CFOTL: What’s your advice to executives who may believe driver-based models are too complex or sophisticated for their teams to manage or execute?
Ethan: First, in trying to create a model, focus on the key elements that influence your numbers. You don’t have to create a model so dynamic that every single line item updates perfectly. Think about the standard 80/20 rule in all things -- 80% of all expenses are probably focused in a handful of key accounts.
The following is an edited abstract from CFO Thought Leader’s “Ask Ethan” podcast featuring Ethan Carlson, CEO, Carlson Management Consulting, and Jack Sweeney, co-host of CFO Thought Leader.
CFOTL What’s your advice to executives who may believe driver-based models are too complex or sophisticated for their teams to manage or execute?
Ethan First, in trying to create a model, focus on the key elements that influence your numbers. You don’t have to create a model so dynamic that every single line item updates perfectly. Think about the standard 80/20 rule in all things -- 80% of all expenses are probably focused in a handful of key accounts.
Models don't have to be complex. Your business will dictate the level of complexity and what you can identify. It’s great if, say, every ten visits to your website translates into a sale and you can categorize how many marketing dollars drive people to your website and that cascades down through sales, or if you can look at salesperson productivity and see how that will translate into sales. If you can't, that's okay too. You should focus on what you can quantify.
Models can vary in complexity. They don't need to be overly complex to be useful. If I were taking an organization that had never done any sort of sophisticated modeling or budgeting but wanted to implement a model, I would take small steps and make sure we focused on a couple of key things and then expand.
All of the topics we talk about here are on budgeting and forecasting, and they are all iterative processes. It's not like you're going to come in and overnight transform a model and a business's way of thinking. It takes time: you start with simple models and drivers and continue to add the sophistication over time.
CFOTL When people want to adopt driver-based methodologies as part of their entire approach what do you recommend?
Ethan A lot of factors that will influence the timing. One question is do you have good information on your business and on your drivers? One problem that often presents itself to a CFO or a director of financial planning and analysis who wants to change how the company budgets is they don't have historical data and enough of a trend to really make informed decisions to determine whether these factors are in fact good drivers or not.
Often you have to start with what you can get your hands on, and then over time – be it three months, six months, a year – with the focus on those topics, you start to gather better information for the basis of your analysis, which then can be improved.
A lot depends on what's available and how long it will take to get it. Again, a focused approach is always best. In almost every business we’ve worked with, we spend a lot of time putting together these models and budgets and forecasts and reports. But at the end of the day, it's always a half dozen line items, or maybe a dozen, which are the big ticket items that really move the needle in their projections.
It's a shift in the thought process. If you've been using historical trends or a budget manager’s input, asking for a kind of “guestimate” as to the projections for next year, shifting to something that's more grounded in these drivers will take time. But again, if you focus in on just those key accounts first and then expand that, you can be very successful.
CFOTL Do you have an example of how some company went about this half-heartedly and then realized it needed to adopt some of these new approaches?
Ethan There are maybe two ways you can run into challenges with the driver-based model. One would be if you build a driver-based model but don't have a collaborative process to use it. In effect, you have a fancy tool for a finance person only and you're not getting input from the right people — meaning the model is good but the inputs are bad —garbage in, garbage out. Involvement is key.
The other challenge would be not to go too far. I have a current client whose driver-based model became too detailed, down to the level where he included how many color copies versus black and white copies a department might use, and what would be the factors in that. It was a lot of work and it didn't really have any meaning. So you have to be careful on that level of granularity.
CFOTL Can you expand on the pitfalls?
Ethan I can think of a number of clients that fall into that first category where you take the time and you build out a very involved model that is very well thought through with all of the right drivers. It produces financial results that clearly show why your revenue and expense numbers are what they are. But then they fail in building a consensus and collaboration with the business. As a result, it takes a couple of iterations to get to a better forecast.
The other key pitfall is building a driver-based model that is so involved that is covers every GL account, every department and the drivers all the way down to that color-versus-black-and-white-copies issue and its impact on expenses. I can tell you now that the impact it had on our client was insignificant. As a result, they've changed that part of the model to focus on things that are more relevant to their budgeting and forecasting projections.
It's an iterative process. You're going to make some mistakes along the way. You're going to collect data that's not useful, and then you iterate and you improve it. It's that evolution over time that separates the clients and the companies that we've worked with that we would say are the best in class and separates leaders from the ones that are average.
CFOTL Ethan Carlson, thank you for joining us on CFO Thought Leader.
Have any comments or questions for the next ASK ETHAN podcast? Please use the form below.