Data Analysts Not Making Recommendations? – Here’s Why

After 25 years working in the technology space and 15 years of leading analytics teams at companies including Amazon, eBay, VMware, and more, I’ve heard one resounding theme from leaders. They want their data analysts to be more proactive and to provide more recommendations to the business. I agree with them. This is a significant gap in most organizations. But there’s reasons for these gaps, and most of the reasons are related to how you run your business.

What’s Desired vs What’s Asked

 Almost every company will say that they want their analysts to analyze data, to be proactive, and to make recommendations. But the reality is that many analytics teams don’t actually do these things, or at least do them to the desired degree. Also, when making these statements, many non-analytics leaders tend to overlook the other things that they want analysts to do. They don’t just want someone to analyze data and proactively provide recommendations. They also want analysts to build reports and pull data on demand.

These requests tend to occur so often, and with such urgency that analysts rarely have time to be proactive and make recommendations. How does an analyst proactively allocate their time to work on un-requested tasks if they are under a constant barrage of urgent requests? And what about the never ending backlog of work? Should we expect that analysts will allocate the necessary time required to think about the data, and think about the business context? Both of these are required to provide proactive information and recommendations.

And what about the details of the request and how it’s received? If context is necessary to provide recommendations, then an analyst would have to understand the real business problem at the onset.

I’ve been triaging stakeholder requests for close to 20 years and during my last year at Amazon, my team received approximately 1,400 ticket requests. I can confidently say that none of those requests contained all of the details necessary to properly understand the business, the problem, and what we really needed to solve. If you want your analysts to provide recommendations and be proactive, you have to start by understanding what analysts are actually being asked and what’s blocking them from doing the higher level work that you desire.

If you want your analysts to provide recommendations and be proactive, you have to start by understanding what analysts are actually being asked and what’s blocking them from doing the higher level work that you desire.

Measuring and Understanding Your Requests

The first step to understanding the problem is to measure it. To measure the problem, I recommend working with your analytics team to understand the following metrics. 

1. Volume of Work

First, look to understand the volume of work requests. While all requests are not equal, the sheer volume of work will tell you a lot about how your stakeholders operate and how your analysts may feel. 

Every request is a disruption of work. Not only is time required to address the request, but these disruptions also cause lost time due to context switching. It’s costly, it’s frustrating, and it’s stressful. I’ve never seen a better example of this cost than this comic from XKCD. 

2. Urgency of Requests

Second, measure the urgency of the requests. While disruptions are costly and can be stressful, there are ways to mitigate some of those issues. However, what is more detrimental is receiving disruptive work that has a high degree of urgency. Not only is there disruption and context switching, but the highly urgent work tends to create high amounts of stress for team members. They were likely under pressure to deliver on a previous task, and now they are put in a position of needing to complete an additional request within the same time frame. If you’re thinking, “why doesn’t the analyst just escalate and ask for re-prioritization?”, the solution isn’t that simple and this situation exposes many other challenges. I’ll save that for another article.

3. Type of Work Requested

Third, measure the type of work requested. While almost every stakeholder will state that they want analysts to be proactive and make recommendations, it’s important to understand if the request actually lends itself to recommendations. For example, if the request was to build a dashboard, then we shouldn’t expect an analyst to provide recommendations. By quantifying how many requests could actually fall into the realm of potential recommendations, you’ll get a better picture of what your analysts are being asked to do.

4. How You Make Requests

Fourth, measure how you asked for the work. Rarely have I seen stakeholders create a request that starts with, “I’m seeing problem x, could you help me understand why this might be happening and help me with some solutions?”. Instead, almost every request comes to analysts as, “I need you to pull this data” or “I need you to build me a report that has these metrics.” These requests are telling the analyst what to do. They aren’t asking the analyst for their help or their opinions. The analysts are simply following orders, and who could blame them.

5. Your Culture

Your culture has a significant impact on how your analysts work. Business stakeholders are frequently under pressure from their managers, and that pressure is frequently pushed downstream to a data analyst. A work request is generated and the analysts are requested to provide metrics and answers, usually in rapid fashion. 

There’s an important question that you must ask yourself as the stakeholder in these moments. When you’re under pressure from your manager, or you're just really busy,  do you really want an analyst asking you 20 questions about why you need this request and what the business problem is? Or, do you simply want your answer with the least amount of effort and resistance?

I’ve seen countless cases where stakeholders become frustrated when analysts ask clarifying questions to understand key details. This isn’t because the stakeholders want to help, but rather the stakeholder is trying to avoid appearing slow or unresponsive to their manager. A data analyst only needs to encounter a few of these situations to condition the analyst to do as they are told and to curb their independent thinking. This is usually the result of dealing with a frustrated stakeholder or hearing negatively about how the analyst “is being difficult” or “asking too many questions”. In some organizational cultures, I’ve seen well-intended leadership principles weaponized as, “the analyst doesn’t have Bias for Action.”

Additional negative cultural conditioning that occurs in some organizations where I’ve seen stakeholders draw a hard line on what they wanted the analyst to do. For example, most organizations, rightfully so, expect their business leaders and program managers to be experts. Where this becomes delicate and challenging is that analysts can also be experts with some of the same aspects of the business. If the business leaders are supposed to be the experts, then why do they need recommendations from a non-expert? This tends to cause analysts to be pushed into “service provider mode” instead of “partner mode”.

Adding It Up

After measuring how your stakeholders operate with the analytics team, and measuring analytics operations, you should be able to create a clear picture of what is getting in the way of the proactive insights and recommendations that you desire. With this understanding, you’ll need to develop the right mechanisms to ensure that analysts have the bandwidth to provide proactive insights and recommendations. 

You’ll also need to develop a culture of partnerships where analysts are empowered to provide recommendations. Additionally, you’ll need to foster a psychologically safe environment for your analysts to provide such recommendations and allocate time to being proactive.

These aren’t easy changes and they require your entire organization to be aligned. They also require your organization to have a strong backbone to avoid regressing when the next urgent issue arises. 

When you elevate your understanding of how your organization is operating with your analytics team, you’ll be taking your first step towards fully utilizing your analytics talent and delivering impactful recommendations and results. 

Brandon Southern, MBA, is a Sr. Director of Analytics and the founder of Analytics Mentor. With a 25-year career in the tech industry, Brandon has excelled in diverse roles encompassing analytics, software development, project management, and more. He has led analytics teams at Amazon, eBay, VMware, GameStop, and more.

You can learn more about Brandon and Analytics Mentor at http://www.analyticsmentor.io/