Budgets, goals, KPIs, STIs and other trivia

Yesterday I mentioned being involved with 20 versions of a budget, implying the practice was a bad idea. Au contraire. I am, in fact, an advocate for testing multiple alternatives under multiple scenarios. It’s just about the only way to find the weaknesses in your understanding of the complexly inter-related system we call ‘mining’.?

What I’m not a fan of is tweaking a spreadsheet in a misguided ‘goal seek’ approach to generate the parameters that result in a preconceived outcome. You need to take a moment to think about that statement…. What preconceived outcomes are buried in your models, estimates and plans? They can be hard to identify blinded as we are by our self-serving bias. Objective honesty might be the rarest commodity on the planet!

Here’s an example. The very first time I was involved in presenting a site budget and plan to our CEO was both exciting and something I looked forward to with trepidation.?As a young Geo still building a career path, I thought it was a big step to be in the room with the big players. When you are new in this game it’s easy to forget that the big boss gets dressed the same as everyone else… one leg at a time!

So there I was. The most junior person in the room. I was there for my knowledge of the resource and because my boss was off being a good husband as his wife gave birth to their second child (that was almost a rare thing in those days.)

A team of geologists, engineers, metallurgists, accountants and other specialists had -slaved- over that budget. Diligently making estimates and plans, testing assumptions and parameters. It was probably the work of a dozen people over 3-4 months, no trivial matter. The plan was supported by a communication pack. No fancy PowerPoint slides, a simple overhead projector and a thick set of transparencies with a strategically placed transparency pen placed to the side so the GM could make highlights as he presented.?

The meeting started. Safety first - a new practice. Then, the second or third transparency. The one with the summary physicals. We believed in the mantra of ‘tell me three times’ when it came to presentations.?Tell ‘em what you are going to tell ‘em, then tell ‘em and finally wrap up by telling ‘em what you told ‘em… having those summary physicals upfront was the first step in locking in the key message.

It failed. Utterly and totally failed… the CEO took one look at the slide, focused on the head grade and said in a clear voice, not to be questioned… “That is not?right. The head grade next year is 1.4% not 1.2%”.?

I’d love to know how he knew that.?

The meeting stopped. I think that was the first of several similar meetings I’ve been involved with over the years that stopped after the first 10 minutes. The meeting stopped and we went back to the drawing board - a new plan. Yup, the work of a dozen people over 3 months was all revised and ‘fixed’ in less than a day. It was just like magic.

Looking back at that event and a number of similar occurrences I often ponder the rationale of the approach. In my experience it’s pretty common. And I can almost guarantee that if we had presented a plan with a head grade of 1.4% the CEO would have insisted that the ‘right’ number was 1.6%.?

My conclusion after more years (I typed ‘tears’ but it was autocorrected to ‘years’) than I care to remember is that the purpose is part of a somewhat crazy and self-deceiving game we play with each other. It goes like this:

Planner: “I must make a plan that we can achieve, based on the best knowledge available. If we don’t meet plan/budget the boss of bosses and other important stakeholders will be angry. I’ll just add a little contingency to make sure we over-deliver. That way no one will be disappointed”

Geo: “this estimate is the basis of the operational plan. It describes the tonnes and grade we can recover and shows the spatial continuity which affects the mining process. I know I’ll be in -big- trouble if we don’t achieve what this model says. Every time we under-perform someone will blame this model. I’ll just be a little conservative to make sure the reconciliation is always positive”

Metallurgist: “Look at this production schedule. It’s crazy! All the metal is back-ended. We don’t see any real grade until Q4! Those geos and engineers must really not understand how this business works. I know they always deliver a different production stream to what they have planned. It’s never the same. Let see… if I allow for some reduction in recovery in the first few months, I can plan for improvements later in the year. That gives us more metal and yet it means we have time to get read for the Q4 grade. That’s a good idea. It increases the chance we can manage this lumpy schedule!”

Manager: “look at me, look at me, look at me… I’m not a technical person anymore. I know better.?I used to be the best at putting those plans and budgets together. I know exactly where those guys have stuffed things up. If we use there plans as they are we probably won’t make budget.?Hey! My bonus depends on hitting all those targets.?I’ll add a few lines just to knock off the edges and make sure I get paid”

CEO: “hmmm… when I was a GM I always padded my budget a bit.?I’ll just tell the, to make everything 20% better and then, when they find 10% it will look like a great result.?And to top it off I know we can do that extra 10%?- they just need to work smarter”

It’s an unspoken and delicate ballet. Some parts are conscious decisions, others are almost like muscle-memory, deeply ingrained and subconsciously influencing the entire ensemble.?

Beneath all the pirouettes the psychology of setting goals and targets changes behaviours. Set good objectives and you are more likely to get good outcomes. Set poor objectives and you will get unexpected and contradictory results.?I remember one KPI I was assigned… to increase the ore reserve by 100kt metal in 3 years. Easy! If there’s a large prize for delivering you make it happen… increase the reserve a bit each year, take the cash and run… if your morals allow.?

This dance of setting goals, making ‘stretch’ targets and designing KPIs is something it it easy to forget or dismiss but its effects can be dramatic. When was the last time a consultant was paid to say a project was NOT feasible??The implicit message in a feasibility study team (and often their bonus) is to make the project successful. Even the dog that everyone knows is a dog looks good when it have a fresh wash and a pretty bow wrapped around its collar.?

This is not new. It should not be news. But it’s another example of the difficulty of human systems. Despite our best intentions our own heads get in the way.?

If this is ‘old news’ to you I guess my question is… what are you doing about it? How are you ensuring that the hidden impact of goals and budgets do not bias your plans and estimates? If you recall my thesis that mining is a complex, interconnected and non-linear system, a transfer function that often resembles a black box when seen from the outside, quantifying this type of bias needs to be addressed across the board. Any single component that has a different risk tolerance is likely to perturb other parts of the system.

And, to keep flogging a dead horse…. How do these factors contribute to the whole question of Competence (with a JORC Code capital ‘C’)??When there is no single ‘right’ answer, no single ‘right’ estimate and when we know these cognitive biases exist in us all, surely the expectation of a singular truth is nonsensical??

Combine natural variability, data-scarcity, parametric uncertainty and cognitive biases into one complex system and it’s amazing we get anywhere at all!

While I personally believe the solution (in part) lies with recognising and quantifying variability and uncertainty I know how difficult that can be. I also know how unsuited our minds are at accepting probabilities as opposed to a singular truth. Doesn’t mean we should not aim high though!

Take bold strides I believe was the cry from at least one stakeholder…