Bridging the Gap - A Data Leader’s Guide To Helping Your Data Team Create Next Level Analysis
One of the biggest challenges data leaders have mentioned to me over the past couple of weeks?
Getting quality work from their analysts.
It’s not that the analysts are bad at their jobs—far from it. But too often, their work gets handed off to stakeholders before it’s truly ready.
Maybe it’s missing a clear story. Maybe it doesn’t drive action. Maybe it’s overly detailed without getting to the point. Or worse, maybe the numbers aren’t even accurate.
And honestly? I get it.
After hours spent analyzing data and writing queries, the last thing you want to do is take another hour to refine the narrative and make it compelling enough that stakeholders do more than just stare blankly. But that final step is what separates work that gets ignored from work that drives real change.
So, how can you get more from your team? Let’s break it down.
1) Make It Clear What “Good” Looks Like
Most people coming out of college aren’t trained to deliver an analysis for executives, create a compelling story, and provide actionable results.
At least, none of my courses covered that. Maybe you did a few case studies, but those didn’t require you to pull together multiple disparate data sources, query everything, and then make sense of it.
And in school, there were usually right answers.
In the real world? Not so much. And to make things even trickier, no one hands you a guidebook on what a “good” analysis looks like. You’re just supposed to figure it out.
Some companies invest heavily in training new hires—and honestly, more should. You do need to set the standard. Yes, analysts need to handle ambiguity, but even if a junior or mid-level analyst had 10 extra hours to polish an analysis, it wouldn’t necessarily make it better.
Why? Because they don’t know what they’re aiming for.
So, show them.
Define what good looks like. Give clear examples. Teach basic techniques. It doesn’t have to be complicated—sometimes, simple visual comparisons work best:
If you’re looking for a good book to buy your analysts(unless they already have it), Story Telling With Data is a classic. The picture above is from their blog and it helps highlight a lot of what I’ve talked about over the past few articles.
Don’t bury the lede - If there is a take away from a chart, make it clear which they do
Provide support after the main point, notice the “because of reasons x,y,z”. You want to have support for your conclusion but you don’t want to bury it behind a bunch of other facts.
To be clear, you’re not trying to hamper their own creativity. You want to hire smart people that can solve problems, but that doesn’t mean you can’t provide frameworks and guidelines to help them better present their ideas.
A few common challenges analysts run into:
They want to show all their work, so they include four charts when one would do.
They focus too much on how they got to the answer instead of why the answer matters.
They overlook small but critical formatting details—like bolding key insights or using color effectively.
These might seem minor, but they make a huge difference in how an analysis is received. And most analysts, myself included, take forever to learn some of these things on their own.
So, don’t make them figure it out the hard way. Show them what works.
2) Peer Reviews
It’s interesting—software engineers regularly review each other’s code, but I’ve rarely seen analysts review each other’s analysis or queries. Instead, they often write up their findings, skip the review process, and hand them off as fact. No second set of eyes. No formal data quality checks.
Now, I get it. You can’t always implement a lengthy peer review process for every ad-hoc request. But you can put a few simple safeguards in place:
Have analysts write basic data quality check scripts on their own work (though ideally, someone else should verify them).
Implement peer reviews by senior analysts to ensure:
A clear and structured story
Well-formed, actionable insights
A quick sniff test on the numbers
Does this take extra time? Yes.
But let’s be honest—we all put in a little more effort when we know someone’s watching and likely to critique our work. If I’m at the gym and someone asks to work-in on my set, you better believe I’m adding extra weight and pushing myself more.
The same goes for analysis. A second review makes everyone more accountable—and leads to stronger, more reliable insights.
3) Teach People the Power of Story
Analytics deliverables come in many forms, but whether you’re building a dashboard, writing a report, or giving a presentation, the goal is the same—you need to tell a story with your data.
You could take the easy route and simply present some numbers. Or, you could guide your audience through a narrative that helps them understand why the data matters and what they should do with it.
To craft a strong story, you should be able to answer the following:
Who is your audience?
What does your audience care about(what pain points do they feel)?
How do you want your audience to feel after seeing your presentation?
What actions do you want them to take?
In fact, I just heard
say something similar on ’s podcast when they were talking about creating content.At the end of the day, you can be just another analyst handing over numbers, or you can be the one who drives real decisions and change.
Tactical Tips
Before wrapping up, let’s go over some tactical tips you can apply—whether you’re a data leader or an individual contributor.
Create a Framework for Analysis: When I did my first analysis, I was all over the place—too many thoughts, a lot of questions, and no clear direction. That’s the nature of analysis: you can always dig deeper. Without structure, it’s easy to get lost. Help your team by creating a framework that acts as a guide. It doesn’t need to be rigid, but it should provide clear steps analysts can follow to organize their work, focus on key insights, and deliver results efficiently.
Onboarding Matters: When I spoke with Joe Reis and Richard Millington, both mentioned the importance of onboarding. They created materials to help consultants understand expectations from day one. Why? Because their reputation—and their business—depends on quality work. The same applies to data leaders. If you want consistent, high-quality analysis, take the time to set expectations upfront. A strong onboarding program ensures new hires understand what’s expected and reduces the trial-and-error phase.
Provide an Example of a Good Analysis: Once someone broke the four-minute mile, others quickly followed (ok, not everyone—I’m definitely not close). The point is that when people know what’s possible, they know what to aim for. If you have an example of a well-done analysis—whether from a past employee or your own work—keep it. Use it as a benchmark during onboarding. Break down why it’s effective so others can learn from it.
Some risks come with introducing a framework—people might stick to it too rigidly, which isn’t ideal. While structure is helpful, you don’t want it to stifle critical thinking. Encourage your team to go beyond the basics but still hold them to a standard.
Perhaps it’s the side of me that started my career in fine-dining, but someone has to call you out when you’re putting out bad or mediocre work.
In the famous words of a sous-chef I once worked for:
“If you’re ok with mediocrity, then I guess so am I.”
Final Thoughts
Many analysts never had anyone show them what a good analysis looks like. No one taught them how to write effective recommendations or highlight key points.
Maybe they were lucky, and they had a professor with business experience who at least emphasized getting to the point quickly. Maybe they didn’t. But if you want to get more from your data team, you need to put in the effort to set clear standards.
If you have an analysis that stood out or a strong example to reference, use it. Consider creating a guidebook on how to write and present metrics and outcomes effectively. Books and resources can help, but having a solid internal example is always a great place to start.
As always, thanks for reading.
Upcoming Member Events
I have been working to add more and more value to the Seattle Data Guy Newsletter. As part of that, I’ve been running member webinars!
Here are the next three I have planned!
How To Pick The Right Data Tools For Your Team - March 27th, 9 AM MT
Breaking Into Data Consulting - From Marketing To Positioning - April 8th, 9 AM MT
Dimensional Data Modeling Fundamentals - April 17th, 9 AM MT
If you’d like to be invited, then consider supporting the newsletter below!
Join My Data Engineering And Data Science Discord
If you’re looking to talk more about data engineering, data science, breaking into your first job, and finding other like minded data specialists. Then you should join the Seattle Data Guy discord!
We are now over 8000 members!
Join My Technical Consultants Community
If you’re a data consultant or considering becoming one then you should join the Technical Freelancer Community! We have over 1500 members!
You’ll find plenty of free resources you can access to expedite your journey as a technical consultant as well as be able to talk to other consultants about questions you may have!
Bring Instant Clarity to Business Objectives
Before I got into data engineering I tried my hand at data analysis. I hated it.
I just couldn’t figure out what people wanted out of data. Whatever questions I could answer led to more questions and whatever features I could add to a dashboard led to even more feature requests.
It was incredibly frustrating.
Even when I asked people what they wanted, gathered data requirements, built and released multiple reports and dashboards it never seemed to satisfy. On top of that, the dashboards wouldn’t even be used.
How To Data Model – Real Life Examples Of How Companies Model Their Data
How companies data model varies widely.
They might say they use Kimball dimensional modeling. However, when you look in their data warehouse the only part you recognize is the word fact and dim.
Over the past near decade, I have worked for and with different companies that have used various methods to capture this data. I wanted to review some of the techniques that are commonly used to model data for analytics.
This is part of my unofficial series on data modeling, so if you’d like to learn more, then you can check out some of the prior articles such as this one.
But for now, let’s dive in.
End Of Day 171
Thanks for checking out our community. We put out 4-5 Newsletters a month discussing data, tech, and start-ups.
One of the reasons I wanted to become an adjunct lecturer was so I could cover things like data storytelling in my courses because, to your point, it's often not adequately covered (or covered at all) in traditional higher education. On top of sharing what constitutes as "good," I've found making a safe space to surface WIP analysis and do peer reviews in addition to manager review to help everyone keep simplicity top of mind!