Break out the popcorn, dim the lights, get yourselves comfortable and settle in for the first instalment of something so spellbinding, something so insightful and riveting you won’t be able to leave your screen. At journey’s end you will experience that lightness of being that is truly transcendental. You will realise that you have found a deeper understanding of how the world works. Alternatively you can read a post about capacity management (CM for short).
Where to start? Well, in true Hollywood style, let’s start at the very beginning, a very good place to start. When you read you begin with a,b,c, when you count you begin with 1, 2, 3 but preferably 1 to 3. If we’re to get anywhere, we need to be able to count. We don’t need to learn from the mathematicians and theoretical physicists about how to make a+b+c = z, although it may be useful to leverage the fine calculus skills honed from the noble professions of economics and accountancy that help make 2+2 = 13.
In simple terms, CM is about trying to balance how much work there is to do (demand) with how much work the team can do (capacity) and what to do when they don’t balance.
And that’s basically where the simplicity ends. This week we’ll focus on the challenges of trying to pin down precisely how much capacity you have/will have. Next week we’ll look at some aspects of the demand side.
A typical text book definition of capacity will say something like: “The sustainable amount of output that can be produced under normal operating conditions over a period of time”. So in true Wodehouse style where “Every day you seem to know less and less about more and more” – here’s the top 10 pain points that will keep you scratching your head, when moments earlier you thought you’d got it nailed:
- Units of Measure: it may sound obvious but balancing the equation requires you to be able to count the demand and capacity and they need to be defined in the same way. The most common problem is that the sales or product teams who prepare forward forecasts, may only forecast in $ not units. But even if you get a forecast in units it may be different to how you’re measuring capacity. For example, if your demand forecast is based on the number of customers who complain, and you measure capacity in terms of the number of complaints, there’s an obvious mismatch where some customers may lodge multiple complaints or multiple customers lodge the same complaint.
- Duration: this may also seem obvious but any definition of capacity requires you to be clear about the measurement timeframe. If you are forecasting demand as payments per month but measure capacity as payments per day you need to agree on a common duration and then convert one side of the equation into the other. Simple as it may sound, allowing for the different numbers of working days per month, public holidays that are in different months from year to year eg Easter can have a material effect.
- Service Mix: most service teams produce more than one type of output so it’s rare that you can aggregate the things that get done and make it meaningful. Similarly for a manager that looks after a diverse range of teams outputting things as diverse as credit assessments, processed payments, resolved enquiries, updated account details etc trying to add these things up and come up with a meaningful number is nonsense. So the solution is to ignore the definition above and count inputs instead e.g. how much resource you have available to deliver the output required e.g. # FTEs
- Multiple Input Resources: most teams require a range of resources to deliver their output, each of which has their own capacity limit e.g. # of desks, # of computers, # of telephones, amount of bandwidth, # of people. So you may have worked your charm and caught the company accountants at a moment of weakness and they’ve said you can have more people, but if you’ve nowhere to seat them, or there aren’t enough phone liens you’re no better off.
- Time Horizon: all capacity is flexible i.e. you can wind it up and dial it down it just depends on the time horizon. It’s pretty easy to ask your team to stay back for an hour or two every now and again, it’s another thing to find new premises because you’ve run out of space. If demand rises by 5-10% most teams can respond promptly, 15-25% is more challenging and you may need more people. 50-100% increase will probably require a whole lot more infrastructure and navigating the vagaries of the investment planning cycles, so it’s critical that both demand and capacity side of the equation are talking about the same time horizon.
- Leakage: there’s a vast difference between theoretical capacity and available capacity. You might have your a full complement of staff available but you have to factor in training, team meetings, unexpected CEO visits, rework, complaint handling, training new hires, absenteeism, systems latency and outages. Some of this may be deliberate and known e.g. targeting a particular utilisation rate (time spent on core work) to ensure staff are fully trained and briefed; other leakage factors can be quite random subject only to Murphy’s Law.
- Individual Productivity: given we’re now measuring capacity as inputs e.g. how many FTEs we have, you need to understand individual productivity to be able to calculate capacity. We know that “all animals are equal, but some animals are more equal than others” but not only does the productivity between individuals vary, the productivity of each individual will vary from day-to-day. There’s a great saying “don’t waste a crisis” – if you want to see what your people are truly capable of producing i.e. their maximum capacity, check their productivity the day after a crisis e.g. when you’ve been in BCP mode, struck by a bout of sickness etc. You’ll be amazed at how much more they can produce. The issue is, is it “sustainable”. So, if we have no idea what we’re going to get on any given day, how can we come up with a capacity measure?
- Standard Time: To turn work into hours or FTEs we have to multiply the number of work items by the time it takes to process them – this is typically referred to as standard time. Sounds pretty straightforward, but when no two items are the same and the rate at which the service consultant operates varies (see 7 above), standard time is anything but standard.
- Output Specification: another interesting trait of services is that it’s a common, but false assumption, that the specification of the service is consistent. In reality, particularly in peak periods, we may change that specification just to get the work done and avoid the ignominy of backlogs or missing a customer deadline. For example, the team may take shortcuts where there is a hard deadline e.g. in payments processing the team may skip some of the checks. If you measure capacity in the last hour of the day, it’s different to the first hour of the day with the same number of people in the team!
- Forecasting: if all of the other pain-points were not enough, knowing how much capacity you had yesterday is not particularly useful. It’s more helpful to know how much capacity you will have tomorrow, next week, next month, next year. So you now need to forecast and, as Neil Bentley once pointed out to me, “there are only two types of forecast: lucky and wrong”.
At the start of this piece, I said “when you count, you begin with 1 to 3”. As I’m sure you’ll now appreciate, this was to emphasise that to measure capacity is challenging, at best. Expecting to get to a single number is setting yourself up for failure. It’s much better to understand the likely range (similarly for demand) and then you can plan around it. For example, if your demand forecast says you probably need 10 people (at standard time) available next Tuesday but it could be as few as 8 or as many as 12. While your plan A is focused on ensuring you have 10 people available to do core work, you allow for the option of pulling in two additional trained staff members or finding extra activities for the two surplus team members if not required for the core work. Then it’s simply a question of actively managing the unfolding realities of the day’s demand. Whilst this may appear as challenging as trying to catch smoke – the only thing I can say for certain is that you have more capacity than you think you have!
So, to avoid accusations of Schadenfreude, the next two posts will focus on how to act on this in practical terms. Next week will talk more about demand and particularly demand profiles, so you can set up your base team. The following week will look at some of the strategies you can use both to reduce the implicit variation and manage the mismatches effectively.
In the meantime, until the next gripping instalment, I’m sure there are many other points to consider and I would love to hear your thoughts and your feedback.