I am still amazed when someone reminds me that Balanced Scorecarding is still a relatively new concept as is Lean Six Sigma (LSS)- At the CFO magazine conference, multiple CFO’s and CEO’s approached me after my presentation to ask how we came up with our KPIs with trepidation as though they were going to expose that vulnerable, "I don’t know" side of them that they rarely wear on their sleeve.
You might be wondering how a presentation delivered at a BSC conference could be missing information on how to define KPIs…Well, that is my trick…It I were to open that can of worms, and believe me, the topic of KPI definition is a can of worms, I would end up rat holing into a lengthy, somewhat off topic discussion that is completely subjective and customized to each organization. I mean if CFOs cant agree whether to report their Gross Profit or Net Revenue, which others may ask "well, aren’t they the same", then one should assume that something as laborious (that should be laborious) as KPI definition would not be easy. In fact, it is, while at the same time, time intensive UP-FRONT. Yet, how much time do your analysts waste collecting data, often viewed from a different lens by different groups, and rarely even showing the same information and numbers? How many times a day do you hear "spreadsheet" or "pivot table" — This time intensive process for them yields arguments about the data integrity and often frustrate your analyst who probably just pulled an "all nighter", getting ready to present these numbers to you, only to leave with their tale between their legs because no one believes their version "of the truth" because Sally So and So Data gal showed an entirely different analysis around the same metrics. Talk about a laborious, time suck, cyclical in nature, often forcing the analyst to reconsider their career path or go home in frustrated tears because of their wasted work.
Think of it this way…have you ever had an employer who gave you so many objectives (I had 43 on my last mid year review), in a melting pot approach covering everything from soup to nuts and then some? How did you feel? Did you have a clear path of prioritization of those 43 objectives when stack ranked against one another? Chances are…Probably not. Which did you tackle first, and in hindsight, do you still agree with your choice?
This is the part where setting expectations that the KPI definition process should take time and should be considered the most vital part of sustaining any quality program or BSC long-term. In fact, this process was extremely difficult for us, and one that we spent more time than most phases of our implementation (except when we finally automated). This is what I meant by "should be as laborious as KPI defintion" from the 1st paragraph– If an organization, like most, races through the BSC planning phase, without defining their strategic pillars or strat map, they will fail, plain and simple. Of course, unless Jack Welsh was your leader with an infused and energetic message of support it or get out, it will fail.
Burying your head in the sand at this point (aka jumping to another URL that says it is ok to not tie into the strat plan), a sin in the BI/PM space, downstream, will leave you with sloppy dashboards, over run with KPIs and will leave the user feeling like the INFORMATION OVERLOAD demon just sucked their brain out. In fact, should the resident PM / BI person who is building the BSC, not take the time to build a strategy map FIRST, which then becomes the roadmap towards KPI definition, YOU should feel personally empowered to stand up and tell them to "take a step back and rise to the 60,000 ft holistic view" of their program. By focusing their attention to the benefits of cascaded indicators and the implications to both the upstream/downstream flow of information within any performance management system not tied to the corporate goals and objectives, you will become an INFORMATION OVERLOAD demolisher and my personal hero for taking my advice.
This simplistic example can quite easily happen to a BSC implementation (and happens all of the time) without a strategy map and crisply, defined KPIs, limited AND I MEAN LIMITED, to just the critical or vital inputs that drive your business, in addition to the voices of your internal and external customers. The diverse voices that the VOC and VOIC (a term I created) relationship offer must be considered, reconciled and balanced to develop a truly successful BSC.
I’m going to stop and make a point here: It is important to realize that an organization does not, nor should they blindly respond to internal or external customer needs and opportunities, the key word being blindly. We’ve all seen those C-levelers go out and start "shooting from the hip" on a whim or on a once off issue that happened to a removed family member that they escalated over Thanksgiving dinner to Mr/Mrs C-level. Thus, NO BSC nor LSS program should ever be allowed to deploy without first defining a strategy map from the bottom to the top line of your business. This, in effect, should define everything your business does from both the business strategy level as well as the day-to-day. It should contain definitions of your internal and external customers and markets to be served, competitors, and competitive strengths, while providing a framework from which to evaluate potential KPIs. As KPIs are identified, there are too many tools out there to use, but I will name 2 that helped us prioritize and capture this information. The techniques used will depend on the nature of that customer relationship and your business. It is this voice of the customer (whether VOC or VOIC) that should be the core element of all perspectives of KPI definition. Sometimes, there will be vital KPIs which you wont have measurement systems in place for, much like with Six Sigma programs…As they say "you measure what you manage, and manage what you measure’ and often that list reflects a lack of line of sight through the organization in which people think they know what they know, all of the while forgetting that they don’t know what they don’t know yet. One might assume what’s important, based on others’ experience, like customer satisfaction index, revenue, bookings/transaction counts, etc…But do you know that those gauges are mutually inclusive of your end-to-end business strategy? How do you know that an increase to customer satisfaction means squat to your business? I mean, intuitively, it makes sense…But, it isn’t true…and I say that with complete authority of knowledge. In satisfaction for the sake of satisfaction will turn even the most staunch supporter away from your satisfaction program and create an environment that propels ney-seighers to the forefront, and will plague any BSC initiative that may follow. Even more puzzling should be the question of interaction and effect: how do you know that these will move the needle the most, representing your Main Effects? What about the 2 and 3 and 4-way interactions between them or whether there exists other inputs that actually influence the output more or exhibit colinearity in their relationship to each other? All of these are very important questions to explore while defining your KPI selection process.
For us, as I have laid out many times before, we took our time with our KPI selection process starting with the strategy map which we used to define those top line scorecard metrics most relevant to our leadership team From there, using a multi-pronged approach that yielded both the scorecard KPIs or the ‘WHAT’ , as I like to call it at the top line, is happening to your business health (from all 4 perspectives) — It shouldn’t include more than a handful of absolutely critical KPIs that your executive team will eventually use to make faster twitch business decisions, and then YOU STOP.
Review, revise and optimize before you go any further…Why replicate a bad process?
Once all iterations were completed, we cascaded down 1 level, and began exploring the linkage options: some do it at the LOB level, while others, at the geographical level. We chose to do it at the LOB level, LOB by LOB, which was then cascaded to the business teams and finally down to the call center and Telesales agents since they too contribute to that KPI and their performance directly influences our hypothetical example of our "Revenue stream" — (As an aside, I cannot understand why so many companies forget about them or think it’s not important to roll it down to their level considering they are the closest to your external customers).
This cascaded view should contain 2 handfuls of metrics max and each has to be tied into the top line scorecard and strategy map, in an effort to answer the ‘WHY’ that the initial user had when the saw the ‘WHAT’ had gone red.
Think of it like this…Scorecards tell you ‘WHAT’ in less than 5 KPIs.
Dashboards tell you ‘WHY’ in less than 10 KPIs (across all lines of business)
Stop, review, revise and optimize before you go any further. (This is where the black belt in me really got its groove on was during these stage gate process stops because they allow the planners to make quick changes without having to change the entire BSC platform. Replication should only occur once optimization is achieved).
Why…? If we saw Revenue had gone Red, and cascaded down to our LOBs, to find Flights = Green, Hotel = Red, and Car = Red, would you be able to tell you CEO what was driving decline in revenue? No, because your root cause could be due to 1 of 2 main effects or the 2 or 3-way interactions between the two: Hotel &/or Car.
Next, you need to cascade your KPIs down to the lowest level you can get to.
For us, we start all Scorecard based KPIs at the top line, and cascade them down to the call center agent….For example, Revenue as the top line metric can flow through to our LOB channels, or flights, hotels etc, which can further be stratified downstream by the units that role into each, like Retail, Marketing and Operations, which can be cascaded again in the case of Operations, to Telesales and Customer Support and downstream one more layer to the actual agents themselves.
How many scorecards and dashboards do we have, I ask, based on my explanation above?
How many KPIs should we have? (use the MAX spec limits that I have laid out for you above to derive both — answer will be on the bottom of the post).
The other prong to our approach of KPI definition was that we wanted to build our audience buy-in and support simultaneously to defining the KPIs…How does one do that? This is where leaning out the process as much as possible helps, but recognizing it to be lengthy upfront, helps take that pressure off, especially knowing that doing it right the 1st time, saves time, sustains quality and reduces cost of updates in the long run. A key to our proverbial LSS and BSC toolkit is that in order for any of these initiatives to be successful and not be seen as a flavor of the month, one has to act as a change agent, culturally shifting perspective of both the negative and positive views that I guarantee exist in your population of internal customers (and perhaps, some of your external).
For us, we started at the top line first (President, CFO and their direct reports all at the V-level of SV-level). Once we had a list of drivers for their business, we met with the leaders of the individual lines of business teams, along with whomever they felt should be in the room. You don’t want to condense these sessions into 1 giant meeting, as you will end up arguing more than solving anything. Instead, we opted to meet with the Flights team first, review with them the metrics that their leaders came up with and brainstormed others to add to that list. Once a full list of roughly 30 kpis that cascade from Revenue were derived (representing 10 per unit and team under the LOB max), we used two Six Sigma tools to help us prioritize and methodically select from the list just the critical or vital KPIs for our dashboard, using the LOB strategy as our driving factor, as well as the voice of your internal customer, or what I have defined as "VOIC", to help differentiate from the external customer voice (or "VOC", as most know it today). These 2 tools, moreover, often used as part of a Six Sigma program or toolkit, are known as the QFD and FMEA, or Quality Function Diagram & Failure Modes and Effects Analysis, one centered around quantifying qualitative measures in an effort to help prioritize possible alternatives. The FMEA helps collect the potential failure modes that can occur during any new process or project roll out, and assigns risk ratings to each, called a RPN score, as well as documented action plans and EOC ratings (Ease of completion) against each.
Create an upfront consensus at where to draw the line: QFD importance rating of 80% or higher only, or a RPN of 100 of higher, so that the team of people has no real wiggle room to argue after the fact, since they agreed upfront.
Because they participated, they will buy into the KPIs more than they would had you planned them in a vacuum or with the C-levelers only.
Just remember, sometimes the best things take a big of pain upfront to launch or achieve, though for me, it is always worth it in the end when both prongs are successful. When taken in conjunction, these two tools are extremely powerful and robust, and will exhibit great big-picture thinking on your part.
Answer to my questions above:
How many scorecards? Answer = 1 (and only one, at the top line strategic view); all others, are dashboards that cascade from that 1 scorecard with no more than 5 KPIs.
How many dashboards? Answer = 1 per LOB (Flights, Hotels, Cars, Packages, Cruises and Tickets = 6 LOB Dashboards, 3 shared services dashboards (retail, marketing and Operations) and 2 for both Telesales and Customer Support, or 6+3+2 = 11 Dashboards that all cascade from the 1 scorecard, and provide a clear line of sight through our entire organization!!!