Skip to content
BotFlowDev
Services Approach Platforms Guides About Contact
Discuss your project
Home/Guides/Reporting
Guide · Reporting

What to measure in a chatbot programme

A chatbot can produce a lot of numbers. These are the measures that usually help programme managers make decisions.

By Morgan CollettPublished 3 October 2026
In this guide Reach and engagement Journey completion and drop-off Opt-outs and failures Data quality Messages and cost per outcome Links to programme outcomes Dashboards people use

Reach and engagement

  • New registrations over time, by channel and language.
  • Active participants: people who replied in a given period, not just those who received messages.

Journey completion and drop-off

For each journey, measure how many people start and finish, and at which step those who leave stop. Drop-off by step is the most useful number for improving a journey.

Opt-outs and failures

Track opt-outs after each message or series, and messages that failed to deliver. A rise in either is an early sign that content, timing, or frequency needs attention.

Data quality

Measure how often answers are invalid or need re-asking, and how often fields are left empty. Poor data quality usually points to a confusing question.

Messages and cost per outcome

Count messages sent per completed journey and, where possible, the cost per registration or per completed survey. This is increasingly important as messaging charges change.

Links to programme outcomes

The most valuable reporting connects chatbot activity to programme indicators, such as clinic visits or course completions. This usually needs an integration with the system where those outcomes are recorded.

Dashboards people use

Agree a short set of measures for each audience. A dashboard with six numbers people check weekly is more useful than one with sixty that nobody opens.

Integrations and reporting

Need help with this?

We connect journeys to your systems and build the reporting your team needs. BotFlow Dev works on your existing Turn.io or RapidPro/TextIt workspace.

Discuss your chatbot projectSee our services

Related guides

AI

Using AI in chatbots responsibly: guardrails, evaluation, and model choice

AI can help a chatbot handle free-text questions, but programmes serving the public need clear boundaries. This guide covers the decisions to make before and after launch.

Read the guide
Platforms

Turn.io or RapidPro/TextIt: choosing a platform for your programme

Turn.io and RapidPro/TextIt are both used for programme chatbots, but they suit different needs. This guide sets out the questions that usually decide the choice.

Read the guide
BotFlowDev

BotFlow Dev is an offering from Morgan Collett Software.

hello@morgancollett.com

BotFlow Dev

  • Services
  • Approach
  • Platforms
  • Guides
  • Contact
  • Privacy & Legal

Morgan Collett Software

  • Morgan Collett Software
  • BotFlow Host
© 2026 Morgan Collett Software Turn.io, RapidPro and TextIt are the property of their respective owners. · Privacy & Legal