Presenting bad news to the client? Unexpected results? Let’s ensure to not be shot as the messenger this week. As usual, let’s start with the definition given by Dave Sturrock on the Simio blog as a starting point:

The modeller should not be criticized if the results promote an unexpected or undesirable conclusion.
— Dave Sturrock, Simio Blog
Bad news to report?

Bad news to report?

This week, we have a very short quote but a very real danger: your results are under everyone’s attack as no one expected them and they are bad news. People are starting to doubt your work and focus on your lack of skills like desert vultures on fresh meat. In your eyes, the work is extra-valuable because you supplied the client with a fresh view out of the box. They should thank you, not attack you… shouldn’t they?

Here are four points to avoid being eaten by the vultures and instead lead the discussion into a useful direction:

ONE: Acknowledge the biases

Both you and the client have particular biases right now. Your client was (or still is) convinced of his solution and expected the awesome simulation model to support his feeling. How could it not? Surely it must be wrong somewhere. You, on the other hand, strived hard to satisfy the client and assumed an honest result would make him thank you for your skills and insights.

In acknowledging these stand points will you be able to guide the discussion much better. Grant your clients their disappointment and surprise: “I know this is not what you expected but we double-checked the results and are confident of the conclusions”. Notice the second bit there: you must be confident about your work and clarify its value.

TWO: Avoid surprises

While these window cleaners are a pleasant surprise in a children hospital, your client CEO might not like surprises too much!

While these window cleaners are a pleasant surprise in a children hospital, your client CEO might not like surprises too much!

There is one thing a CEO does not like in meetings: a surprise. If you have bad news, by no means ever share them first time in a large project meeting. Discuss it beforehand, if necessary. Ideally, you should be in a close work relation with your client (see previous posts on agile modelling). If you talk progress every week, all major surprises should have evolved over the project. This means smaller chunks of bad news to swallow. On top, you will have started a mitigation process already. News should have reached the client CEO but double-check before the final meeting. Again: they hate surprises.


If I tell you that 7% of Americans believe that Elvis is still alive, it might be an entertaining (and true) fact. But if you did the poll yourself, you’d probably not forget it for the rest of your live. Similarly, it can be useful to provide the simulation tool to your client but let him arrive at the unexpected conclusion. This could happen as part of a staged training session where you guide him through the model step by step such that the problem in question becomes apparent. Ensure to offer plenty of alternative exploration alleys so they can convince themselves and learn about their erroneous conceptions.


If you need to find water on Mars, suggesting solutions is important.

If you need to find water on Mars, suggesting solutions is important.

When presenting unexpected simulation results, you should naturally have thought about solutions or next steps. Present these pro-actively (even when not asked by the client). This will show you did not stop thinking beyond what you were paid for and also diverts their attention from blaming you to start thinking about possible solutions. Ideally, you can start a discussion on that rather than fighting of blaming.



Four ways to avoid being shot as the messenger of bad news: Be aware of client bias when presenting unexpected results. Avoid making it a surprise in the first place. Let the client arrive at the counter-intuitive conclusion himself. Suggest solutions.