Business and Data Science - Managing Expectations

Posted 2017-06-01T Posted by Tom

The role of a data scientist is an incredibly diverse one – demanding varied responsibilities and a natural inclination for both analytical skills and business acumen. Little wonder then that skilled data scientists are compared to being as rare as unicorns.

On par with mythical beings they may be, yet for businesses and data scientists to be aligned, expectations must be realistic, and structures must be put in place if even the most talented of data scientists is to flourish.


The role of a data science leader demands skills unlike those gained in management or analytics roles in and of themselves. Such managers must hold both a depth and breadth of understanding of data science – with depth coming in the form of first-hand experience of complete data science product life-cycles - hands-on experience from working in a project from the ground up. Breadth then comes from a comprehensive understanding of the data science landscape – something that can prove the critical difference between an acceptable manager and an exceptional analyst.


The starting blocks – Be selective about the data that is to be analysed

First and foremost, businesses should avoid becoming data greedy. Overwhelming amounts of data may well be generated by the average corporation, yet forming actionable insights from vast swathes of data is often needlessly time consuming. For this reason, data strategy demands a flexible approach – one where data points are continually assessed, with data needs prioritised by every relevant stake holder in a firm, not just by data scientists.

Embrace realism – and prepare to commit the necessary budget

Data science can represent the future of a company, and so businesses must be realistic about the talents that they require and the budgets that are needed. Taking an example – if a company is data-heavy, it may be that separate hires are required for the role of managing the data, and the role of analysing it.

Before you hire: Step back and appreciate the importance of the data scientist

Realism must also extend to evaluating the role for which you are hiring – data scientists are there to not only uncover insights, but to also identify and address problem areas. Their role is complex, pressing and challenging – and you must provide the business support structures they need to thrive in their role.


Communicate: Facilitate teamwork

The success of a data science project relies on solid communication – a team must be cohesive and co-ordinated, central to which is the rare breed that is the natural data science leader of which we spoke of earlier.

Harness business intuition

Businesses are home to business intuition and extensive commercial talent – and data scientists mustn’t be afraid to seek out these people and draw on their knowledge, which can shed light on the customer, the product and the market. Companies can assist this by opening up lines of communication and providing time for these departments to collaborate.

Measure results

Why are you doing what you’re doing and what defines success? These are key questions that must be answered before so much as a single data set is looked at. And based upon your measure of success, must be metrics that will then define whether or not your data science efforts have achieved all that you set out to, fell miserably flat, or reached somewhere in-between.

Work in an agile manner

Business intuition and input should continually feed a data science project – and as insight is gained this must influence the project’s next steps – making agile working a must if fluid project deliverables are to be met. Minimum viable products and short delivery cycles are key.

Create a community and a culture of sharing knowledge

A successful data science programme begins and ends with open, healthy and productive relationships between the right people and the right departments, yet this doesn’t happen without a concerted effort. In order for a culture of collaboration to be created, it must be nurtured – meetings must benefit all, and involve everyone – providing chance for key stakeholders to share knowledge – rather than for a leader to simply dictate.


Data scientists are rightly placed on a pedestal – held in high esteem and capable of being at the centre of smarter decisions, improving productivity and commercial direction informed through data insight. Yet the commercial advantages of utilising data scientists rely not only on realistic expectations being placed on them, but likewise the realistic expectations of others – specifically the senior management and leadership teams. Beyond this symbiotic relationship, there lies the challenge of fitting data science with business strategy – something that demands a suitably prudent appetite for data, as data greed can quickly devour the prospects of project success.

The strategic preparatory groundwork for any data science endeavour is as important as the nuts and bolts of a data science project - the methodologies, the data sets, the models. Ensuring data scientists work collaboratively with the business <u>and</u> with those in other core departments is key to optimal ROI being derived from what is, after all, a significant investment in the future of an organisation.

Get priority access to Pivigo news, features, events and networking opportunities