The world’s leading publication for data science, AI, and ML professionals.

5 Tips for Analysts (and their Managers)

Add more value as an analyst and ways managers can support and retain talent

Office Hours

I’ve worked professionally as an Analyst for just over two years in large retail organisations. Depending on the company you work for, analysts are either seen as an asset or just a numbers grunt, and this pigeon holing in my experience is driven by the insecurity of some managers when it comes to being more data driven and not decisions based on their gut feeling, elsewise known as the HiPPO (Highest Paid Person’s Opinion).

Photo by Isaac Smith on Unsplash
Photo by Isaac Smith on Unsplash

The inspiration for this article stems from that one weekly meeting we all sit through where various teams present their functions KPI metrics for the previous week. You know the meeting, where manager’s explain their figures with a spreadsheet prepared by an analyst in their team, with KPI boxes, or worse slabs of unformatted tables. Is anyone paying attention, and how can we do better? Here I provide 5 tips for analysts to increase the value they bring to their team, and advice for managers to better engage with your analysts.

Analyst Tip #1— Know your end user

The life of an analyst is tough. We need to be strong communicators, extremely logical thinkers and creative. Before you go diving into your data warehouse (or data lake), spend some time considering your end user needs. This may even mean setting aside time to take a brief from them, and ask meaningful questions, because in the end this will make writing that query much easier. Whatever they’ve asked for add another 50% in scope. Explore adjacent datasets that may add more value to your analysis. I always have ‘what if?’ circling in my mind when writing queries for new analysis.

Manager Tip: Be very specific about what your asking your analyst to do, and be patient as they seek to flesh out your idea. Set clear expectations and give reasonable timelines for delivery. Treat this as a collaboration and not an exercise in top-down management.

Analyst Tip #2 – Be impactful with visual analytics

Data visualisation is an art form. Most people are visual, and the endeavour of an analyst is to make insight easily perceptible to their target audience by using well designed visualisations. So many times have I seen people have four weeks worth of figures in a table, when your audience would perceive changes easier and better with a time series chart. Visualisations should be designed with a specific question in mind that they are seeking to answer. Understanding trends over time is incredibly important for managers, and understanding how and why trends are shifting or staying the same is even more important. Reporting aggregated figures tends to hide things that are shifting in the underlying sands that simple KPI reporting doesn’t reflect, tell the story with a series of well designed visualisations.

Manager Tip: Be brave and turn up to your next stand up with a handful of visualisations that report the same figures with better detail and improve engagement with your audience. It may prompt a discussion that leads to better business outcomes.

Analyst Tip #3 – Develop flexible and dynamic dashboards

An extension of the above, is the dashboard design process. Who doesn’t love a good dashboard? Dashboards are collections of metrics and visualisations, with the intention of being able to provide deeper level analysis and enable underlying trends not immediately visible with KPIs to be perceived. I’ve worked with Tableau, Microstrategy, PowerBI and Quicksight, and by in large they are all variations on the same theme. They become even more powerful when you start developing means of interactivity. This is usually in the form of controllable filters and parameters.

Parameters are my absolute favourite, particularly when working with time series data, being able to change how the dimension that the group by clause uses, or changing how the metric is calculated is so powerful. A recent dashboard I’ve developed enables you to change the time series from a week to week figures, cumulative year to date, percent change week on week and a 4 week moving average. Dashboards should be easy and intuitive to use, and enable your customer to answer an array of questions using the data you’ve curated. Design is an incredibly thoughtful and creative process and shouldn’t be rushed. Look back at your work from time to time and consider ways it can be improved

Manager Tip: Spend quality time with your analysts so that they can explain the features of a new dashboard and how you can use it most effectively. Make sure to understand the scope of the questions it can answer, and provide meaningful feedback. It’s a collaboration.

Analyst Tip #4 – Don’t enable access to raw data

This is a pet peeve of mine, when a manager doesn’t trust the figures you’ve presented so they insist on seeing the data or having raw data visible in a dashboard. What’s the point really? The point actually is, you haven’t done your job properly. If ever your approached to provide large datasets, always ask "What are you seeking to understand?". The last thing you want is a manager or colleague wasting their time playing with a spreadsheet trying to find something that isn’t there, or worse, is there and you haven’t surfaced it. For the sake of transparency if someone would like to understand how the dataset was generated, explain (where reasonable) the table logic. This goes back to our first tip, properly understand the data you’ve collated after you’ve understood your end user’s needs.

Manager Tip: Don’t ask to see raw data, use this as an opportunity for feedback and explore with your analyst the question you’d like answered. 9 times out of 10 it may require a further query to be generated.

Analyst Tip #5 – Value yourself

Sometimes analysts don’t get the praise or acknowledgement we deserve. We slave over complex queries, work with data engineers and architects so that we can access additional data sources, develop visualisations and dashboards. The last thing you want after all that effort is to not be valued by your team and management. What we do takes time, but when analytics is done reproducibly adds tremendous value to our respective business’. It has never been a better time to be an analyst, the next wave of technological uplift is being accelerated with the impacts of COVID-19 being felt worldwide, and business’ are shifting to better utilise their most precious asset, data. So value yourself, your skills and if you feel as if you aren’t being valued, (discretely) find another job. You’ve got nothing to lose.

Manager Tip: Keep your analysts engaged in meaningful work, and facilitate their ongoing development. Find them an internal mentor and/or if your organisation can afford to do so, enable further training. The job market is going through a generational shift and is increasingly competitive as organisations seek to grow their internal analytics capabilities. Employee churn is generally the result of poor management and organisational culture.

Concluding Remarks

Having worked in environments that take different views on the role of the analyst here’s mine. Analysts are inherently business partners; the interface between our line of business and data. You are playing a growing and important role in your organisations future direction. You enable data driven decision making in how you communicate your findings and enable self service analytics. You have a unique understanding of business systems and see links that others may not. You will collaborate across functional areas and create professional networks internal and external to your organisation. Think about ways you can add value, and how you can work smarter. Lastly be visible and take credit for your work.

To the managers who read along, I recommend establishing a regular 1:1 meeting to maintain open lines of communication with your analyst. Collaborate constructively with them and they’ll provide you with the data firepower you require to drive business outcomes.

Today’s analysts are tomorrow’s business leaders, informed by data and unique business understanding, we represent a new generation of managers that will make better and faster decisions.


Related Articles