Lies, Damn Lies, and Statistics: Logistics and e-commerce

So what is to be done?

Start with the core business

When we use 101 statistics to measure the health of our operations, we quickly find ourselves in a state of analysis paralysis. Every company and operation has unique value propositions, and therefore KPIs that are more significant to the business than others. That is, the universal and most important thing is also the simplest: what did I tell my customer I would do and how well? In the world of logistics, this is also known as timeliness, leaving aside the complexities of the perfect order and the associated inventory and pickup accuracy.

There is no e-commerce customer who doesn't want to know when their order will be delivered when they check out.

How well the supply chain meets these expectations is more important than any other data and should therefore be known and understood by everyone - at least in the logistics department.

Prioritize in two categories

.

Next, prioritize into two categories:

1. On which orders can we take immediate action to avoid a critical Customer Experience and
.
2. which orders are more useful for post-mortem analysis?

Orders that are currently being processed and may fail are much more critical. These shipments can represent opportunities and problems in equal measure. With the help of various trigger points over the life of an order, we can see where risks lie or even where an order will fail before the customer even notices. If we have these data points and react to them, in the best case we can actively intervene and avoid a critical customer experience. In the worst case, we can communicate with the customer before a disappointing customer experience to soften the blow.

Now to the retrospective, but still important, post-mortem analysis. Again, we don't want to overload the statistical data for quick review of operations, but the view should be broadened. Again, we want to take advantage of both objective and subjective information. Objective information is the what, where, and when of call center volume and other capture points related to the duration of a job (i.e., how many weather events can Atlanta possibly have?). Subjective information is customer feedback in the form of surveys, driver ratings, and detailed phone conversations.

Both objective and subjective information are important. But the most sophisticated KPIs are a poor measure against an unhappy customer base if one KPI doesn't support the other. Then that would mean the wrong objective statistics are probably being used.

But this doesn't mean there isn't a need and tremendous value for deeper insights into operational data. Especially in rapidly evolving and low-margin industries like e-commerce, there are opportunities and benefits that can be found in the depths of statistical analysis. Consultancies can provide comprehensive insights and guidance. Academic institutions offer cutting-edge research, and business intelligence tools can shorten time and reduce costs,

...and that's how millions of data points make sense.

But in terms of managing the day-to-day operations of e-commerce logistics, with its myriad demands and time constraints, a direct view of a few Key Service Indicators has the most immediate and positive impact. Indicators should be prioritized based on proactive, transparent themes and subjective customer feedback.