Martijn Moret

As an (internal) data team you need to continuously show the value you are adding. You don't want to become the next "IT support" department, taking ages to solve requests and often without really solving the problem. How do you show your team adds value? By defining your deliverables, services and metrics to the "clients", and by measuring the effects of "data-backed management" vs the situation before.

A data team along with non-data savvy people on the business side doesn't work and only leads to division. How often have I not heard that any question to "the data team" only leads to answer that don't answer the question, after 5-6 weeks. Making everyone data-savvy (but specifically the leadership of teams and executives) is extremely important to get value from a data team.

In my experience there is a natural gap between dedicated data teams and the business. On a high level there may be understanding of business processes and goals but for in-depth analysis the business (unit) executive and expert are required. Centralised data teams run the risk to be completely disconnected from wealth of experience in the business, and just focus on their own priorities. There needs to be at least one data-savvy person on the business side to make the connection fluid. In addition, the data team needs to understand long term and short term data analysis needs. The business is not served if every question takes 6 weeks to answer.

This is a typical queuing theory problem! Start by getting people to into different queues for different types of actions: eg separate inquiries from tasks. Then use Data Intelligence to see the processing time of subcategories of tasks/inquiries and identify handling steps and optimize those by making them non-sequential. eg: a check-in consists of finding a booking, checking documents, filling forms, marking accompanying baggage, explain next step etc. A number of these can be done beforehand, or in parallel, or completely digitized.

The right tool to use depends on the requirements and how well you know the "behaviour" of the data. It starts with finding out what is in there. An easy (also for non-tech users) tool is DataSquirrel.ai which allow instant exploring, cleaning and visualising of data.

The starting point imho is finding out what exactly is the problem with the supply chain. Only than you can create a set of requirements to begin selecting a suitable software package. Is the struggle with lead times, inventory control, pricing, waste, material quality, billing, end to end reconciliation, etc. This involves talking to stakeholders and data analysis (with, most likely, data from many sources) covering the past year or so. One of the handy tools you can/may use in this initial step is DataSquirrel.ai as it combines, cleans and visualises data from all kinds of sources.