Maturity is therefore one of the key facets of momentum which needs to be constantly considered and demonstrated. Unless the business is making incremental forward steps up the maturity curve, it may start to believe that data and analytics itself is at fault as a practice, rather than other process or cultural issues.
At the same time, CDOs and CAOs need to be aware of another facet which is outside of their control – shareholder value. Although data and analytics are constantly seeking to draw a straight-line between their activities and the company P&L, one of the biggest pressures on the C-suite is share price and investor sentiment.
That ought to mean a high turnover at the very top. But one of the unexpected upsides of the UK’s dull economic performance in the last decade has been to suppress the turnover of chief executive officers (CEOs) in the UK. Average job tenure for CEOs in FTSE 100 companies has lengthened to 5.5 years over the past decade.
With relatively few of the “go-go” business drivers and City mavens who typified the 1990s and 2000s, the argument for keeping an incumbent in place has been stronger than for pushing them out, especially against an uncertain political backdrop. Shockingly, the average tenure for female executive directors is just 3.3 years, however, half that for men (6.6 years), according to statista.com, making the headline average misleading if your organisation has a female boss.
So the data and analytics function can count on a degree of stability and a long-term perspective. But when a CEO does get replaced, it usually heralds considerable disruption and re-organisation, not least because the incoming chief executive will want to put their stamp on the business (and often as not flush out anything they consider not to have worked or to be tainted by association with their predecessor). This has been witnessed directly by several DataIQ Leaders members in the retail, insurance and logistics sectors.
Job tenure for CDOs and CAOs themselves is a little harder to calculate robustly since there is a much smaller cohort with a shorter job history than the C-suite. But across the 28 individuals holding either of these job titles that were client-side members of the DataIQ 100 in 2019, 12 have changed companies in the 12 subsequent months – 43% of the group. It also appears that female CDOs are more loyal – while they represent 46% of the group, only 38% switched organisations, below the overall average.
If CDOs and CAOs are more mobile than CEOs, this presents an additional challenge – while a fresh face with a 100-day plan is often the specific reason for a data and analytics hire, an incoming leader arrives without a political base. Against an established chief executive who may or may not share the vision (depending on who led the hiring), this can make traction hard to achieve.
An example of momentum shift
Under a previous CEO, one of the largest insurance providers in the country invested extensively into data and analytics. It established a standalone digital operation with the culture of a start-up that was very distinct from the more traditional head office. This is a sector that is very risk-averse, yet the new function created in 2016 was given a mission to “put a rocket up the company”. To accelerate its impact, 70 data scientists were recruited and tasked to “move fast and break things”.
Significant changes did start to be realised, including a reduction in the friction during insurance pricing and quotation through predictive modelling based on customer data. Robust data underpinnings and platforms were put in place to support both customer science and data science with strong growth in internal demand as business units – which had been set against each other to create a competitive culture – looked for support for ongoing innovation and optimisation.
Despite all of this progress, a new CEO arrived in 2019 who has a more traditional view and has set about reviewing the business benefits being derived from data and analytics. Although there is clearly a potential culture clash, one of the senior leaders in the digital operation recognises that the data function needs to help the CEO to understand what it can do for the business, having already taken it from zero up a steep learning curve. That will involve continuing to do great work and also good internal PR in order to maintain momentum.
Another example of internal conflict
It is easy to assume that the creation of a data and analytics function involves a singular, usually centralised operation through which all such projects flow. Even if a decentralised model is used, wrapping individual teams into communities of practice will create a virtual, unified function.
But that is only likely to be the case if the way resources are adopted and managed is properly governed. One media organisation has experienced significant issues as a result of its approach to resourcing which involves an internal market – stakeholders are free to create or source data and analytics however they like, which has led to a “free-for-all”. This was made worse when one of the few central data teams responsible for data governance left the organisation and were not replaced. As a result, practitioners are routinely poached from one area of the business to another, or use this internal competition to gain promotion or higher pay without having to leave their employer.
The organisation also has a legacy data management system and related staff whose value is now being questioned. While it is delivery value to customers, it does not have any internally-focused outputs, such as business intelligence.
As part of a new data tech strategy, which is likely to see AI being embraced, moves are afoot to address these issues. In part, this is being done through resistance by practitioners towards working with stakeholders who only view this resource as reactive to their demands, rather than wanting to work collaboratively. By trying to change the culture from within, it is hoped that the momentum around what data and analytics can achieve – which will be critical to the adoption of AI – will be maintained on the back of existing solutions.
Using new projects to gain momentum
As in the example above, there are cycles within the adaption of and enthusiasm for data and analytics. Many of the existing functions were the result of the explosion of big data as a theme from 2012 onwards. The current decade will be led by AI and machine learning.
The challenge for CDOs and CAOs is to use these waves of hype to pull along foundational projects or to close out activities that have been lagging behind. This is particularly important given the new wave of chief data scientists and the emerging role of chief AI officer which risk overshadowing incumbent practitioners.
Internal communications and PR will form part of this. More critical still is to associate these core tasks with new strategies, such a business transformations. If data can get written into strategy, policy, values and more, then it is that much harder to separate out and downgrade. This does not mitigate the risk from competing divisional CEOs with non-aligned views on what purpose data and analytics should have. But those are perennial political problems that will exist across the organisation and are not unique to this sector. As long as the function leader knows how to tell a good story and is convincing, there is every chance of keeping things moving in the right direction.