What a process map shows, and what it hides
A process map is a graphical representation of the chain of activities that turn a request into a result: who is involved, in what order, with which tools, which decisions and which delays. Flowchart, flow diagram, BPMN notation (the standard for modelling processes), swimlanes by department: formats abound, and they are roughly equivalent. The format decides nothing. What decides everything is the source.
There are two possible maps of the same process. The first describes the prescribed process: the one in quality procedures, standard operating procedures, tool configuration. It gets drawn in a workshop, with the managers, in half a day. The second describes the real flow: the one a purchase order, a case file or a complaint travels through on an ordinary Tuesday, with its re-keying between systems that ignore each other, its unofficial approvals by email, its parallel spreadsheets and its detours via "the person who knows". The first map reassures. The second explains.
Prescribed work never covers all of real work
This gap is not the sign of a failing organization. Francophone ergonomics established it thirty years ago: in the founding work of the field, Comprendre le travail pour le transformer (Guérin, Laville, Daniellou, Duraffourg, Kerguelen, Éditions de l'Anact), the actual activity of workers constantly translates and reworks the material and organizational conditions of the work. In other words: no process runs the way it is written, because reality (the incomplete client, the tool that crashes, the absent colleague) demands trade-offs the procedure never anticipated. Real work fills the gaps left by prescribed work. That is precisely how you recognize that work is happening.
The market offers involuntary proof of the scale of this. Process mining software, whose sole promise is to reconstruct how processes actually unfold from system logs, is now worth more than a billion dollars a year, growing 32% (Gartner, Market Share Analysis: Process Mining Software, 2024), with Europe the top spender. Thousands of organizations therefore pay, every year, to discover what their own procedures do not tell them. That settles the question of whether the gap exists in your organization.
The methodological consequence follows on its own: mapping the prescribed process takes only a workshop; mapping the real one takes an investigation.
The five-step method
Step 1: bound it with a question, not a scope. "Map all our processes" condemns the project to decorative inventory. Start from a question a leadership team will have to settle: why do our quotes take eleven days to go out? Where does the client file get lost between sales and invoicing? The question sets the scope, the level of detail and the stopping point.
Step 2: collect the prescribed process. Procedures, standard operating procedures, tool configuration, job descriptions: gather the official version and draw it without correcting it. This map will serve as a comparison baseline, not a deliverable.
Step 3: listen to the people who do the work. The real flow is written in no document; it is told. Individual interviews with the people who perform each link (not only their managers), observation of a few concrete cases end to end, and one golden rule: follow the file, not the org chart. The quality of the map depends directly on how candid this collection is; anonymity and full coverage of the scope matter more than the sophistication of the drawing tool.
Step 4: compare the two maps. Overlay the prescribed and the real, then document each divergence: re-keying, back-and-forths, queues, redundant approvals, skipped steps, parallel circuits. For each, note its frequency and its apparent cost in time. This list of gaps is the real product of the mapping: the drawing is only its support.
Step 5: decide gap by gap. Each divergence calls for a trade-off, and only two exist: correct the real flow to bring it back toward the prescribed one, or revise the prescribed process to make official what the field has invented better. A mapping that does not lead to this series of decisions will have produced a nice diagram, nothing more.
Organisation & Diagnostics
Structurez votre organisation avec un diagnostic clair
Nos audits révèlent les leviers concrets pour aligner vos équipes et vos process.
Reading the gaps: workarounds as raw material
The managerial reflex when facing a workaround is to stop it. The diagnostic reflex is to read it. A spreadsheet running parallel to the CRM does not express indiscipline: it expresses a specification, the exact list of what the official tool cannot do, written for free by those who needed it. An approval that goes through email rather than the intended circuit traces the shortest path the organization still refuses to admit. Urban planners call desire paths those trails pedestrians wear into the grass by cutting across the paved walkways; your processes are full of them, and they show where the pavement should have gone.
Some organizations have turned this into a discipline. Since 2020, AT&T has invited its employees to flag the outdated processes, rules and tools that complicate their work: more than 270 suggestions handled and 3.6 million hours recovered in three and a half years, starting from irritants as tiny as an eight-dollar expense (Legg, MIT Sloan Management Review, 2024). The raw material already existed; it was enough to open a channel to collect it.
One last bias deserves attention before drawing anything: mapping led from the top inherits the biases of upward reporting. Harvard Business Review recently described the "two organizations" problem, the one in the dashboards and the one teams live, showing how each hierarchical layer smooths, summarizes and polishes the information moving up (Wolpert, 2026). A map built in a meeting room with managers alone undergoes exactly this smoothing: it documents the reported organization. The real flow is collected at the level where it flows.
From the map to the process audit
Mapping describes; the process audit judges. The first exercise produces a faithful picture of the flow and the list of gaps; the second quantifies those gaps, traces them to their causes, compares them against requirements (regulatory, contractual, economic) and ranks the fixes. Many organizations commission an audit when they do not even have a map of the real flow: the auditor then spends most of the engagement reconstructing it. The efficient order reverses the sequence. Map first, broadly and honestly; audit second, precisely, on the areas the map flagged as costly. And when several successive mappings surface the same gaps across different departments, the subject changes nature: it no longer belongs to process auditing but to organizational diagnosis.
Tools draw, they do not listen
There remains the question that dominates searches on the subject: which tool to choose? Honest answer: the one you already know. Miro, Lucidchart, draw.io or a photographed whiteboard all draw a decent map; BPMN notation brings rigour to teams that practice it and jargon to the rest. The bottleneck of a mapping is never in the drawing software: it is in collecting the real flow, which takes ten times more effort than tidying it up. Process mining deserves a status of its own: it objectively reconstructs flows from digital traces, which makes it an excellent gap detector, within the limits of what systems record. Yet the most instructive workarounds happen precisely outside the systems: the phone call, the corridor, the local file. The tool sees the traces; listening gathers the rest.
A successful process map is recognized by a simple test: show it to the people who do the work. If they answer "yes, that is exactly it, even the embarrassing detour", you have a decision-making instrument. If they answer "that is the theory", you have a poster.
Foire aux questions
What is the difference between process mapping and a flowchart?
What is the difference between process mapping and a flowchart?
A flowchart is a representation format: standardized shapes connected by arrows. Process mapping is the whole approach: bound, collect, compare the prescribed with the real, decide. A flowchart can render a mapping; it does not replace it.
How long does process mapping take?
How long does process mapping take?
The drawing, a few hours. Collecting the real flow, most of the time: from a few days to a few weeks depending on the size of the scope and the listening method chosen, sampled interviews or exhaustive listening. Beware of mappings delivered in a single meeting: they describe the prescribed process.
Who should take part in process mapping?
Who should take part in process mapping?
First, the people who do the work, link by link, not only their managers. Managers know the process as it should run; only front-line staff know how it actually runs. Involve both levels, but collect separately.
Process mapping and process mining, what is the difference?
Process mapping and process mining, what is the difference?
Process mining automatically reconstructs flows from system logs; it excels on heavily tooled, high-volume processes. Mapping by listening additionally captures what escapes the systems: trade-offs, workarounds, informal circuits. The two approaches complement each other, the second often explaining what the first detects.
Sources
Guérin F., Laville A., Daniellou F., Duraffourg J., Kerguelen A., Comprendre le travail pour le transformer. La pratique de l'ergonomie, Éditions de l'Anact (repr. Octarès): the reference work on the gap between prescribed and real work.
Gartner, Market Share Analysis: Process Mining Software, Worldwide, 2024: global market at $1.1B in 2024, +31.7% year on year. gartner.com
Wolpert I., "The Two-Organizations Problem", Harvard Business Review, June 26, 2026, reprint H098H2. hbr.org
Legg J., "How AT&T Employees Turned Process Gripes Into $230 Million Saved", MIT Sloan Management Review, 2024, reprint 65404. sloanreview.mit.edu
