.png)
AMA, or “Ask Me Anything” events are a great way to surface burning questions within a community and let an expert voice to weigh in with their perspective. This promotes continuous learning and creates psychological safety within the community, enabling people to ask questions outside of the AMA events, too.
Aside from building connections between members, it allows a culture of curiosity to flourish. That’s why the GOATs analytics community held an AMA with June Dershewitz, where we explored some wonderful questions from the members about career paths, data visualization, decision-making, and how the analytics field is evolving with new advancements.
Here are the results of the GOATs AMA. Enjoy!
Thanks for reading! Subscribe for free to receive new posts and support my work.
From Olga Berezovsky:
“When stakes are high and stakeholders are pressuring for quick decisions, how do you stay rational and avoid emotional or political bias?”
Your question reminds me of a quote from the 19th-century Scottish scholar Andrew Lang: “Most people use statistics like a drunk man uses a lamppost; more for support than illumination.” Data folks have latched onto this quote because it points to a real problem. In high-pressure situations, business stakeholders sometimes want data to confirm a decision they’ve already made. That can make it hard for analysts to remain neutral.
So how do you stay rational under pressure? Here are a couple of ideas:
First, even if time is tight, take a step back to understand the problem the stakeholder is trying to solve. I call this “the question behind the question” (which I’ve written about here). A rushed request might get conveyed to you as a narrowly scoped data pull, but once you know what decision the stakeholder is actually trying to make, you can think more creatively about the best way to respond. Sometimes the most valuable thing you can do is reframe the problem.
Second, resist the urge to provide more certainty than the data allows. It’s okay to say, “Here’s what I know right now,” and stop there. If it’s helpful, propose what you could do with more time or more data. Maybe that’s a follow-up analysis or a more rigorous method like A/B testing. Offering a phased approach gives stakeholders something they can act on now, without compromising the integrity of the bigger picture.
Finally, empathy can help, too. If someone is pressing you hard, it’s probably because they’re under pressure themselves. Recognizing that dynamic (without getting caught up in it) can help you stay calm.
From Maya B.:
“Is self-serve analytics the optimal direction for business users? What are the critical enablers for its widespread success in addressing most analytical questions? How will the role of business analysts evolve amidst these rapid advancements?”
Yes, self-service is the optimal direction - with caveats.
In companies that place a high value on data, there will come a point when the demand for analysts’ time eventually outpaces supply. Rather than growing the analyst team larger and larger, a natural way to remove the bottleneck is to make it possible for business users to access data directly through self-service tools. As a result, I’ve found that data-savvy companies tend to invest heavily in self-service.
Here’s where it can break down, though: business users may not know what pitfalls to look out for. Maybe the data quality is poor and they don’t notice. Maybe they mistake correlation for causation. Or they interpret the data in a way that confirms their existing beliefs, without realizing they’ve skipped a few logical steps. There are lots of ways that self-service can cause more harm than good.
Here are a couple of ways to avoid these problems:
If, as an analyst, you begin to see self-service starting to absorb the simpler ad hoc requests you used to handle, I encourage you to think of it as a good thing. It frees you up to take on higher-impact work that draws on your deeper knowledge of the data and more advanced analytical skills. It also paves the way to longer-term, more strategic projects.
I recommend that you keep a running list of valuable work that you COULD do if you had more time. That way, when self-service starts to reduce the data-pulling load, you’ll know exactly where to reinvest your energy.
From SaiTiger:
“One key factor in dashboard design is understanding the audience. When you’re working with a diverse group of stakeholders, how do you simplify without overwhelming them with a large set of metrics?”
First of all, you’re absolutely right: understanding your audience is the foundation of good dashboard design. It can definitely be challenging, especially when your stakeholders have different goals or don’t agree on what success looks like.
My advice would be to treat this as a product management exercise, because that’s exactly what it is.
Although there might be a lot of people involved, make sure that you identify a primary business partner for the dashboard. Work with that person (or team) to align on purpose and scope. As you gather input from the larger group, take the time to document all of the requirements in one place. Then work with your primary partner to review those requirements and prioritize them. Some metrics will clearly tie back to important business goals, while others may be simply “good to know” but unlikely to drive any meaningful action. Through this exercise, you want to make sure that everyone feels heard, while also reinforcing that not everything and the kitchen sink will make it into the final dashboard.
I have sometimes found that when users ask for a large number of metrics, it can mean a few things. Perhaps the business is still unclear about how to measure success, and they’re casting a wide net in the hopes that something will stick. That can be a signal for the data team to step in and offer more support around metric strategy. Also, stakeholders may not fully understand the cost associated with including more and more data. Every metric displayed in a dashboard has to be collected, cleaned, validated, and maintained over time. That ongoing investment isn’t always visible to the end user, but it does matter.
If you get pushback, it helps to explain that placing boundaries on what’s collected and shown can actually improve the quality and trustworthiness of results. A lighter-weight dashboard is easier to interpret, more likely to stay accurate over time, and ultimately more useful for decision-making. When we limit what we include to what’s truly necessary, we’re doing everyone a favor.
From Prashant Raturi:
“If I’ve always been a people manager and want to pivot into people analytics, what skill set would it take? How does the path differ for someone who is working vs. someone who is laid off?”
From a general-purpose analytics perspective, I would recommend learning SQL, Python, and a BI tool like Tableau. On top of that, add on skills that are more specific to people analytics: definitely HR systems like Workday, and perhaps also survey design and analysis. Finally, I encourage you to do some forward-looking research to understand how this space is evolving. As businesses begin to integrate AI agents into workforce operations, the definition of “people analytics” may shift. In a few years, we might not even call it that anymore!
To your question about different paths based on your employment status: yes, you can build these skills whether you are working or not, although your tactics will likely differ.
If you’re currently employed, ideally you’re already having conversations with your manager about your long-term career goals. If not, it’s never too late to start. Together with your manager, you can look for opportunities to develop relevant skills within your current role. For example, you might set aside time to partner with the HR team on a project that involves people analytics. If there is already a dedicated people analytics team within your company, reach out for an informational interview to learn how they work and where your experience might fit. One advantage that you already have is an understanding of how the business operates, and likely a strong foundation of trust with your colleagues. That gives you a head start.
If you’re in between jobs, focus on upskilling through public resources. Many are free, like Youtube tutorials or documentation from tool vendors. Others are relatively affordable, like Maven or Coursera content. Keep an eye on job listings that sound interesting and assess your current skills vs what’s being asked. Use that gap to shape your learning plan. You might need to take a step down in seniority temporarily in order to break into a new area. I know it’s a tough time to be a job-seeker, but your advantage is that you have more time to invest in growth than someone who’s trying to make the switch while holding down a full-time job.
From Prashant Raturi:
“How difficult is it to break into analytics? Is there demand for it? Apart from theory, are there hard skills to focus on—coding, data visualization, etc.?”
To answer your question directly: yes, there’s demand for strong analytics professionals, but it’s also a tough time to break in, simply because of the state of the job market.
Compared to, say, five years ago, there’s a lot more competition. It’s not necessarily that there are fewer analytics jobs, but rather that more people are trying to find employment at once.That means job seekers need to be especially thoughtful about how they stand out.
The long-term outlook is still strong, though. According to the U.S. Bureau of Labor Statistics, employment of data scientists (used here as an umbrella term that includes many analytics roles) is projected to grow 36% between 2023 and 2033, which is much faster than the average for all occupations. (Source)
In terms of hard skills, as I mentioned above, I recommend starting with SQL, Python, and a BI tool like Tableau. Those are foundational. On top of that, I think there will be a greater demand for analytics professionals who demonstrate their ability to leverage AI in their work. Not only will it help you improve the efficiency of what you do, it will also help you understand how your own role might evolve. Tasks like rote data pulling or basic reporting might diminish, while skills like stakeholder management and strategic thinking will become even more important.
From Prashant Raturi:
“Are there any courses, textbooks, or communities to follow that could provide a foot in the door or show that someone is serious about people analytics?”
Here are a few recommendations to get you started:
I’m not aware of any people-analytics-specific communities, but broader analytics groups like GOATs are great for finding like-minded folks.
From Zack Martin:
“Do you have a favorite goat?”
Why yes I do! Her name was Maxine. She was a purebred French Alpine dairy goat, and she was my showmanship partner during the formative years I spent in 4-H club (which is basically a youth group for farmers). Maxine and I spent a lot of time together on the farm and in the competition ring. My dad was a software engineer and my mom raised show goats, which explains how I came to be a data person who loves goats!
Here’s Maxine (and her kid, Melody) out in the pasture, back in the day.
.png)
