Real life measures: Measuring efficiency in offices

Image of huge vintage office - metaphor for too many KPIs

 Who is this article for?

This article is a fairly technical article aimed at people who are looking to measure the efficiency or productivity of teams in an office-based environment, want a guide on how to measure efficiency and the system options for doing this.

 

Article facts:

Word count: 1320

Approximate time to read: 6 minutes

 

What is office efficiency?

The aim of most efficiency measures is to show how we are performing and how much room for improvement there is. An efficiency measure can:

  • Show us trends in our operational performance
  • Enable comparisons between areas (although with caution – small changes in measurement standards can make big differences in final figures)
  • Show us the opportunity and (with some arithmetic) the financial value of improvement
  • How to measure office efficiency

Using manufacturing techniques and applying them in the office environment is a very popular theme at the moment, as you can see from the popularity of Lean and Six Sigma courses in financial services organisations. One of the foundations of any process improvement approach is knowing how well you are performing at the moment – how efficient your operation is. In fact this can be surprisingly hard to implement and many organisations either give up or don’t even try. This article walks you through the options and some of the pitfalls for the unwary.

 

There are a number things we need to understand when we measure office efficiency:

1) How much of the available time was used for activity

2) How much resource was available

3) How much of the active time was used for “value-add” tasks

4) How many “things” or tasks were performed (by task type)

5) How long those “value-add” activities should have taken

 

What efficiency calculation should we use?

You can use OEE as a measure. The advantage of this approach is that it is very commonly used and it is a “standard”. The downsides are that:

  • The language used can feel more applicable to a manufacturing operation than an office
  • Explaining the measure can be complicated
  • The “manufacturing” association with the measure can be counterproductive in some office environments

Another approach is to use a “simple efficiency” measure. For a full description follow this link. Put simply we work out how much of our day was occupied with “value add” activity. You then divide that value by the total number of hours available. Here’s an example calculation…

 

Non-defined activity

Q: What about the situation where I have a task that doesn’t have a standard time, like an unscheduled meeting with a customer. It is value-add but doesn’t have a standard time!

A: We can tackle this by either:

1) Excluding that time from the measure, as we know that we won’t hit 100% and it’s just one of the things that we accept we have to do and will impact on our efficiency. The good thing about this approach is that it presents this task as an improvement opportunity. In the case of a customer meeting it might prompt us to ask why that customer is contacting us and think about ways in which the customer could “self serve” or we could anticipate their needs, removing the need for the meeting.

2) Just take the total time taken for these “undefined” tasks and add it in to the calculation as “value add” time. This generally makes staff happier, as they feel like they are getting credit for activities they can’t avoid or control – but it does hide potential improvement opportunity (i.e. how could we reduce that “undefined task” duration?)

 

Where it gets complicated

Unlike a manufacturing process, getting a measure of how much work we do, how long it takes and the type of tasks were done in an office environment can be tricky. You would think that the progressive movement of back office processes onto computers would help. From experience - it doesn’t.

The reason? Most back office areas use multiple platforms, so it can be incredibly difficult (sometimes impossible) to piece things together. This leaves us with the following options:

1) Activity sampling

2) Time sheeting

3) Desktop activity software

4) Forcing everything onto one platform

 

Here’s a quick analysis of each option:

Activity sampling: A 3rd party, e.g. a local Six Sigma practitioner, sits in an area and every few minutes notes what proportion of the team are visibly engaged in apparently work-related activity.

Pros: The simplest, cheapest and quickest method. Based on observation rather than self-reporting, so less open to manipulation.

Cons: Needs to be done over a decent period, otherwise the presence of the observer is likely to disrupt the measurement. It is impossible to tell if the “apparently work-related activity” is work-related or is value added. The method is only suitable for “snap-shot” measurements; it’s not a day-to-day method of measurement. It can be viewed as “spying” on the observed teams.

 

 

Time-sheeting: Team members note down what tasks they did and how long the task took. The can be recorded on paper, a spreadsheet or through a dedicated timesheet system.

Pros: Simple. Self-completed by team, so less risk of perceived “spying”. Can be used to gather other interesting information, such as rework and dirty data (incomplete or inaccurate date being fed into a process) – a great start for process improvement.

Cons: Data must be collated manually if it’s a paper-based time sheeting system – taking up precious process time and annoying the staff who are required to do it. Volumes and times are open to manipulation and genuine mistakes. Teams are normally very reluctant to record idle time or non-value-add activities.

 

 

Desktop activity monitoring software: A small desktop application sits on the desktop where the process is being performed, reporting activity, duration and process steps to a database located on the network. The database can then be interrogated with a business intelligence tool, generating efficiency figures and analysis.

Pros: Can be fully automated by identifying “trigger” activities – potential zero impact on user. This approach can be highly accurate. Fast to implement (days possibly weeks) compared with major system change options.

Cons: End-to-end measures cannot be implemented as the tool is “unaware” of processes linking multiple users. There are employee relations issues to be addressed in some environments before staff will accept this kind of tool. Often will not record off-PC activity such as phone calls and meetings.

 

 

Forcing everything onto one platform: This is the most appealing to anyone with a tidy mind. It involves sweeping up the selection of applications and replacing them all with a “super system”. This will normally involve a class of tools referred to as BPM (sometime also called Workflow, although technically incorrect).

Pros: Has the potential to bring everything onto one system. An extra benefit is the process redesign/improvement opportunity that this presents. The data is then all in one place. There’s the potential to develop end-to-end measures.

Cons: It can be massively expensive, invasive and risky to do. Often takes months/years to do in large complex environments Many BPM systems intentionally make the data on the system difficult to access to avoid compromising the operational capability of the system through badly constructed queries – many require that you “mirror” the data onto another server before you can access it for efficiency calculations.

 

 

The practicalities

When choosing your preferred options you need to be really clear about what your objectives are. If you are simply looking to allocate costs to a number of different internal customers then a simple time-sheeting system may be fine. If you want an end-to-end measure of efficiency in a complex multi-stage process then you may have no choice but to go with a heavyweight BPM solution. Looking for a straight forward efficiency and capacity measurement? In that situation my preference is a desktop analytics solution like OpenSpan Desktop Analytics, Verint Impact 360 or NICE RTAM.

 

For help in evaluating the best options for your situation contact Bernie for a free 20 minute investigative chat.

5 Comments

  1. Avatar James Lawther on May 2, 2012 at 05:06

    Very interesting article Bernie, thank you for it.

    The question I would love to know the answer to is should you try to measure everything? Or just measure what you can get?

    James

    • Avatar Bernie on May 2, 2012 at 08:01

      An interesting question. I always think it is worth going with what you can get (even if it isn’t comprehensive) unless it is rapantly misleading. One potential risk is where we cannot distinguish activity from value-add activity. If you have a significant amount of waste and rework hidden in there then it can be really risky to draw any conclusions from activity levels. Don’t be put off though, some meausurement is always useful in an area that has none, as at least it starts to build the discipline and familiarity that you will need for a more thorough set of measures in the long-term.

  2. Avatar Larry R. on May 16, 2012 at 19:31

    Understanding all the time variables in my opinion is critical to understanding efficiency of an employee or project. Either way, thanks for the break down in your analysis.

  3. Avatar Suraj Timilsina on July 23, 2013 at 06:42

    I am MBA finance student and have go through thesis writing which need an efficiency measurement of the financial derivative market. What model of efficiency do u suggest me to use and why?

    • Avatar Bernie on July 24, 2013 at 09:51

      Hi Suraj,

      Most efficiency measures are about measuring actual performance against best-possible-theoretical performance. For manufacturing this is fairly obvious – actual output compared with bottleneck-speed x duration. For derivatives it’s going to depend on the specific kind and whose perspective you are looking at performance from. Do you have details of any particular efficiency models you are looking at? What is the objective of measuring efficiency? Perhaps with a few more details we can help with your thesis.

      Bernie

Leave a Reply Cancel Reply