Watch your KPIs and use your rudder!

In order to get the most out of Big Data and AI corporations must implement and nurture a data-driven corporate culture. This takes time and abundant leadership to establish, nurture, embrace and grow. Becoming data driven is not just an issue of spending money in analytics software and graphical displays, it is a thoughtful, immersive and iterative process that largely depends on being able to trust the data. Let’s touch a bit on the “iterative” part of the process.

As part of my blog series on how to “Transform Your Corporate Culture to Best Use Big Data”, below are three suggestions on how to use KPIs and metrics effectively and successfully through a thoughtful iterative process, to increase trust in the continued monitoring and use of KPIs.


One of the business “cool words” (I can’t say “trending” because ‘what is this internet thing?’…) of the 80s and 90s was “KAIZEN” – meaning “change is good”. While I don’t want to dive into its three pillars, I do want to focus on one of its five elements; the “suggestions for improvement”, in the context of KPIs that we may have selected to monitor our business’ performance.

On the previous blog I proposed that in selecting KPIs we must:

1. Understand the audience to whom you present them.

2. Define them mathematically and verbally.

3. Identify the LEVERS that impact (can change) their behavior.

Identifying the LEVERS that impact KPI’s behavior is of particular interest because they will tell you “what” and “where” to look, “how” to make positive change, and “how much” change to make in order for the organization to absorb it well. These are critical things to know, and to act upon in order to get the most out of the business.


Fact - Most if not all office functions are related. What a corporation does sourcing raw materials or components, selecting vendors or partners, has an impact that ripples throughout the organization. That being said, most levers are inter-related. If you want to validate the impact of any one action on a lever, select only ONE LEVER at the time.

Here is a scenario – You are in a meeting where the team identifies a problem with a project that is beginning to run late, but is still in a “recoverable” place. The folks in procurement say that they will go to the one vendor that has been running late on their delivery (the lever) as it may be impacting the production line. The folks in engineering say that they will go back to the drawing board to look for an alternate supplier (a lever) or to change the assembly and test process (a lever) of one of the product sub-systems.

The above are all good actions to take, but they need to be prioritized and also “isolated” such that their individual effect can be measured and proven. If you don’t use the levers carefully, you may fix the problem and yet not know how, thus leaving yourself open to suffering the same problem in the future.

Make sure that when you pull/push a lever, you can directly isolate cause and effect. The more you can do that, the better (clearer) “lever” you have selected.


To avoid collisions at sea, vessels on a “CBDR” relationship (The Titanic and the iceberg had one of those relationships) are supposed to change course in an obvious manner. When you pull or push on a KPI lever, you should give enough “input” such that you see the change. Further, you should be able to see the change that you expected.

Here is a scenario – You are running a special offer on your web sales and decide to run a promotion on your new product offering a 5% discount by using a coupon. After a week, sales are relatively flat and far short of campaign goals, so you re-run the campaign offering a 7% discount now. Another week goes by and that 40% change has zero impact on sales. You consult with your product management team, they with the senior leadership and you decide to adjust the discount offered to 25% for a short and limited time, and only for the first 100 sales (for example). All of the sudden sales take off and you are off to the races.

The last change delivers results (a change in sales pattern) and it was as expected (sales increased). Further, the change is such that when the input expires (the first 100 sales are booked) you can see the impact on sales (cause and effect).

Surely the above is oversimplified, but it is to make a point. Making measurable changes can be daunting and rightfully so. Remember that the point of making these changes is to “get better”. Once you have selected the best lever to pull, make changes timely and measurably


When you pull or push on a KPI lever take the time to critically observe their impact every single time. Not once, not once every quarter, or year – always. Why? So glad you asked –

Here is a scenario – You process orders for a given product in 14 days, order to shipping. During the busies part of the year, your cycle time creeps to 18 days because of order volume, you add 15% additional workforce like last year expecting smooth sailing, but no luck; you barely make an improvement. Unfortunately, you were distracted by other issues and did not realize that the change had not had the previous and expected impact until 7 days later.

If you don’t observe the change (magnitude, time, direction) you can’t validate that the lever is delivering the expected results. Worst yet, you fail to take corrective action and miss goals.

When you use a lever, document the result every time, and compare it to historical data. When you do this, you will validate the usefulness of the lever and will be in a position to validate the inter-relationships of the business we spoke about. If the results are not the same, inspect the change and adjust KPI or levers accordingly.

Use monitor your KPIs and use your levers as frequently as needed. When you make timely and meaningful corrections those around you will notice and believe in the processes and success metrics you have established. This is one of the key elements of attaining a data-driven corporate culture.

Use your tools for more than having a cool screen saver!

5 views0 comments

Recent Posts

See All