From Stuck to Strategy: A Project Manager’s Guide to Visual Thinking with AI
- Andrew A. Rosado Hartline
- 7 minutes ago
- 8 min read
Rewiring Communication and Work
In 2023, I read a book that genuinely shifted the way I think, work, and communicate: Stuck? Diagrams Help by Abby Covert. It wasn’t just an instructional book, it was an empowering invitation to think differently, a practical framework for how to get “unstuck” by drawing your way to clarity.
And yes, that’s exactly what I mean: drawing. But not in the artistic sense. This book helped me realize something we’re never really taught: how to actually make diagrams. Sure, we’re asked to read flowcharts, Gantt charts, mind maps, roadmaps, swimlanes, and product maps. But few of us are ever taught how to create them well or how to use them in ways that truly clarify thinking, align teams, and communicate ideas to others.
It wasn’t just an instructional book, it was an empowering invitation to think differently, a practical framework for how to get “unstuck” by drawing your way to clarity.
Why This Book Hit So Hard
What Covert does brilliantly is break down diagramming as both an individual thinking tool and a shared communication tool. This hit especially close to home for me as a project manager, because simplifying complex systems or ideas is half the battle. We often assume visuals like Gantt charts are inherently useful. But useful to whom? Diagrams only work if someone else can actually read them, understand them, and do something with them. “Explaining things so a wide range of people can understand it is a mixture of art, skill, experience" and truly a science as Covert clearly points out.
Covert treats this as a professional craft. Diagramming isn’t just something creative people do for fun; it’s a practical methodology. And yes, there is a science behind it... Rather, several
In Chapter 5, Covert references dual coding theory—a principle from cognitive psychology that suggests information is better understood when presented through both verbal and visual formats (Paivio, 1986). That explains why diagrams are often more effective than text alone.
Taking that into consideration, we can also talk about Cognitive Load Theory which suggests that people can only hold a limited amount of information in their working memory at once. Diagrams reduce this load by organizing complex ideas visually, making them easier to process (Sweller, 1988). While Covert doesn't name this theory, she does mention it numerous times. Her approach to breaking diagrams into “containers, connections, labels, metadata, and visual variables” is about managing cognitive complexity and thus reducing the load.
Finally, Covert also weaves in information architecture principles—how people navigate, interpret, and assign meaning to information throughout the whole book, but predominantly in Chapter 4 "Diagram Recipes". This overlaps with semiotics, the study of signs and symbols. Her advice on using consistent labels, scalable structures, and visual hierarchy is grounded in these disciplines. She treats diagrams as information environments, not just pictures. That’s why context, audience, and intent matter so much.
A New Way to Collaborate and Think
Another key insight for me was the collaborative potential of diagrams. These visuals aren't just personal tools for self-reflection (though they can be!); they are meant to be shared. Covert emphasizes the need to design diagrams that deliver a message, not just capture a moment of insight. Chapter 4 especially stood out—she calls them “diagram recipes,” a brilliant metaphor for thinking through intent, audience, scope, and format.
Before reading this, I never really stopped to ask:
Who is this for?
What are they trying to do with it?
Will they understand it at a glance?
That kind of intentionality transforms how I build timelines, process maps, or stakeholder visuals today. It’s no longer about completing a task, it’s about communicating something clearly.
AI + Diagrams: From Thought to Clarity
Pairing Abby Covert’s philosophy with generative AI and diagramming syntax like Mermaid gives PMs like me a thinking edge. It allows for:
Quick externalization of ideas
Iteration without needing graphic design skills
Sharable clarity, not just internal insight
Prompting an AI like Qwen, Mistral, and DeepSeek to generate a flowchart is one of the fastest ways I now think through problems. It's not just time-saving, it's thought-sharpening.
Here’s a typical example:
```mermaid
flowchart TD
Sales --> Qualification
Qualification --> Proposal
Proposal --> Kickoff
Kickoff --> Delivery
Delivery --> QA
QA --> Handoff
```
It’s not about perfection. It’s about clarity. But let's expand on it!
✅ Use Case 1: Cross-Team Handoff Confusion
The Problem:
At a mid-sized SaaS company, the handoff from Sales to Implementation is a repeated pain point. Sales closes the deal and sets expectations with clients, but the delivery team receives little context, no timeline awareness, and minimal documentation. Projects kick off late or scramble to recover.
AI + Flowchart Solution:
The PM prompts ChatGPT:
“Create a flowchart to visualize our current process for onboarding clients from Sales to Implementation.”
Initial Draft (Iteration 0):
```mermaid
flowchart TD
Sales -->|Closes Deal| AccountManager
AccountManager -->|Briefs| ProjectManager
ProjectManager -->|Schedules Kickoff| ImplementationTeam
ImplementationTeam -->|Delivers Project| Client
```
Iteration & Refinement
First Iteration:
```mermaid
Flowchart TD
Sales -->|Deal Closed| AccountManager
AccountManager -->|Discovery Call| SolutionsEngineer
SolutionsEngineer -->|Creates Requirements Doc| ProjectManager
ProjectManager -->|Schedules Kickoff| ImplementationTeam
ProjectManager -->|Sends Welcome Packet| Client
ImplementationTeam -->|Executes Project Plan| Client
```
Final Diagram (Iteration 2):
```mermaid
flowchart TD
Sales -->|Closes Deal| AccountManager
subgraph Parallel_Onboarding
AccountManager -->|Discovery Call| SolutionsEngineer
SolutionsEngineer -->|Req Doc| ProjectManager
AccountManager -->|Triggers Autom. Welcome Email| MarketingOps
end
ProjectManager -->|Capacity Check & Sprint Plan| PMO
PMO -->|Green Light| ImplementationTeam
MarketingOps --> Client
ImplementationTeam -->|Delivery & Handoff| Client
Client -->|Survey| CustomerSuccess
```
Why this matters: Each iteration was drafted by an AI, but can be iterated based on team feedback, practice, experience, and feedback, marked-up asynchronously by the relevant team, then regenerated in minutes so the diagram itself became the shared single source of truth.
Potential Impacts (with real-world data)
Before we dive into diagrams and workflows, it’s worth pausing to ask “Does this actually move the needle?” The short answer is yes, substantially so. When teams make their hand-offs explicit, visual, and repeatable, the benefits show up not just in anecdotal “feels better” feedback but in hard-number performance metrics across industries as varied as software, construction, and healthcare.
The research below highlights what happens when a rough sketch evolves into a living, shared process map—complete with discovery steps, capacity gates, and documented scripts. Spoiler: even single-digit percentage gains balloon into weeks of schedule recovery, six-figure cost savings, and dramatically fewer “dropped baton” errors. That’s why the next section unpacks the data—and why your first draft diagram is already worth the effort.
Where standardized handoffs have been introduced in other industries, researchers report:
Schedule adherence improves by ~4 percentage points (63 % vs 59 %) and budget adherence by ~5 points (73 % vs 68 %) among teams that share cross-functional “business acumen”—essentially a common process language like the diagram above.
Project failure rates drop from 11 % to 8 %.
Rework costs can consume ≈ 9 % of total project spend when communication breaks down; clearer, gated processes directly attack that waste.
When a structured hand-off script (I-PASS) was rolled out across 32 hospitals, adherence to all required elements jumped 231 % for verbal handoffs (and > 600 % for written), showing how fast a well-documented workflow can scale once codified.^2
Even modest percentage-point gains translate into days shaved off go-live, thousands saved in rework, and a measurably better client experience and all by iterating a handoff diagram in public with the teams who live it.
The feedback loop and fresh perspectives you can get from AI and your teams can add another dimension to your process planning at an organization regardless of the maturity.
✅ Use Case 2: Risk Mitigation Planning for Data Migration
Problem:
A project team is preparing a data migration with known risks: incomplete data audit, third-party API reliability, and rigid go-live dates. Leadership wants a clear mitigation strategy—text documents haven’t been working.
AI + Flowchart Solution:
Prompt: “Generate a Mermaid flowchart to represent the following project decision and process tree. The flow should begin at Project Kickoff, branch based on data audit status, classify data, trigger backup/cleansing operations, and include conditional branches for API stability, dry-run validation, go-live timing, and post-migration monitoring”
Generated Flowchart:
```mermaid
flowchart TD
A[Project Kickoff] --> B{Data Audit Complete?}
B -- No --> B1[Blocker: Run Full Audit] --> B
B -- Yes --> C[Classify Data Severity]
C --> C1[Low-Risk Tables] & C2[High-Risk Tables]
C1 --> D[Backup Snapshots]
C2 --> E[Backup + Cleansing Scripts] --> E1{PII Compliant?}
E1 -- No --> E2[Mask & Encrypt]
E1 -- Yes --> F[Integrity Verified]
D & F --> G{External API Stable?}
G -- No (Sev 1) --> G1[Activate Retry + Failover Vendor]
G -- Yes --> H[Dry-Run Migration]
H --> I{Dry-Run Passes SLOs?}
I -- No --> I1[Roll-Back • Adjust Buffer]
I -- Yes --> J[Schedule Go-Live Window]
J --> K{Go-Live < 10 hrs?}
K -- No (Sev 2) --> K1[Trigger Contingency Buffer]
K -- Yes --> L[Final Migration]
L --> M[Post-Migration Monitoring]
M --> N{Error Rate < 1 %?}
N -- No --> N1[Hot-Fix Patch & Roll-Forward]
N -- Yes --> O[Project Closeout]
```
Impact:
Leadership understood the plan at a glance.
Teams assigned mitigation owners.
Project delivered on time despite mid-migration API errors.
When in Doubt, Draw It
We’ve all seen a bad diagram. Some are beautiful but confusing. Others are cluttered but only understandable to the person who made them. Covert teaches us to simplify, iterate, and check for clarity at every step. She calls this the balance of clarity over creativity. That one idea alone could fix 80% of slide decks out there.
On a personal note, this was one of the most fun nonfiction books I’ve ever read. It blends art, science, and storytelling in a hands-on, highly engaging way. Even the book itself is a diagram it starts at the back, leading you through a flowchart that helps you figure out where to begin depending on your level: beginner, practitioner, or diagram nerd. Covert’s curated literature review and resource list at the end is also gold. I've already started reading a few, and I can see myself going back to this book again and again, not just for diagrams, but for thinking better in general.
A friend initially recommended this to me when they were stuck. They said, “I know you love diagrams, but I used this to get unstuck personally and I know you could use it for so much more.” They were right. Since reading it, I haven’t stopped diagramming and I’m proud of that. I’ve built countless workflows, storyboards, decision trees, and even personal timelines (and those who know me personally, know that I use this for D&D as well!). With tools like Mermaid and AI, I now iterate both content and form faster than ever.
I turn to diagrams first when I need clarity, not last. And I’ve learned to ask:
Does this make sense to others?
Is the message clear, not just complete?
Am I showing or just telling?
I recommend this book to every project manager, process thinker, or just anyone who’s ever said, “I don’t know where to start.” Because the truth is, you start by drawing it out.
Call to action: Grab a scrap of paper (or open your favorite diagramming tool like Freeform, draw.io, MindMup, MindMeister or whatever you have and sketch a quick flowchart of the biggest problem on your plate today. Then drop a note in the comments and share what you discovered and where you got stuck or unstuck.
Cover image created using ChatGPT