Skip to main content

Common Data-Related Challenges in the Transport Sector

Pre-conceived notions in the transport sector are causing data challenges that can prevent organisations from innovating, becoming more efficient, and treating data as an asset in its own right.

Bolly Williams


When most people leave their homes nowadays, they assume that the journey planner apps on their smartphones or vehicles will get them to their destination in the safest and most stress-free manner possible. This new way of travelling is commonplace thanks to the proliferation of advancements in AI and data over the past few decades.

Yet, many transport organisations are struggling to cope with the veracity, velocity and volume of this data. Many do not consider data to be a tangible asset in its own right, or truly understand the value of being insight-driven to achieve growth and improve efficiency. In fact, 45 per cent of organisations measure the success of data and analytics primarily only as part of wider solutions and functions, and fewer than 10 per cent of organisations have a consistent and systematic mechanism with clearly defined metrics to measure data and analytical return-on-investment - according to the latest edition of our Global Insight Driven Organisation survey here . Three of the most prevalent reasons for this are due to challenges related to data strategy, data integration, and data quality

Data Strategy: Go from the business problem to the data and technology, not the other way around

Many data leaders want the latest piece of AI-powered technology that solves all their business problems. They believe that these ‘silver bullets’ will help organisations with their every data need - from privacy to governance, all the way through to architecture and modelling.

Yet, a lot of organisations end up not leveraging these technologies to their full potential and are often no longer using their ‘silver bullet’ post-implementation and rollout. That is a lot of capital expenditure without delivering on-going business value. Much of this is because organisations tend to downplay the critical importance of cultural and business change in the success of any implementation - but all too often, organisations have not laid the groundwork that is required well before implementation.

A coherent data strategy is imperative to provide the guardrails for ensuring data and AI initiatives remain on track to deliver business value against stated objectives. The key questions your data strategy should consider are: Whatcrunchy questions** are the business trying to solve; what data is available/required to solve them; what internal governance and competencies do we have/need to manage the data throughout its lifecycle; and what technologies can support a number of strategic imperatives for long-term value generation? The point is to take a strategic rather than tactical view of your organisations’ data construct.

We recently helped a transport client make sense of the plethora of tools and technologies on their estate. They had tools that were not being utilised - ultimately because these tools were “solving” business problems they no longer had. We went back to basics by identifying the organisational strategic priorities and assessing the current challenges with the existing technology set-up – combining these to inform the requirements guiding the future state architecture. With these guardrails in place, we were able to create a list of short-, medium- and long-term initiatives to address the challenges, supported by a technology improvement roadmap. In summary, we had to turn how the business thought about technology on its head in order to arrive at a sustainable business-led strategy for tech and data.

Data Integration: Tackle it early and lay the foundations for scalability

The sheer amount of data that organisations have to handle on a periodic basis means that data is now coming from a wide variety of different structured and unstructured sources. Whilst some organisations are now addressing this by moving to the Cloud and leveraging native architecture capabilities, some in transport are still wary of these approaches – both from a complexity and perceived lack of security standpoint. These perceptions are leading many organisations to continue integrating their ever-expanding data sources and architectures manually, leading to lots of inefficiencies and higher levels of operational expenditure.

Yes, data integration can be a difficult task – but with the advancements in technological capabilities and technical expertise in intelligent automation, difficult does not mean impossible! The benefits of future-proofing your data landscape far outweigh the capital expenditure from long-term operational efficiencies. And the reality is, the avoidance of this is simply kicking an ever-growing can down the road. This is evident in today’s world where the travel customer expects a seamlessly integrated travel experience across transport modalities – it is not possible to serve this need with poor data integration capabilities.

We recently helped one organisation bite the bullet and automate the integration of data from 10 different sources – each with a different complexity in the structured to unstructured data spectrum. They were integrating these sources manually, and often this took so long that the data was already ‘out of date’ by the time front-end reports were ready. We designed and built a platform that not only introduced automated integration and ingestion of data, but also data analytics on key business KPIs to enable accelerated decision making. And what’s more, we were able to achieve this within 6 months. The real value add for the business was that we lay the foundations for scalability, which is enabling them to continue developing and expanding the platform to address additional use cases long after project conclusion.

Data Quality: Define what “good enough” means for you

The search for perfection is the reason why a lot of organisations get overwhelmed when they attempt to solve their data quality issues and do not know where to start. No organisation can boast to have perfect data quality, but some will be able to boast about having “good enough” data quality. So, what does “good enough” mean?

“Good enough” is having a good understanding of the quality of your data and ensuring that this quality is factored into the decision-making process. For example, the level of data quality required to enable financial reporting is different to that required to enable traveller demand modelling and forecasting. “Horses for courses” should be the mantra here – prioritise the data quality challenges required to support a specific business problem first, and then build out from there. This enables the seemingly mammoth task to be broken down into bite-size chunks without attempting to boil the ocean – start with “good enough” and then steadily progress towards “perfect” (a never-ending journey!).

We have recently helped an organisation focus their data quality remediation initiatives to support enhanced decision making. The business generated over 50 KPI reports – however a deep-rooted lack of trust in the data meant a vast majority were not being used at all. We helped the organisation strategize and focus on the L0 reports that were crucial to running the business. We established the critical data for these reports, established the relevant data quality rules and built various data quality index trackers. We segmented the data into 3 quality buckets – bronze, silver, and gold. This helped the organisation prioritise accordingly and focus on improving the data quality of business critical KPI reports first i.e. what “good enough” means for them. And importantly, the data quality segmentation gave business users qualitative evidence to understand how much credence they should put in certain reports based on the objective data quality assessment.


The first step should always be establishing an overarching data strategy for the company. This gives a clear direction for the organisation on its data journey as it clearly establishes the end goal and guardrails required to become more insight driven.

Armed with the right data strategy, addressing challenges in data integration and data quality will allow your organisation to innovate, become more efficient, and begin to treat data as an asset in its own right. Through tackling these challenges through a strategic, scalable, and pragmatic approach, transport organisations can unlock tremendous value from their data assets to enable enhanced decision making, operational efficiencies, and an improved customer experience.

For more information on how to tackle data challenges and become an Insight Driven Organisation, check out our IDO Playbook

** A crunchy question is a specific, quantifiable, and narrowly focused business question that is directly tied to strategic objectives and can deliver demonstrable value once answered