1. Own the win, thank the players – and let go
Success stories do not tell themselves. It is a common assumption among analysts that effective work will gain the recognition it deserves organically. Unfortunately, the noise within most organisations means this is rarely the case – or positive impacts get claimed by other functions without due credit being given to the role played by analytics.
Leaders can achieve a positive shift in the status of the function – and therefore the type of demands being placed on it – by constantly talking within the organisation (and even externally) about what it has delivered against specific objectives. This requires good data storytelling skills – and even the use of a specific data storyteller to capture and transform projects into compelling business narratives.
As part of this, individual practitioners and the team as a whole should be given due credit. Small acts of recognition – from just a thank-you through to prizes – have a big impact on morale and self-belief. Making these actions part of routine meetings ensures that recognition is embedded into the culture and expected to be given by all concerned.
An extension of this is to persuade stakeholders to attribute a proportion of any uplift (incremental revenue, cost saving, time saving, etc) to the analytics or data input. Even if this is a best guess – such as 5 or 10% – it establishes a clear belief that the function is a motor of change.
From there, leaders need to focus on up-skilling their teams with desirable communication and behavioural abilities. It is often commented on that analysts struggle to pushback on stakeholder demands or to say no, even when this is appropriate. Leaders can play a significant role by establishing the right questions to ask in these circumstances, such as whether the demand has the right backing, is framed correctly, or is genuinely the next best action to take.
As a final step, analytics leaders need to let go of the hands-on work and trust their teams to be capable. It is often tempting for the leader to step in – “I’ll just do it” – rather than spending the time to explain fully and delegate, then to provide appropriate support and oversight. But only by taking this step back will the team realise it has its leader’s confidence and start to operate positively.
2. Spend time with the business – and the team
In line with the need to communicate success stories outlined above, analytics leaders need to focus on developing a deep understanding of their business and their stakeholders’ needs and pressures. This is where the difference between routine queries and genuinely stretching demands emerges, since there is a more profound understanding on both sides of what is achievable.
Part of this involves explaining to the business the art of the possible – where analytics efforts could be focused for transformative effect. It is easy to assume there is a baseline of understanding, but this is often not the case. Exactly what analytics can (and can not) do for the business requires consistent explanation and exposure.
Alongside this, analysts need to be embedded alongside their stakeholders as much as possible. Whether this happens as part of a project team (perhaps in a squads and tribes approach), on a secondment basis, or as part of the permanent organisational structure, the dialogue that happens between embedded analysts and stakeholders builds a profound understanding on both sides.
It will be critical, however, to ensure that as a leader, you also have regular, structured contact with individual practitioners and the team, as well as more informal involvement. The sense of identity which comes from being part of a community of practice (see this briefing for more) helps to sustain the self-belief and skills base of analysts. They may operate as part of the business, but their role is not to do what their stakeholders do, it is to deliver great analytics. Leaders must continue to encourage, celebrate and support this.
3. Stick to the process
One of the most consistent complaints from analytics leaders and teams is that their ability to deliver the best possible work gets constantly disrupted by ad-hoc demands or is bogged down in business-as-usual tasks that should be offloaded (generally through automation). The fault for this can lie as much with the function itself as with its stakeholders.
A key dimension of good practice is to establish a proper process by which requests enter the analytics workstream. This involves a clear briefing process during which all requirements are reviewed and checked, with any that do not meet the right threshold required to be redeveloped and resubmitted. Sustaining this process will require firm gatekeeping by the leader and a willingness to turn down projects and stand up to pressure.
While this is desirable, there are clear challenges, especially arising from the seniority of the person making the request (or the use of seniority-by-proxy). Another disruptive factor are casual requests made by stakeholders to their favourite analyst without going through the formal briefing process. This is where enabling analysts to resist and say no – with the backing of their leader – will pay dividends.
As a final step, the prioritisation of projects needs to be addressed, especially where resources or time will not allow for all current demands to be met. The ideal is to reach a point where stakeholders themselves agree on priorities, especially if they agree among themselves which projects should go forward. Leaders can create regular work stream review meetings with the business during which this can be addressed.
4. Embrace the power of no
One of the challenges reported by a DataIQ Leaders analytics boss came from a stakeholder who asked, “why are you the only department that is allowed to say no?” Leaving aside the obvious untruth in this – finance almost certainly trumps analytics in this regard – it is a sign of the maturity of the analytics function when it feels fully empowered to decline tasks.
Indeed, to reach the desired level of maturity, it must actively pushback against certain demands and, in so doing, bring about a culture change. As noted in habit 3 above, one of the most disruptive factors for an analytics team can be stakeholders who by-pass the formal briefing process and try to get their task carried out as a favour. Pushing back on the basis of an agreed process will eliminate this behaviour.
More difficult to achieve is the status for the analytics leader as ultimate authority on what gets done by the team. In early stages of maturity, it is commonplace to be told “just do it” and for the seniority (or seniority-by-proxy) of the stakeholder to be used as a lever.
Only when the business understanding and technical expertise of the analytics leader has been established will their decision be accepted. To reach that point, strategic refusals – back up with strong evidence – will be necessary. Being able to demonstrate that the ROI for a task will be limited, that alternatives offer a better return, or that other questions need to be answered first will progressively build both credibility and authority.
5. Encourage experiments
Constant pressure to deliver against deadlines and repetitive, low-value tasks are the enemy of mature, innovative, value-driving analytics functions. While it is crucial to ensure that agreed, properly formed, prioritised tasks get delivered on-time and to specification, analytics leaders also need to create space in which experimentation can be carried out.
In some large organisations, there is a formal acceptance that analytics (or data science) teams can spend an agreed amount of their time on non-core tasks. Side projects are generally to be encouraged because of their ability to build skills, keep analysts curious and engaged, and as part of making their workplace feel exciting and dynamic.
What is most effective, however, is where these side projects are genuine business challenges or in some way aligned to the needs of the business, particularly if this experimentation time is carried out during regular office hours and using company equipment. During communication with the business, therefore, it is more than worthwhile soliciting moonshot projects or blue-skies objectives which can be set up as formal experiments. Alternatively, these can be scheduled into hackathon-style events where the whole team takes on a stretch project
6. Remember the vision
As the leader of an analytics function or team, one of your most important tasks is to keep the vision for the organisation and its use of analytics front of mind. As with BAU tasks, it is easy to let the day-to-day overwhelm any other considerations. But it is vital to be able to show how everything the function does is a step towards the ultimate goal and progression towards maturity.
Part of this requires the leader to live by example – “being the change you want to see”. While this can be challenging, especially if your natural personality type does not lend itself to being front-and-centre all the time, it is what your team and practitioners are looking for.
Conclusion
Being the ideal leader who brings about culture change and helps to deliver the vision for the organisation is not an easy task. There are certain to be mis-steps along the way. But one of the key metrics to track, which will help to keep you focused along the way, will be when stakeholders stops making the type of requests that have to be refused. Nobody wants to be in the department that always says no. Which is why changing the culture across the organisation is what will get you closer to being able always to say yes.
