Modellers can be shrinking violets. They like to do their modelling in private, in a darkened room, showing nothing to anyone, except a summary of the outputs.
They believe the model needs to be kept as an impenetrable ‘black box’, away from prying eyes who might ask too many awkward questions, uncover that the model is actually quite simple or simply not very good.
To wow your clients, (or your boss) and to be a truly great modeller, you need to have the confidence, when called upon to bear all in public!
Most clients will be used to being told how big and complicated the model is; that it will take days to run the numbers; that you need to go away and think about it; that there is a complicated optimisation process or quality assurance process you need to go through, etc.
How cool will it look to these clients when you rock up, roll up your sleeves and say, let’s take a look at the model? Let’s get together and run a few scenarios, try a few things out?
Pretty cool, would be my guess. They will get a real feeling of confidence that you know what you are doing.
Uncover Problems / share risk
It’s good to get the model ‘out there’ so decision makers can see what is really going on and not just the headlines. Making sure people have a real understanding of the workings of the model will help drive better decision making and will reduce the risk of points of detail being missed.
Also, as the modeller, getting the model ‘out there’ as much as possible reduces your personal risk if any details do end up getting missed. It becomes a collective problem.
Add value and get recognition
If you are sat in a meeting, driving the model on screen, playing with variables, testing scenarios, you are really adding value to the team. You are the key part of the decision making process. People can clearly see what you do, that it is hard and that you are good at it.
You have no choice
That horrible feeling, when you have to present the model to the team and model live, when you are not confident that the model will do what it should. You need to be prepared for this.
What to do.
Although there are benefits to bearing all, it can be risky. If you or your model are not up to scratch, proceed with caution. Here are some key considerations / suggestions:
Time and place
There is a time and a place where it is appropriate to bear all. If you genuinely do have a situation where there are lots of input changes needed and the update process is lengthy and painful, don’t try and do this with an audience.
Have a good model
If you are going to bear all, you need to be able to deliver.
The models you build must be robust and you must have a high degree of confidence that they work correctly without you having to do lots of ‘tweaks’. This is good modelling practice in any case, but the thought that you are going to be running the model in public might help focus the mind.
It must be easy to trace through from outputs back to inputs. If you are running the model and the output doesn’t change in the way you would expect, you need to be able to quickly work through the model to explain why.
The model has to look good
If you are showing your model off in public, it has to look the part. You should have a clear dashboard showing what your audience want to see. You should have a clear and consistent layout of sheets, your colour scheme should be appropriate, you should not have highlighting, #REF and #DIV/0s all over the place.
Know your audience
You should consider who your audience are and what they are likely to want to see from the model.
What are the key variables they will be changing and what are the key outputs they are interested in seeing. A good way of running a model for an audience is to bring together these key inputs and outputs on a single sheet so you can amend an input and immediately see the impact.
Prepare and test
If you know that you are going to be demonstrating certain bits of your model and amending particular inputs, check out in advance that the model behaves as it should when you amend those inputs. You will look even more impressive if you have anticipated the likely changes and have an expectation of the results.
Please let me know if you have had any positive (or negative) experiences of sharing a model with an audience and if you have any tips on how best to present a model.
To find out more about presenting your model, read our blog post “10 Presentation pitfalls to avoid”