As the last session showed, what we really need to use analytics for is ultimately to choose action plans. How can I take this analysis to say how can I as a manager actually implement something that's going to improve some non-financial dimension that ultimately leads to higher financial performance? Again, how can I pick out the thing that's got that biggest bang for the buck? And again there's lots of ways we can do this. So, ultimately you should really think about this as, again, peeling back the onion. Right. You find these initial statistical results, but what you need to do is say," Okay. If I find that, what exactly is driving that relationship, and what can we as a company do to actually improve this." So even though you may start out thinking, "Hey, what I'm looking at is the relationship between a customer measure and financial performance." In the end what actually may be driving that has nothing to do with that, it may have something to do with your employees on this. It may have something to do with your facilities. So looking at these non-financial metrics in isolation of other metrics is not especially sophisticated. You really need to put all these metrics in the broader business model that your company is working with. Try not to do it just functionally, because what you find a lot is marketing does some analyses, operations does some analyses, quality does some analyses. Nobody is talking to each other. And it's not like the actions in one function doesn't impact the other one. So if you really want to say how are these metrics related to each other, how are the actions going to impact financial results? You need to start thinking a little broader, and then peeling it back and saying, even if I'm looking at this function, maybe it's the function over here where I really have to take the action plan on this. Right. So, you may think of starting out say, I'm looking at customer satisfaction financial performance, but ultimately as we saw on some of the earlier examples, it may be the case of things like employees are related to other financial or non-financial customers. It may be that these non-financial numbers ultimately lead to the financials, but there's intermediate financial numbers you need to think about. Right. It could be they lead to cost reductions. It could actually be they lead to cost increases because I have to spend money to make these things go up, and then ultimately it's that cost increase, or investment in something like RND that leads to the financial results. So you again need this causal model. I predict that if A happens B happens, if B happens C happens, and lay that out naked either be financial or non-financial. So, let's walk through an example of how you would keep peeling back the onion until you get to this action-plan, that one you can do the financial analysis on, and to decide this is something that I think a manager can actually do. So let's go back to a financial services firm that we looked at before. So this was the firm that when they did the initial analysis trying to link customer satisfaction at the firm level to firm financials, they found nothing. What they did find was there was, there was a big relationship between how satisfied were you with your investment adviser. Which had a relationship with how much asset did you have invested with them. And this was an example of on this one to seven scale, you wanted people be really satisfied with your investment adviser. That's where the biggest bang for the buck is. That sounds great. What does it mean to say," I'm satisfied with my investment adviser" What can this company do to actually make you more satisfied as a customer? Because if I stopped here, if I was a manager, and my boss told me increase satisfaction with your investment advisor, I have no idea what to do. I have no idea when the customer answered that what they meant by saying I'm satisfied. So what we are going to have to do is now let's step back. What do we mean by satisfied with the investment adviser, and what can we actually do to implement an action plan? Okay. So here's what this company wanted to do. We had that initial result. What they wanted to understand, what are the drivers that I can actually take an action on that's going to lead to these future financial performance. Because again, we want to use it to develop our strategy, and two, we need to select these action plans with the largest expected economic payoff. As we saw before, increases in customer retention, and assets invested, or under management had a direct impact on future economic success. But again, this company didn't know what the drivers of retention and assets investor were. And without knowing that, I don't know where to invest money, I don't know how to change my strategy to take advantage of this information, and I don't know how to tell my managers which dimensions of satisfaction do we think we need to focus on. Okay. Well it turns out, once you start peeling this back, additional analysis found out this customer satisfaction with the investment advisor was related to three things; trustworthiness, responsiveness, and knowledge. Okay. The big thing though was not those, it was investor adviser turnover. Not just what do I think the investment advisor I have now is, really can I deal with the same person over and over again. Think about it. I'm taking my pension fund, I'm taking all my savings, I want to give all this to an investment adviser, I want to deal with that same person, I want them to feel like part of the family here. So the big issue here was not just satisfaction. Now I peeled it back, and now you have to start looking at investment adviser turnover. What can you do to reduce turnover? That's one step back when you do the analytics from what we had before. Okay. Well we still have to go further, because I still don't know how to reduce turnover. So again, you can start doing analytics on this. So here's the model they estimated. Now again if you look on the right hand side, there are the ultimate outcomes I wanted. We started out with customer satisfaction with the investment adviser. When it goes up, it impacts whether I stick with you, and how much money I invest with you. Okay, that's fine. We took that as a step back, and we said, "Okay, what's the big driver of customer satisfaction that matters?" Investment adviser turnover. Let's take one step back from there. What are the things that we can do as a company that are going to impact investment adviser turnover. So what you see here, the little pluses is it's a positive and significant relationship. Okay. Because the company didn't want us to give the coefficients. If it's got more little plus signs, it's more significant than if it's got fewer. So if you look here, there are certain things that have a stronger statistical relationship than other ones, which you would like to do again and say, given the ones that have a big impact on adviser turnover, and given how much money it's going to cost you to actually do that, where would we like as a company to change our strategy, and change our action plans? Okay. So, here's what they ended up doing based on that. You find out that the level of compensation is one of the most important drivers adviser turnout. Not really surprising. Right. If I pay him enough. Right. We are going to get that. The question is how much are you willing to pay them. Right. But based on this, they started looking at what are some of the other things in addition to pay to develop these action plans. And based on that, what they found out is some of these things are, things like work environment, is the work environment better, in terms of work environment as perceived by the investment adviser. Knowing that, now you can start coming up with one of these net present value models that we had before. We started out with customer satisfaction leading a financial performance, we ended up having nothing to do with customers. We came up with a human resource initiative, and linked that up to economic value. So, here's where if you are going to start doing the action plans you get to keep peeling it back, and it may turn out you end up in a place you never started, because when we started working on this, we started working with the marketing department. We ended up working with the HR department, because that's where the actual driver was, that was going to lead to this financial performance. And that's where you have to put in this causal model of what drives it, and keep working backwards as you go through the model on this.