Background
Do you have any letters after your name? Forget about having a PhD, MSc, MBA or BSc – is there something that contains the letter I for institute or C for chartered? There are professions in which this question would not be asked because the possession of such a qualification is automatically assumed – accountancy, law, engineering, for example. Yet within data and analytics, there is no such presumption. Indeed, it can not even be assumed that practitioners all share a common educational pathway since a wide variety of STEM courses can be found in their background.
At a DataIQ Leaders breakfast briefing in September 2018, the qualifications of six chief data officers (or similar senior roles) offered a proof point. Academic training included computer science, business and law, and even international marketing. In most cases, the individuals had followed zig-zag careers to reach their present positions.
Professional certifications held by this group included chartered accountancy, banking and project management. The depth of study which some of these require is remarkable. Accountancy requires a four-year post-degree study to become chartered, for example, then three years of work experience which has to be written up and signed-off. Practitioners move into specialisms which have their own set of exams. Once qualified, that is the start of the career, even though individuals will already have an established practice. Continuous professional development is an ongoing requirement to maintain the qualification or practitioners lose their badge.
Yet despite their undoubted skills and abilities, even this senior group does not have anything which certifies them specifically to work on data. This is not to question their capabilities, skills or commitment.
Instead, it is a question of due diligence. Consider the paradoxical gap within the banking sector in how practitioners in different functions are required to prove their skills. A fund manager will have to spend three years studying to get a CFA certificate. But a data scientist may build the algorithms that run fund management without any such qualification.
The risk is of unqualified data practitioners developing something which they present to the business or externally and which has not been filtered through their understanding of core standards and requirements that apply to data. When a chartered accountant signs off on a set of books, their adherence to a set of fundamentals is not questioned. In data and analytics, this is not the case.
In the realm of recruitment and career development, the absence of a common qualification can also present difficulties. Comparing candidates on paper is challenging since even the new raft of data science degrees and above lack commonality and may not check that students are across the basics of data governance, data models and the like.
Even the broad workforce might see a benefit in being able to pursue a data qualification since, as one participant in the briefing noted, “the future of work = data + technology + vertical knowledge”.
So, should there be an Institute of Data or Chartered Practitioner in Data qualification? And, if so, what would it need to cover and how would it operate?
Chartering data practitioners and creating a data institute
In discussing the issue of a professional body to certify practitioners, one thing emerged clearly – the breadth of activities undertaken by data and analytics functions is so extensive that a single institute could not possibly embrace all of it. It was also strongly argued that data scientists are likely to consider that possessing a PhD is sufficient qualification in its own right.
In considering the boundaries for a data institute, therefore, it was felt to be important to identify a sweet spot where any certification was both meaningful and also relevant to a large number of people. This is most likely to be at the intersection in the T-shaped skills set practitioners need where practice-specific knowledge coincides with the domain-specific. Given the horizontal relevance of data across multiple business functions and also industry sectors, a certification sitting in this space could gain mass adoption.
Some key aspects of an institute were also identified:
1 – Setting industry standards
As anybody who has ever tried to achieve agreement on an industry standard will know, it is one of the hardest things to resolve. Vested interests and embedded processes often dictate the shape of such standards, while new entrants will attempt to disrupt the market by using their own.
Given the pace of change that data and analytics has undergone over the last decade, this may explain the dearth of existing standards. Relevant ones include BS10015 Personal information management system and ISO27001 Information security, both of which focus on technical systems and their implementation. There is also the IST36 Standards Committee for geospatial data which has yet to deliver a report.
BSI was reported in mid-2017 to be considering a standard for data ethics, which would move it closer to the behavioural and cultural dimensions which were identified as necessary. After all, you expect a lawyer to have a good technical understanding of applicable legislation and precendents – what you also need to be reassured about is that they will behave ethically, in your best interests and to the best of their ability.
But the gap is clear – there is no commonly-agreed standard for what constitutes good practice in data. A putative institute would need to address this as a matter of urgency.
2 – Accrediting practitioners
Professional institutes are primarily educational bodies which measure their members against a set of standards and award certificates. In the best examples, these become common currency and a benchmark for knowledge and performance.
For that to happen with a data institute, the consensus required to set industry standards would then need to be leveraged in order to build adoption. If employers start to expect candidates to hold a certificate which it has issued – and reward incumbents who decide to achieve certification – then this will embed both the institute and its accreditation into the industry.
3 – Continuous assessment and sanctions
While the baseline aspects of data, governance and models are relatively stable, things do change which will require refresher or update courses. It should be expected of certified members that they will maintain their skills and knowledge in this way.
To reinforce the professional standards being pursued, there should also be a mechanism for the institute to review any complaints or failures by members. Sanctions should be available which might include removing accreditation, either for a specific period and subject to remedial actions or even permanently. Achieving that level of enforcement usually requires recognition of an institute by government in order to avoid the challenge that it is restraining trade.
4 – Member and industry lobbying
Trade bodies play an important part in presenting the case of their sector to government, regulators or other parties who are considering actions that might have an impact. Professional bodies can also play this role where the effects are likely to be felt specifically by individuals.
Data practitioners have not enjoyed this type of representation except through the prism of specific sectors, such as techUK or the Direct Marketing Association. A professional institute with strong membership across a wide range of organisations would be likely to have a powerful voice.
Next steps
As with any action that is essentially political, the key questions to ask are, “if not now, when?” and “if not me, then who?”
On the first of these, DataIQ Leaders members were in agreement that there is a clear, current need for a professional institute. In some form or other, it seems likely that this will be addressed in the near term.
On the second, a specific challenge was issued to DataIQ Leaders to be the convening body, even if it is not the ultimate place where the institute gets founded.
To progress the concept, funding is necessary to support the research involved in understanding what standards are appropriate, the willingness of organisations to recognise them, and the effect on recruitment and salaries of accrediting practitioners. A potential source for this was identified as the Information Commissioner’s Office which makes grants for data protection and rights-related initiatives.
Initial scoping will now be undertaken into the opportunity, feasibility and support for such a move. Watch this space!