Over the past few blogs
, we’ve explored creating a consistent digital thread to connect every aspect of the value chain. Integrating data from R&D to customer experience and the after-market can deliver stand-out returns for auto companies. But how to go about it? We’ve seen how department-level analytics add some value, but also can allow costs to multiply as systems and models are duplicated. We have also discussed how integration of data across the whole enterprise can have radical impacts on agility – enabling profitability and growth. But building an inter-departmental data set is hard.
Relationship status = ?
The problem that many face is organisational. The short tenure of chief data and analytics officers often forces them to focus on fast iteration and immediate ‘wins’ rather than foundational work of building governance, data quality and an integrated data platform. It's like being trapped in an unending series of first dates and never working on a longer-term relationship!
Individual analytics, AI or data science projects are sexy, look great on CVs and can briefly grab the attention of the board. But this constant pressure to ‘look good’ distracts from the more mundane but critical investments in the foundations of longer-term success. Architectural work that focusses on the availability of high-quality data will define consistent and repeatable analytics as a driver of long-term value in the business. Finding Mr or Mrs Right takes commitment.
Step by step
As with any relationship you need some early wins. CEO’s and senior executives are bombarded by data – what they are looking for are answers. They know the data is out there in the organisation, but they don’t know how to find it. Showing the rapid value of connecting a few disparate data sources to deliver an answer that can drive new revenue or cost savings will quickly win over any CEO. Seeing their investments pay-off builds trust and creates the conditions to extend and expand commitment. Encouraging them to invest, piece by piece, in a data architecture that links data into insights will demonstrate how integrated data is the key to the closed-loop, agile and reactive enterprise that CEOs want to create and win you that second date.
Don’t bring old baggage
No one likes to be roller-coasted into things, nor do they want to feel trapped. The low cost and agility of cloud architectures are among their most compelling attributes. But diving into a new cloud relationship with all your existing baggage can undercut both cost and flexibility advantages. Haphazard moves risk replicating old inefficiencies on new virtual hardware. At Teradata, we have identified six critical capabilities to consider for a modern cloud platform
. Integrated data management is one of these six capabilities, which directly impacts agility, responsiveness, and security to ensure a smooth migration and longer-lasting relationship.
Old baggage in the form of inefficient and duplicated systems also effects sustainability – which is rapidly rising up the CEOs’ agendas. Scrutiny on carbon footprint is rapidly moving to include IT’s contribution to environmental damage. As automotive companies work towards reducing waste in the physical world, they should look to do so in the world of compute power – for which corporate demand is growing rapidly. An integrated data set supported on a platform with dynamic resource allocation and workload management is far more efficient than individual, separately managed systems.
A recent Google paper Everyone wants to do the model work, not the data work
highlights the danger of paying too much attention to models and not enough on getting high quality data ready to use. We’ve all fallen victim to propositions that look good at first but have nothing behind them. The Google paper highlights this issue for data modelling. ‘Cascades’ of issues stemming from poor data foundations have unforeseen, but significant, impacts that go well beyond simply changing the model. CEOs of any automotive business relying on predictive models need to be assured of the firm data foundations of any model, however ‘flash’ they look. They will thank the data architects who can deliver assurance, confidence, and reliability.
Making a good first impression is important, but strong relationships are built on concerted efforts and mutual trust. Just like in relationship you cannot build a strong foundation if you skirt around the hard problems. Once trust in the data and the data platform approach is established in one area of the business, the digital thread will grow longer, stronger and more valuable with every new integration. Showing how the digital thread impacts the measurements important to the CEO (profits, growth, corporate reputation) will provide the CDO with the opportunity to get beyond the first date and build the data foundations for long term value.
For our next, and final blog in this series, we’ll look at how data scientists can pull upon these same digital threads to build data product factories that consistently supply the business with added value from machine learning and AI.