Eight steps to better chatbot conversation design

Skyline view of Paris with Eiffel Tower in background.

If you’re involved in writing content for chatbots, or are looking to develop your chatbot conversation design skills, we recently spotted this presentation from the PLAIN 2019 conference which has some clear and useful guidance.

In her talk, plain language consultant Andrea Othilie Rognan outlines why robots get human interaction so wrong: unlike humans, machines have “no social intelligence, layers of meaning, or real understanding of context”.

To get around this, she offers eight simple but useful instructions for better chatbot conversation design and setting up machine-human interactions that are more likely to succeed.

In summary they are:

  1. Manage expectations – be clear on what the chatbot can do
  2. Use personality – be human in your responses
  3. Take turns – this is a conversation
  4. Be brief – the golden rule
  5. Speak the truth – make sure your answers are accurate
  6. Know what’s relevant – context is everything
  7. Be polite – Ps & Qs matter
  8. Fail graciously – how to get it wrong

See the full presentation here: The robots are here – how to combine plain language and conversational design.

Find out more about new jobs, partnerships and opportunities in the era of AI.

Leave the first comment

CJ Walker

Related Posts

Call to action

Why does AI want to know what’s in your sandwich?

Jerry Bartlett of Content Pro Tech Ltd poses a classification question that brings semantics home in this guest blog post. Enjoy! Is a hotdog a sandwich? And what the answer got to do with AI or technical communication? My answer…...

CJ Walker

Happy 2026! And a Question

I hope your long winter’s nap was enjoyable and your festivities brought you close to family and loved ones this season. .I hope you’re happy to be back at it—at least a little—in 2026. All of us here at Firehead…...

CJ Walker

Unlocking New Career Paths: How Knowledge Graphs Empower Technical Communicators

Part 4 of the Knowledge Organisation Systems Chain in our Skills for Modern Technical Communicators series In the KOS ladder—raw data → structured data → vocabularies/taxonomies → ontologies → knowledge graphs → semantically‑enabled services, we’re stepping from the ontology (the…...

CJ Walker