But while digital transformation seems an appropriate response to these challenges, the same report discovered that 26% of organisations have experienced a failure in a project. Underlying causes were typically a lack of technology understanding (51%), lack of skills (31%) or lack of budget (23%).
Research carried out by DataIQ in early 2019 suggested another reason for the problems experienced – a failure to undertake a data transformation at the same time as pursuing a digital-first strategy. While 60% of organisations surveyed said a single view of the customer (SCV) was essential to the success of their digital transformation, only 24% actually have a SCV in place. Paradoxically, pursuing data integration was only identified by 33% as a critical success factor, even though 76% are undertaking this as part of their programme.
What seems evident is that digital transformation and datafication of an organisation are not considered to be interlinked or mutually beneficial, but rather separate and only partially overlapping projects. Yet discussion of this issue with DataIQ Leaders during the course of 2019 has made it very clear that undertaking digital change without addressing data is only tackling half of the problem which organisations are trying to solve.
The five gaps in digital transformations
Any organisational change will be far-reaching and cross-functional by nature. Digital transformation is especially complex because of the way it reaches across functional boundaries. These five gaps are repeatedly identified by senior data practitioners in the strategy and execution of digital-first programmes within their organisations and across industry.
1. Leadership gaps
Having a visionary leader to initiate and drive through a project is regularly identified as essential, whether that is the CEO, CTO, CIO or another CxO. Setting out a vision and using the political capital that comes with sitting in the C-suite makes significant change programmes harder to resist than when these arise from within a specific line of business or division.
But with this upside there also come several significant downside risks. First of these is the appeal of “shiny new toys”. Many DataIQ Leaders members have spoken about senior executives returning from technology conferences as newly-converted advocates for the latest solution. In recent years, everything from artificial intelligence to blockchain and cloud-based services has inspired this response. These can be highly disruptive of business as usual – 47% of finance decision-makers have stopped an existing service to pay for an unexpected large technology purchase, according to Econocom – even when the return on investment is not clear.
Second of the downsides is the abrupt departure of the visionary leader. Either as a result of technology overspend, revenue falls, service disruption or other non-related political issues, C-suite executives are in exposed positions in most organisations. Multiple DataIQ Leaders members have had the experience of such a departure mid-transformation, often leading to a loss of momentum and even the stalling of the project (leading to those failures identified in the research).
2. Revenue gaps
In certain sectors, such as publishing, there is an ongoing digital transformation as customers and revenues migrate out of print products. In retail, multi-channel brands are trying to balance rising revenues from digital (which are often not yet the majority of sales) with falling physical retail sales, and also the impact of print products such as catalogues. In insurance, the last two decades have seen price comparison sites transform customer loyalty in ways the industry has still not fully responded to, while open banking, energy switching and data portability are all bringing the same pressures to bear in multiple sectors.
Digital needs to have its own profit and loss account, as one major insurance provider has done, to ensure its revenues and costs are accounted for within its own terms, rather than in a cross-subsidy model. This will highlight or expose the digital transformation as a success or failure and make evidence-based decision-making on continued progress much clearer.
3. Process gaps
To be genuinely transformative, digital-first projects should introduce new ways of working, rather than digitising existing processes. This is where many such transformations fail because the established process is simply not suited to a digital environment.
Cycle times are a common example of where fundamental change is required. In the case of the multi-channel retailer which continues to produce a print catalogue twice a year, new product launches – from partner relationships through to marketing decisions – are profoundly rooted in a six-month cycle that is not optimal for the realm of e-commerce. This is even written into systems, such as how periodicity is defined.
Where some of these processes are transformed to be digital – for example, real-time stock indicators online – but others are not – such as stock replenishment supply chains – there are likely to be gaps into which the customer experience, revenues and efficiencies can fall.
4. Data flow gaps
Data silos are the perennial challenge to all types of transformation, often leading to problems with new digital processes because data is not available in real-time. A notable feature of conversations with DataIQ Leaders over the last 18 months is the scale of data re-platforming that is now taking place to address this.
As cloud-based services have become more trusted and even the norm, CTOs have come to understand that the core data platforms within organisations need to be modernised and converged. One of the powerful concepts helping to drive this is that of the data product – data as a key element within any new product or service which enables it to be delivered – rather than considering data as a consumable in the manufacturing process which is sourced once the new service has already been specified.
This is one arena in which the visionary leader – especially a CDO – can be a significant advantage and with less downside risk, since there is a good match between the next generation of business IT architecture and that of data platforms.
5. Measurement gaps
Within the data realm as part of digital transformations, business intelligence is often in the avant-garde, transforming how reporting takes place, who can access insights into business performance and how this happens. BI has been at the forefront of data transformation to introduce self-service solutions that support integrated data sets and data visualisation.
What also emerges are often problems with key performance indicators (KPIs) across the organisation. DataIQ Leaders members often discover hundreds, even thousands of KPIs across the business, many of them of limited value or even unused, yet still being generated by the BI function. Reducing this count or eliminating unnecessary KPIs is an important step in a data transformation.
At the same time, internal definitions and data descriptions used across the organisation are often discovered to be mis-aligned or even contradictory. Resolving these and imposing common standards is one of the most powerful steps in the data transformation which can have a major impact.
Conclusion
Digital transformations are set to continue as organisations modernise their infrastructure, seek to align the customer experience with changing consumer behaviour, or face market disruption from new digital-first entrants. At the same time, failures of some transformations seem inevitable because they do not address the underlying issues which need to be resolved or the critical success factors (CSFs) which have to be in place.
Prime among these CSFs are leadership, revenue measures and data. A digital transformation that has political cover, can identify and demonstrate a return on investment, and has a fit-for-purpose data platform as its underpinning will likely achieve its goals. Those which do not are likely to fall short.
Even more likely to fail are digital transformations that simply overlay new systems onto existing business processes, rather than building a digital-first process from the ground up. Ignoring the central role of data will also make it much more difficult to see progress for the business. Data and analytics leaders are well placed to ensure this does not happen.
4 – Impose a data tax
Where foundational data projects are required, funding for them can often be difficult to achieve. This is because of an experience which many DataIQ Leaders and members of the DataIQ 100 have had and which is best summarised as, “the first person to cross the river pays for the bridge.” What this means is that the full cost of a data project, such as a major data integration, is often imposed on the first new business project which needs it during a digital transformation. This can be a barrier to winning investment or certainly an obstacle to demonstrating positive return on investment.
Instead, an incremental levy or data tax on all business projects in the run-up to and during digital transformations can remove this problem. By gaining smaller contributions towards a larger project, no one business process or function leader is left facing the whole bill. This can also establish the data office as a standalone function with cross-functional support from within the business, giving it greater independence and resilience.
Alongside this, the data function should reach agreement with the stakeholders it supports to recognise its contribution. This can be an actual financial return in hard numbers or a percentage of any incremental gains being recognised. The key issue is to have defined ROI attributed to data in some way, since the function itself often struggles to identify the specific uplifts and efficiencies it has created.
5 – Retune the engine
Digital transformations often rush towards new technologies and solutions. In many cases, this is inevitable if there is a migration into the cloud or the adoption of AI, for example. But it is also often true that incumbent technology could be repurposed at lower cost and without the productivity gap which follows from new solutions as users get familiar and skilled with them.
Typically, legacy systems have not been updated or new features have yet to be adopted, even when the organisation is paying for perpetual licences. This is a quick win during any project and avoids the need for a major procurement process, only updating and a small degree of additional training.
Data can benefit significant from this approach since many of these updates and new features are specifically data-related, either generating or absorbing new data sources. This gives the data function an opportunity for low-cost retooling and future-proofing of its own platforms.