Data can drive innovation powering business change, but without careful management and governance of the collection of data, it can fail to deliver the promised value.
With data storage costs falling especially in the cloud, the default position with many manufacturers is to collect and store as much data as possible. If it doesn’t cost much to “just stick it in the data lake” then “why not?” after all, the business may need it to support a data science initiative in the future, so it is better to be safe than sorry surely and collect the data just in case? While the sentiment behind this hoarding of data can be understood, data is often collected without a defined purpose or consideration for the implications. Storage is not the only cost associated with collecting data. To realise the benefits that data can bring inevitably requires initiatives for security, quality, cataloguing, analytics to name just a few. If data is collected “just because” and not in line with a coherent data strategy, data won’t be the only thing being stored up for the future. Watch out for forthcoming headaches from how to manage data collected with limited purpose. Ask yourself this question ‘who wants to be the person that deleted some data that could have been useful?’
Measuring the value that data can bring is one approach to support the business case for collecting it in the first place, but defining this value presents its own challenges. Firstly, can the business agree on what these measures are, is there a consistent language understood, agreed, communicated and used throughout the enterprise? Is the logic behind KPI calculations implemented consistently? If there are decisions that need to be made when defining the value & meaning in business data domains, are there empowered individuals or teams with the authority to make these decisions? Using standards for data, internal and external (e.g. VDMA) can add significant value and can be used to break across organisational silos as well as silos of data. All these points can be addressed by an effective data strategy that looks holistically at processes, organisational factors, technology and Information requirements.
But building a data strategy that puts data at the centre of an organisation, takes education sponsorship and change, none of which come particularly easily and all of which will take time. With most manufacturers recognising the benefits of an agile approach to support innovation, a common challenge is often securing the budget to deliver a data project. If this data project is not part of a wider data strategy, having a senior stakeholder to champion the cause for data in this context will be essential. Typically, data has many cross-functional uses so being able to navigate skilfully through organisational and political issues in your organisation will be essential for addressing the often thorny subject of who pays for the project.
Treating data as a corporate asset is therefore not as straightforward as it might sound and being asset rich is not as easy as just collecting more (data) assets. Business value should be the watchword for every project. Start small, be very focused and deliver value quickly.
In a recent conversation an industrial manufacturer was describing how by using data effectively, they were able to increase the effectiveness of their production lines by a factor of 4 by using data to focus on the key factors of quality and accuracy. The short-term pain of collecting data for a well-defined purpose – fitting sensors to the production line to measure key performance metrics – and the resultant business value generated by doing so were easily demonstrable to the business decision-makers and budget holders.
In a virtuous circle, data can very often be the solution to the challenge of how to secure a budget for more data. Using data to tell the story of the business benefits of better data can be very effective. As a storyteller, you can only go so far with tales of efficiency savings but by bringing the production improvements to life through effective visualisation you can clearly demonstrate how a data initiative could pay for itself. However, as one automotive manufacturer agrees, an MVP project with a solid roadmap for improvement will be more effective than a “throwaway” proof of concept. In conclusion: collect data with purpose and deliver value and do it quickly and set this up as an approach that can be followed for future initiatives is a tried and tested approach for many manufacturers.