The Analytics Power Hour #291: The Data Work that Lives in the Shadows
Release Date: February 17, 2026
Hosts: Michael Helbling, Moe (Mo) Kiss, Tim Wilson, Val Kroll
Main Theme:
Exploring the "shadow work" in analytics—the vital but under-acknowledged tasks data professionals do that fall outside formal job descriptions. The hosts dissect, debate, and affirm the value, frustrations, and nuances of this unseen labor, highlighting its criticality to successful data-driven organizations.
Episode Overview
- The conversation delves into the hidden, often-overlooked work done by data professionals—the "shadow work" that is essential for organizational effectiveness but rarely recognized or celebrated.
- Topics include the types of shadow work, who does it, its impact, the interplay with personality and team structure, pitfalls around recognition, and strategies to bring invisible labor into the light.
Key Discussion Points
What Constitutes Shadow Work in Analytics?
00:14–08:45
- Admin/Project Management:
- Val Kroll opens with "admin" as a major shadow work category—project management, chasing stakeholders for contributions, filling out spreadsheets, and maintaining momentum.
- "Once you assume responsibility for it, it’s almost impossible to ever roll it back." – Val (02:16)
- Tim differentiates between necessary accountability and pure admin, advocating that follow-up after recommendations is analyst territory.
- "There’s an accountability mechanism for the analyst to say, 'hey, did you do it?'" – Tim (03:22)
- Val Kroll opens with "admin" as a major shadow work category—project management, chasing stakeholders for contributions, filling out spreadsheets, and maintaining momentum.
- Integration with Implementation:
- Michael shares stories of data teams unintentionally morphing into project managers when following analysis through to implementation.
- "You start out being the analyst, next thing you know, you’re running the integration task force." – Michael (05:37)
- Both Mo and Val agree taking ownership can be empowering, but warn about being drawn too far from specialist roles—there’s a balance.
- Michael shares stories of data teams unintentionally morphing into project managers when following analysis through to implementation.
Explaining and Interpreting Others’ Data
08:45–12:51
- Data teams are often tasked with deciphering, explaining, and documenting data (and reports) originating from other agencies, departments, or predecessors—with little to no handover or documentation.
- "Now you need me to reverse engineer how they pulled these numbers together. Oh boy." – Michael (08:47)
- This is exacerbated by poor documentation and alignment between business units.
- "Getting everyone using the same data, everyone trusts the data, whether it’s internal or external… sometimes it falls in your lap in a weird way." – Michael (10:03)
- "Not all shadow work is shit. Some shadow work is actually very valuable—alignment is one." – Val (10:29)
Alignment, Consensus Building & Data Culture
11:14–14:57
- Alignment:
- Data professionals often perform crucial consensus-building work—aligning definitions, clarifying objectives, unifying measurements—that falls outside formal requirements.
- Tim frames it as “business partners on the same page,” essential yet invisible.
- Experimentation Culture:
- Mo reflects on the extensive, informal labor in fostering a culture of experimentation—internal advocacy, education, and maintaining enthusiasm—that’s essential but rarely spelled out.
Educating the Business & Organizational Data Fluency
13:17–15:07
- Analysts frequently build business stakeholders’ data fluency—but this, too, is undervalued and rarely acknowledged as core work.
- "The shadow work is building trust, building relationships—walking [stakeholders] at the appropriate pace." – Tim (14:57)
- The challenge: the expectation that data is pristinely clean or simple, and that the analyst can answer highly nuanced questions at a glance.
Technical Debt, Data Cleaning, and Re-engineering
16:19–23:41
- Stories abound of inheriting poor data warehouses, unstructured datasets, or chaotic pipelines, requiring vast cleanup/rebuilding efforts before value can be realized.
- "You walk into an org, they want to do amazing things, but the Snowflake instance isn’t going to do any of that until we clean it up." – Michael (17:45)
- "These are often huge and time intensive and unlock a heap of value, but people don't see the value until months later, sometimes a year." – Val (18:08)
- Business users and developers often underestimate the architectural work needed, assuming AI or better queries can fix bad structures.
- "Everyone thinks you can overcome a shitty data architecture with AI, which is just so fucked and hard to manage." – Val (21:53)
The Analyst as Organizational Glue—and Bottleneck
23:20–24:20
- Analysts become the cross-functional “glue” bridging business needs, technical delivery, and organizational siloes—but sometimes end up sole domain-holders and accidental bottlenecks.
- "You end up with one or two people in the business who know one area and no one else can do it... you’ve created your own bottleneck." – Val (23:20)
Who Should Do the Shadow Work? Personality, Team Fit, and Equity
24:20–36:00
- Generalists vs. Specialists:
- Michael and Mo reflect on how personality shapes comfort with shadow work; generalists often find it stimulating, while specialists may resent it.
- Value of Curiosity:
- Mo: "It always ends up feeling like it adds to the mosaic of my understanding, which pays dividends in the future."
- Documentation No One Reads:
- Tim shares a story about compiling a comprehensive process document that was greatly educational for him, ignored by others.
- Leadership and Recognition:
- As teams grow, leaders must actively surface shadow work, recognize who does it, and distribute it fairly—avoiding overburdening those who are agreeable or less likely to object (noting this often impacts women disproportionately).
- "Half of your job is unseen shadow work. And you can’t advance in your career…good old Jane is always there for it, but it’s not visible." – Tim (33:43)
- As teams grow, leaders must actively surface shadow work, recognize who does it, and distribute it fairly—avoiding overburdening those who are agreeable or less likely to object (noting this often impacts women disproportionately).
- Team Composition:
- Teams should mindfully hire to fill gaps in strengths and manage shadow work equitably.
Pitfalls: Recognition, Job Descriptions, and Team Visibility
36:00–52:21
- Reporting & Recognition:
- Val: "If you do that shit well, you can unlock a lot for your team or business. I want to make sure that’s rewarded and reflected."
- The 'Conveyor Belt' Fallacy:
- Mo: "It was always hard to message up...Do you think we just sit there like a conveyor belt, analyze, analyze?"
- When Shadow Work Reveals Workflow Flaws:
- Michael discusses how shadow work can expose larger organizational process gaps that need system-level fixes.
- Job Descriptions:
- The group notes that shadow work is often absent from formal job postings—especially when non-analysts do the hiring.
Data Quality, Hygiene, and The Eternal Struggle
39:13–47:47
- Data Quality as Shadow Work:
- Cleaning up poorly formatted spreadsheets from agencies or finance is a perennial burden—and business stakeholders often have no idea how significant the lift is.
- "Every city they run media in is a completely different format. That is a huge lift...very senior data scientists are spending their time basically QAing data." – Val (40:31)
- Cleaning up poorly formatted spreadsheets from agencies or finance is a perennial burden—and business stakeholders often have no idea how significant the lift is.
- Alerting, Anomalies & Getting Blamed:
- The expectation that analysts should have "caught" every spike or drop in metrics is ever-present, but unrealistic given data complexity and noise.
- "The analyst gets blamed if the data...wasn’t there for weeks...How did you not notice?" – Tim (43:16)
- Many maintain “secret dashboards” for self-monitoring.
- The expectation that analysts should have "caught" every spike or drop in metrics is ever-present, but unrealistic given data complexity and noise.
Data Literacy and Training – Shadow Work or Not?
47:58–56:21
- Discussing whether data fluency/education is shadow work or core work; agreement that it's critically important, ongoing, and rarely done solely through "tick the box" programs.
- "Every time I talk to a stakeholder, I’m trying to help them get a little bit further in how they think and understand data...heavy cognitive load, but incredibly important." – Val (50:15)
- Mo: "If your role is to help the business make smarter decisions, making sure you’re connecting findings, observations, recommendations with what the business can actually do...that’s core, not shadow work." (50:51)
Notable Quotes & Memorable Moments
- On the persistence and importance of shadow work:
- "Not all shadow work is shit. Some shadow work is actually very valuable. The business just doesn’t realize how necessary it is." – Val (10:29)
- On bottlenecks and documentation:
- "You end up with one or two people in the business who know one area and no one else can do it...you’ve created your own bottleneck." – Val (23:41)
- On gender dynamics and admin burdens:
- "I do think I’ve seen stuff written that women are much more likely to get screwed on this one...‘Mo’s really good at that’...it’s absolute shit work and she’s not going to speak up." – Tim (33:43)
- On leadership and surfacing hidden labor:
- "When you turn from an individual practitioner...into a leader, you need to take the shadow work and expose it to the light." – Michael (35:01)
- On job descriptions:
- "The ones that actually have the shadow work articulated...describe a realistic and practical role." – Tim (54:35)
- On data quality pain:
- "Every single city they run media in is in a completely different format...very senior data scientists spending their time QA’ing data. It’s really frustrating." – Val (40:31)
Important Timestamps
- 00:14–03:22 — Defining shadow work and the admin trap
- 05:37–07:59 — Data teams as accidental project managers
- 10:29–13:17 — Alignment, consensus and business impact
- 17:45–23:59 — Technical debt, data re-engineering, unsung labor
- 24:20–36:00 — Personality fit, team structure, and the equity of invisible work
- 39:13–46:36 — Data quality, anomaly detection and getting blamed
- 47:58–56:21 — Data literacy: is educating the organization shadow work?
- 56:30–60:00 — Wrapping up: organizational recognition, job descriptions, and the value of shining a light on shadow work
Takeaways & Conclusion
- Shadow work is deeply embedded in analytics, often necessary for real business impact—but seldom visible or celebrated.
- There’s a fine line between being “unwaveringly useful” (Cassie Kozyrkov, via Val, 28:15) and being trapped in menial work that stalls your progression.
- Leaders should illuminate shadow work, recognize and distribute it fairly, and ensure it's reflected in both job descriptions and rewards.
- Team composition, communication, and organization design can help balance necessary shadow work—making the invisible, visible.
- Data literacy/fluency efforts, technical debt cleanup, and process alignment are all examples where shadow work has high leverage—deserving recognition and resourcing as much as front-line analysis.
Final Thought
Shadow work may live in the margins, but it's essential for turning analytics insights into business action. As Michael concludes:
"A lot of work is really important, but doesn’t necessarily get recognized for what it is. The work has value, and who does it—and how it’s surfaced—matters."
