Analytics: understand what your customers are asking

In the dashboard you see how many conversations there were and how that develops over time. You also see how many conversations the bot resolved on its own and how often it handed over to your team. The most frequent questions have their own list.

If the bot captures leads, you can also see how many conversations led to a contact. Analysis is GDPR-compliant.

If an answer isn't in any of your sources, the bot says so instead of guessing. The question then shows up under unanswered questions. For most businesses this is the most useful list in the whole dashboard, because it shows quite precisely what is missing from the documents.

We work this way ourselves. PixelConsult uses My-Chatify for support on its own software, where the bot answers 2,100 questions a month automatically. What it couldn't answer, we see in analytics, and with it the places where the documentation has gaps.

Basic analytics are included even on the free plan. They cover message count, active sessions and frequent questions.

From the Professional plan, advanced analytics are added, with conversation histories, usage trends and performance metrics. That is especially worthwhile when conversations pile up and you want to take a closer look.

Analytics is only worth what you do with it. For an unanswered question there are usually three options: create a Q&A pair, add to a document, or decide that the question belongs with a person anyway.

In the first weeks after launch it pays to look regularly; after that, once a month is usually enough. After each change, check in the Playground that the answer now comes through.