting claim. As software is increasingly operated by agents rather than people, where those agents fail is a product question in the same way user friction is, and the failures differ: an agent does not get confused by unclear labelling but does fail on ambiguous states and undocumented behaviour.

Pricing is published on the vendor's pricing page with a free start offered, though no figure was captured.