What Interviewing 20 PMs Taught Us About the Difference Between "Project Management" and the Real Work

Project management is over architected. We have many tools, methodologies and schools of thought, yet PMs are still drown in endless work. We sat down with 20 PMs to dig deeper.

1. "Where it can kind of break down is when you have updates happening in Slack that really should be noted in the project-management tool."
2. "We have like four different note-taker tools--it's really annoying."
3. "In ClickUp people were not able to understand where the source of the ticket is, so everyone was here and there and the context was lost."
4. "Updating tools--checking the tickets, sending the right email... that's 40 % of the job."
5. "No, I have to manually do everything."
6. Updating a Gantt is always a hassle... after a month it's already a mess."
7. "There's going to be inevitable slip everywhere... being proactive as opposed to reactive."

These quotes come from twenty PMs we interviewed across SaaS, fintech, biotech, and deep-tech. Although many PMs we talked to never admitted their pain, but we sure sensed their headaches, "sweet, and tear" across the screen.

We believe LLMs finally make it possible to close that gap. With the right embeddings, chunking strategy, and knowledge graph, an AI can serve as a context-aware lubricant that keeps projects moving without adding "yet another tool" to maintain.

Instead of finding an industry vertical, we may have a found a functional vertical, that is, many PMs experience the tireless communication across siloed teams, scattered data sources, and stakeholders who never share a single source of truth.

We also deeply understand that a great product doesn't start with ideation, but with validation. Below is a synthesis of what those twenty PMs told us on what is holding them back or making their life harder.

---

Here is what we found

Out of the 20 PMs we interviewed, we captured and fully transcribed 17 in-depth sessions (≈ 40 min each). The remaining 3 were short exploratory chats that lacked actionable detail. Across the 17 interviews, 100 % (17 / 17) confirmed that manual, low-value work consumes a majority of their week. The next most pervasive pain was tool fragmentation: 16 / 17 (94 %) complained about information silos and switching fatigue.

Shared Pain-Point Themes

Manual grunt work and tool sprawl aren't just the top two complaints, they're statistically inseparable. The only PM who didn't mention about silos runs a three-tool stack while everyone else with ≥ 5 tools reports fragmentation and, by extension, manual patch-up duty. In other words, the minute a PM's stack hits a specific complexity, the tedious, copy-pasting heavy manual work kicks in.

We finally understand why people say the majority of PM work is still "dirty work" or "glorified babysitting." People are consumed by manually stitching information and communicating it to different stakeholders. Honestly, I was shocked by how similar their situations were--even though we intentionally spoke with PMs across vastly different verticals, from biotech to SaaS.

Industry Breakdown

Across the 17 in-depth interviews, Enterprise and PE-backed operations account for the largest share (24 %) of our sample, averaging nearly 5 tools per PM. Independent power-users (solo practitioners or side-hustle use cases with no clear employer context) form the next-largest slice (18 %), they juggle the leanest stacks at just 3 tools on average. Digital-product startups, service agencies, fintech vendors, and vertical-industry saas firms each represent roughly 12 % of the cohort, with tool counts around 4-5. At the long-tail, government-research and manufacturing PMs make up 6 % apiece, but their work demands far heftier arsenals, 8 and 10 tools respectively. We are convinced that domain complexity, rather than company size alone, drives tool proliferation.

---

Food for thought

1. Manual work is still the default, not the outlier.

We watched every PM in our study reserve calendar blocks for manual data migration and stakeholder meetings. It is recurring, scheduled labor that crowds out discovery and strategic thinking. Once a team labels these activities as avoidable waste, the conversation naturally shifts to automation and process redesign rather than incremental speed. The first win is not how fast we push tickets but how often we can avoid touching them at all.

2. Tool counts grow faster than head counts.

The average PM uses nearly five core apps before counting shared drives and shadow spreadsheets, and the curve is steeper in young companies than in large ones. Each additional tool promises efficiency but inserts another place for context to hide. A deliberate slowdown in tool adoption can produce a faster overall cadence.

3. Fragmentation and re-keying are inseparable twins.

Wherever silos appear, we find manual stitching usually performed by the most senior PM on the project. Trying to consolidate tools without eliminating the copy paste rituals simply forces those rituals into a different format. A sustainable fix must lower both the number of tools and the human glue that binds them.

4. Reliability outranks depth for integrations.

One failed Slack to board rule can erase months of trust because everyone scrambles to recheck work that was assumed complete. Integrations that self monitor and alert users when they fall behind build more trust than a flashy interface with unpredictable uptime. Depth can follow later, first we prove that data always arrives where it is promised.

5. Solve the core, the edge cases soften.

Our data show that once manual effort and silos are reduced, secondary pains like scope creep, compliance friction, and deadline slippage decline sharply in both frequency and severity. These edge cases thrive in the noise created by fragmented systems and repetitive updates.

6. Context loss thrives in medium sized stacks.

Teams with two or three tools rely on shared memory while those with nine or ten lean on full text search across documentation. The danger zone is a stack of five to eight applications where information hops just enough to scatter but not enough to be searchable.

7. Domain complexity drives the outliers.

Government labs and heavy industry shops manage eight or more systems not out of preference but out of necessity. Standard tool consolidation playbooks fail here unless they account for strict data provenance and cross domain language differences.

8. Recovered time converts directly into morale and retention.

Saving a PM three hours per week returns roughly 150 hours per year, which is equivalent to nearly a full month of work. This can translate into numbers like reduced turnover costs, faster onboarding for new hires, and steadier release cycles. Showing leaders the compounded value of each hour reclaimed turns process improvement from a nice to have into a budgeted priority.

---

We are still interviewing

Building a product is the ultimate truth-seeking exercise that never ends. I don't think we will ever find the truth but will be increasingly closer to it (maybe super intelligence will change this paradigm?)

We are aiming to talk to 100, 200, 500, 1000 (as many as we can) team leads, PMs, and founders to dig deep into this functional vertical. We would love your input.

Feel free to schedule an interview with me to help us reach the first 1000-PM milestone