Sometimes, actuaries have problems effectively communicating the results and recommendations from their analyses. There are several reasons for this: lack of training, insufficient time to adequately investigate the impacts of the analysis and compose a compelling message, and a general haze of uncertainty surrounding just what actuaries do, anyway.
The problem is, when you’re not effective with your communications, your company inefficiently implements your insights or misses an opportunity. Your professionalism may be called into question. Worse, you may be viewed as nothing more than a glorified number-cruncher, who must leave the interesting activities of strategy, prioritization, and creative option design to the finance analysts and the data scientists.
In order to help you improve the effectiveness of your communications, keep these ideas in mind whenever you’re drafting your next memo, report, or presentation.
1. They are not you
This is, simply put, just a different way of saying Know your audience. Are you presenting the quarter’s financial results to the division’s department heads? That’s a different conversation than if you’re presenting the quarter’s financial results to the CFO, or your Chief Actuary. They’re going to want to see, hear, and be influenced by different elements of your analysis.
Just as important, though, is to remember that they are not you. They don’t have the skills that you do. They don’t have the background that you do. They don’t use the same jargon and terminology that you do. They’re not as intimately familiar with the data and the model as you are. As a result, they are going to interpret whatever you present through a very different lens than you may have intended.
In these situations, it’s often better to reframe whatever you want to say using terms and examples that will be familiar to them. That way, they can immediately understand what your results mean. When you do this translation for them, they’re much more likely to understand what you’re saying, writing, or displaying.
2. Start at the end
There’s a great temptation to start where you started and walk through whatever you did. You think of beginning with the data downloads that had some idiosyncrasies this month, the extract-transform-load operations that had to be re-worked because of those idiosyncrasies, the models which produced some slightly differing results than expected, and so on.
Why shouldn’t your report explain, in excruciating detail, all the steps you took to get to the end point?
Because, and this references reminder #1 from above, they don’t care what you did. (In general. Sure, the auditors might care that you followed all the controls, but that’s a different issue.) Their interest lies in the results you’re presenting, not the tasks you completed to get the results. They want to eat the sausage, they don’t want to be told how it was made.
If you do it the other way around, by describing every single step and action you’ve taken, you’re going to lose your audience. They will get bored, ignore the rest of your report, and ask someone else to summarize it for you.
This creates problems. It adds an unnecessary delay between your analysis and the implementation of any recommendations you’re wanting. And it potentially dilutes the power of your analyses due to additional influences adding their own commentary.
So my recommendation is to begin with the end in mind. Aim to distill it all down to one or two most critical points. What is the real impact of this report you’re making? What’s the meaning behind this presentation? If all the fluff were boiled away, and you only had 2 sentences to hook their attention, what would you say?
If you structure your communications this way, you’ll earn the right to justify your critical points later with all of the sausage-making details that support it.
3. Emphasize utility over precision
Another challenge many actuaries run into is offering way more precision than their audience can handle. A former chief actuary said, “You get two significant digits. Anything else is going to allow your audience to concentrate on the wrong things.”
Significant digits are the numbers to the left of the decimal point, if you’re greater than 1, and anything after the leading 0s if you’re smaller than 1, and you go from left to right. So $1,465,308 has 7 significant digits, from the 1 through the 8. 0.0037 has 2 significant digits, the 3 and the 7.
The reason this matters is because people generally focus on the last thing you told them. If you’re dealing in numbers that are on the order of millions, the ones digits don’t matter. And if you’re looking at things that are fractions of a percent (0.0037), increasingly smaller fractions don’t matter, either. Practically, 0.0037418 is the same as 0.0037. Any decisions you want to make based on 0.0037418 will be just as useful and just as accurate as those based on 0.0037, and then you don’t let your audience get distracted by focusing on the wrong thing.
Here’s how that looks in practice. Suppose you’re asked to update the expected claims for the month now that an error in the IBNR calculation has been corrected. Let’s say your original values were $31,465,380 of expected claims including an error-filled IBNR of $93,475. And the IBNR after correction is restated to $103,650. Now your total value of expected claims is $31,475,555.
If I’m speaking to anyone about this analysis, I’m saying the following: “We found an error in our IBNR process and corrected it, but no worries. Expected claims are still 31 million dollars. Last quarter’s expected claims were 27 million, and it was only 21 the quarter before that. What’s going on?” I’m never going to spend time talking about the ones digits (much less the pennies), because as soon as I do, we’ll soon start arguing about the change in IBNR, and that’s the wrong message to take away from this situation.
4. Make it actionable
In marketing, and in great fantasy stories like Star Wars, there’s always some kind of important call to action or adventure that really starts the work. This is the critical moment where the authority asks the audience to do something. Could be to [CLICK HERE] to get some discount. Or, like in Star Wars, when Luke intercepts Leia’s distress call intended for Obi Wan Kenobi. Now, he’s got something to do!
Similarly, your communications often should conclude with a request for action from your audience. Without some kind of definite request for action, even if it’s nothing more than just accepting a report, you can get stuck in limbo where you’re waiting for the audience to do a thing, but they don’t know you’re waiting on them to do the thing. Let’s go through one last example.
You’ve been tasked to critically evaluate the current pricing model, with an eye to either adopting a new platform or initiating a project to bring the current one up to requirements. After your evaluation, you estimate that you could adopt a new platform in 40% less time and for 60% - 70% lower total cost of ownership than going through the hassle of upgrading your current one.
Do you just throw this at your boss, like a data grenade, and say, Good luck understanding this, see you later? Nope! You follow-up this analysis with a call to action: “Should I initiate a conversation with this new platform vendor to get a firm quote? Or is that on you?”
Now, the boss has something to do. She can be the hero of the situation. She can either delegate the next step to you again or take on the responsibility herself.
Either way, there’s no ambiguity about what’s happening next. Someone is going to do something, and other people know how to hold them accountable to it.
There. Four tips for better communication:
- Your audience is not you;
- Start at the end;
- Emphasize utility over precision; and
- Make it actionable.
When you keep these in mind as you craft your communications, you’re much more likely to be effective.
Think I’m way off base? Send me a message at email@example.com and tell me all the ways I’m wrong. I’d love to have that conversation.
See that? That was a call to action. Available wherever effective communications are found.