DCwise insights - How to measure your warehouse operations?
Updated: Jun 23
DCwise.eu shares in this post ideas on how you can improve your solutions ➟
When running a warehouse operation - due to its nature often a very complex environment - there is a major benefit in setting up the right measurement methods.
In this post we give some guidelines & pointers on how to bring structure in the measurement and monitoring development challenge.
1️⃣ The Number of metrics
Before going into too much depth an important element is "the number of metrics". When monitoring a basic operation vs. a complex operation this should be reflected in the number of metrics used.
Even with a complex environment realize that a person can only take in a limited amount of information; and that it's not about the metrics, but the ultimate goal is to trigger the right actions.
Ideally the view on the measurements mimics the dashboard of for example your car;
- Where the majority of screen is your view on the actual processes (the windshield...)
- A selected set of metrics is key and are most clear (like the speed...)
- Some metrics are smaller and only trigger attention when they go out of bounds (fuel...)
- And some metrics should only draw attention when there is an issue (tire pressure, etc.)
Note that the reference to the windshield can be seen as "going on the floor" or "having a live view on the operations" which is never can be replaced with just metrics.
One way to organize the monitoring is to have on the top level a limited set of key metrics (opposed to a space shuttle looking cockpit) and on a lower level there are more metrics. If need be one can drill-down from the top level to the lower level to these more detailed ones.
As of now we will refer to the monitoring of the set of measurements as a "dashboard" or "cockpit".
2️⃣ The Content
Next step is filling-up the dashboard with good metrics.
Essential element in defining what needs to be part of the dashboard is the process-control and the link to the (corrective) actions.
What we suggest is to take the following steps:
Start with "listing" which information could be interesting for your operation. Note that we are not yet taking about data!
Immediately add what types of action are driven by this information.
Metrics without action need review and either refined or removed.
Only in this step start looking at how the information can be generated.
In a next step indicate the relevance of the metrics (see ️⃣1️⃣) to determine on what level they should be displayed.
After development of the dashboard validate if the metrics are being used and start the improvement process.
Even when the information can not immediately be harnessed leave the element in the dashboard such its clear that it should be gathered to drive a complete overview. Also don't be afraid to start manually populating some metrics or without all being filled in.
3️⃣ The Data
Probably one of the hardest elements is getting the right data and turning it in proper metrics or information. This while most data are not generated with the analysis in mind.
Either you have too much data, or its not available in your "reporting tool", or the specific data needed is not available at all, or the challenge is complicated while the data is polluted, full of old information, unclear, and so on.
In this the tools that can assist on the market have grown massively recently. E.g., MS Power BI that allows to bring together data from different sources in an easy way, allows for data purification (or clean-up) and supports at the same time proper visualization in e.g., dashboards or any needed ad hoc analysis. But also "Excel", or other basic tools, up to whiteboards (being manually filled) can work in this respect.
4️⃣ The Visualization
Key after all this work is to turn all into a proper visualization that can be easily understood by all. See in this also our post on visualization https://www.dcwise.eu/post/dcwise-insights-how-visual-works-best-4-basic-tips-for-warehousing?
One proper way to visualize is - as mentioned before - through a "car cockpit" approach. Here the metric is displayed in way dependent of the goal, e.g., driving attention and consequent action.
Some examples - and make systematically the link with "when" and "how much" attention should be drawn and if an action should be triggered.
- 🔴🟠🟢 Usage of coloring: off / on with on being green, red or blue dependent of the impact (e.g. "being out of gas”)
- 📈 evolution: some elements should be seen in relation to time. As such outliers, or future ...
Additionally, make sure to communicate openly on what is happening. See it as a scoreboard... in this the home-team needs to know what the score is.
Once you have a first concept; leverage this to validate its functioning. Next go through iterative improvement steps where the concept is matured.
👉 In the end the metrics are meant to facilitate the management of the operation.
So, watch out for the very common trap of falling into the "over monitoring" or "over analyzing".
👉 Note also that more and more systems can be setup that they become self-driven, including automatically taking the rights corrective actions. Think in this e.g., of steering resources away from over-crowed areas, or pulling in more when the workload is high &/ the due is not made. In this the more you grow in these area's the better the management can become, and less stress lies on the metrics.
More interesting information can be found @ our blog.
☞ To further enable your transformation journey and understand better the potential feel free to get in touch in!