Most large operational teams undertake change initiatives to increase the efficiency of their processes. But does making processes efficient help the organization achieve its goals? Possibly, but may not as well.
To understand this better, let me go to the root of all business objectives – to make money. To do more of that, they must generate their outputs (could be products or services) faster and cheaper. But in many scenarios, faster means costlier and cheaper means slower. So, a balance between them is generally sought after. While searching for this balance, most organisations target the wrong measures.
Efficiency and Productivity are the most common performance measurements – both are either incorrectly or interchangeably measured – without realizing there is significant difference between them. Even very experienced heads of operations in large organizations sometimes ignore or misinterpret this difference.
What is Efficiency?
Efficiency is the output of something in comparison to its maximum potential.
Simply stated, it is the ratio of actual outputs generated to the theoretical outputs that could be delivered.
For example, in a health insurance claims settlement process, the employees can verify authenticity of submitted bills for 100 claims in a day, but in practice bills of only 80 claims are verified, the efficiency of this activity is 80%. This method of efficiency measurement provides good insights when measured at an activity level. Extending the same to the process level, may give conflicting insights. Continuing with the same example, if theoretically the company can settle 100 claims in a day, while actual settlement is only 80 claims a day, would the efficiency still be 80%? It would not, it would be possibly much higher, because all the other dependent factors in the end-to-end process were taken care of to settle 80 claims.
What is Productivity?
Productivity is the rate of output that is created for a unit of input.
It is used to measure how much output one can get out of a duration worked or for an unit of investment.
Using the same insurance claims settlement example, if the company settled 80 claims per day that is the productivity of the process. If the number of claims settled increases to 85 per day, the productivity has increased by 5%. But this does not necessarily mean that efficiency of the process has increased.
Let us use some visuals to understand the difference between Efficiency and Productivity better. A simple linear end-to-end process is depicted below:
This process assumes that there are no multiple touch-points and no wait times between the participants. In this case, the total processing time for the end-to-end process is the sum of processing times of the individual activities. Even if the efficiency of each activity improves by 10%, the efficiency of entire process also improves by 10% and so does the productivity. This is because the production is very linear and not dependent on factors external to the process.
Now, consider a more complex process as below:
In this process, there are back-and-forth between process participants and lead times in some activities. The efficiency of each individual activity can be improved in this process too. But this does not result in a proportional improvement in delivering the final outputs. In fact, even if the efficiency of each activity is 100% (maximum), the process could have poor productivity levels. This is because of the dependencies on factors external to the process, which are not proportional to the efficiency improvements within the process. If the complexity increases due to more changes in the variables. And therefore, the productivity as well.
Current ways of measuring performance:
There are two very common metrics used for measuring efficiency and / or productivity.
- Rate per Hour (RPH)
Number of times an activity is performed in an hour. Higher the RPH, better its efficiency and therefore the performance. For example – the number of invoices verified in an hour, number of follow-ups made with a hospital in an hour, etc. This is usually an average, calculated as “the total number of instances of an activity divided by the total duration in hours”.
Pros: Good when measured at an activity level. The average times are indicative of the speed at which the activity can be done, and whether it can be improved on an ongoing basis. This also can be tagged to individual performance for improving their skills and identifying training needs.
Cons: The average values may be skewed due to wide range of times, including a few very good or bad outlier times. The benchmarking of an ideal / potential time is missed out in this measure.
- Unit Time (UT)
The time taken to produce one unit of output over a specific period. The lesser the Unit Time, the higher is productivity. For example – the time taken to settle an insurance claim. This is also an average, calculated as “the total time taken across all the activities of the process (in a specific period) divided by the total number of complete units produced in that period”
Pros: Good when measured at an end-to-end process level. The average times are indicative of the turn-around times for the output and help to understand the dependencies on external factors can be minimised. This can be tagged at a product / service level and analysed in conjunction with lead times.
Cons: The average values may be skewed due to wide range of times, including a few very good or bad outlier times. The lead times may tend to be ignored, thus missing out on finding ways to avoid the dependencies and re-work. Cost factor is being ignored by considering only time.
Suggested ways of measuring performance:
- Efficiency and Productivity are great measures, but they should to be measured jointly and severally to make best use of what the data says.
- Efficiency should be measured as actual performance in comparison with a benchmark performance. The benchmarking should not be theoretical but based on historical data. It will be even better if the historical data was referenced not on averages, but by using percentile / quartile performances so that the better performance can be benchmarked (I will write about this concept in a separate post).
- Efficiency should be measured at activity level, and not aggregated at a process / product level. For example, some activities themselves are waste and causing the process to be inefficient. But as long as the waste activities are being performed (for whatever reason), they can still be performed efficiently.
- Efficiency appears to be a good indicator for measuring individual performance. This does not take into consideration the rework and external dependency factors. Removing or reducing those factors can itself be a performance measure for other teams such as strategy or product.
- Productivity should be measured both in terms of time and costs. Sometimes increase in productivity by time may result in increased costs or vice versa, so a balanced approach needs to be taken.
- Productivity may not be a good measure for comparing performance over consecutive periods, as the results may vary by season or month or due to unknown, uncontrollable factors. However, it can be a very good measure for comparing similar datasets. Say Q1 2017 vs Q1 2018, etc.
- Productivity can be used to understand what factors are controllable by the operations, and those that are not.
The Operations Think-tank must analyse and determine which processes are to be measured by Productivity and which activities for Efficiency and balance these measures with other key metrics at a department, company, or industry level.
Note: As you may have noticed, this article focused on the what and how of the said measures only, and not about how to implement the changes using these measures. I have covered the tools that can be used to manage the process changes in another article – Common Tools used in BPM.
