Look Before You Leap: Avoid Unintended Consequences with Your Logistics Strategy

What do you do if your logistics strategy is just not working? While there are ways to reverse course after a faulty strategy is deployed, you must first admit it was faulty. This never happens. It is much better to look before you leap and test against historical data before you deploy.

Testing is not only a good idea, it is crucial to developing a logistics strategy. It’s the key to avoiding unintended consequences. Without testing you are not going to know the full impact of the strategy, positive or negative, until it is deployed. Once the strategy is deployed, even though it may be negatively impacting some areas, it will be much harder to make changes to it.

When Standardized Data is not Standardized Data

To simulate potential strategies prior to deploying a strategy requires data. But it’s more than having data, it has to be the right data. This data needs to be complete, clean and standardized. We oftentimes see companies that have lots of data and usually it is the correct data but it’s not clean, standardized data. What happens then is they standardize the data by (incorrectly) removing data that isn’t in agreement with the entire set. We hear things like “that was a one-time expense” or “that load was an anomaly and that’s why it cost so much.” By standardizing in this way, the completeness is lost and the analytics are now subjective which usually leads to justifying the strategy that a person wanted to deploy. Using a clean, standardized dataset that is complete should guide your strategy or even prescribe your strategy instead of being dissected to justify the strategy.

Simulating the strategy on the historical dataset prior to deployment is the most accurate way to test. By applying the strategy to a clean, standardized and complete historical dataset, the company will see the exact effect their new rules will have. They can then compare these results to how they were performing historically to understand the financial impact, customer service impact, and any operational impacts the strategy will have. If the simulation reveals any negative impacts to the historical dataset, the company can change or modify the strategy to avoid those. 

Sins of Omission

When developing strategy, what you exclude is as important as what you include. The omissions are always related to the exceptions that occur on a daily basis. When using incomplete data where the exceptions were removed because they were deemed to be unusual shipments, the results of the analysis are also applied (incorrectly) to the exceptions. For example, if 5% of your shipments require expedited freight, this could account for 20% of your overall cost. Let’s say the company’s freight spend is $10 million, and the incorrectly cleaned and standardized data represented the remaining 80%, you would be analyzing $8 million. If the analysis of the strategy reveals a 10% savings and that is applied to the entire spend of $10 million, the company would expect to save $1 million. The reality is they would save only $800,000 because the strategy doesn’t apply to the $2 million in expedited freight.

Another example I’ve seen of unintended consequence is when a company creates a new way to price their freight such as changing the rate base for LTL. They model the results and find huge savings, sign the new contracts with the carriers, then try to deploy the rates to find out their system can’t support it. I’ve seen a 15% savings turn into a 15% rate increase. This would have been prevented if tested with the same constraints that exist in production on clean, standardized and historical data.

Do Not “Set and Forget”

“Set and Forget” does not apply to logistics strategy. After a strategy is tested and deployed, there is still more work to be done. You have to make sure you have a solid data foundation in place that maintains the data-driven approach. This is the governance layer that provides real-time reporting and alerting to ensure that you are achieving your goal and it is made up of two pieces:

1) Savings KPI—measures the “old” rate compared to the “new” rate and

 2) Lost Savings KPI—measures the compliance to the new strategy by showing the financial impact of shipments that didn’t follow the new strategy. The goal is to have zero lost savings which means you are maximizing the strategy.

Only Complete, Standardized Data Makes the Grade

The most important step to take prior to deploying a strategy is testing. But to make the grade and provide the insights you need to avoid unintentional consequences, the strategy must be tested against complete, clean and standardized data. This allows you to create a strategy that creates the savings you’ve been looking for.