Mapping the current process (as-is): step by step
Mapping the current process records how a process really runs today, not how the manual says it should. You get there by talking to the people who do the work, looking at their workplace and the documents they use, and walking through one real case together: “what happens next?”. Times and volumes are estimated as ranges in the same conversation.
Why the current process comes first
Every improvement is measured against a starting point. If the starting point is the process as described in an old procedure, the analysis will miss the detours that cost the most: the phone call because a field is missing, the printout for the signature, the invoice that sits in someone’s inbox during their holiday.
Mapping the current process also creates a shared picture. Each department usually knows its own part well and the rest only vaguely. Seeing the whole path in one diagram is often the moment when people notice that a case crosses the same two desks three times.
Preparation: scope and participants
Before the first conversation, agree on where the process starts and ends, who the result is for and roughly how many cases there are. This takes half an hour and prevents the mapping from drifting into neighboring processes. See scoping a process with SIPOC.
Then choose the participants: at least one person per role who does the steps personally, the process owner, and someone who moderates. Managers who only know the process from reports are welcome to listen but should not be the ones describing it.
Four sources of information
A good mapping of the current process combines several sources, because each one alone is incomplete:
- Interviews: short individual conversations with each role. People speak more openly about workarounds when their manager is not in the room.
- Walking the workplace: watch a case being handled at the desk. You see the second screen, the paper list next to the keyboard and the folder that is not in any diagram.
- Documents and forms: collect the forms, templates, lists and emails that are used. Each one belongs to a step and often explains a query or a media break. See media breaks and forms.
- The workshop: all roles together, walking through the process from trigger to result. This is where handoffs and disagreements become visible.
Running the workshop
Many teams still work with brown paper and sticky notes on the wall, then type everything up afterwards. Mapping directly into a tool on a projector saves the typing and lets the group see the numbers while they talk. Either way, follow one real, recent case rather than the general idea of the process.
Start with the trigger and ask “what happens next?” until you reach the result. Put each step in the lane of the role that does it. At every decision, ask what the alternatives are and how often each one occurs. Write problems on the step where they occur, not on a separate list – that keeps them attached to the place where a measure will have to act.
Record open questions instead of guessing
Almost every mapping session leaves something open: nobody in the room knows who approves above which amount, or the volume has to be pulled from the system first. Do not guess. Write the question down and get it answered before the next session.
In the editor, add a “To clarify” entry to the step (in the step’s details under “Problems, ideas, measures”), type the question and, if you like, who answers it by when. The step then shows a blue flag. Once the answer is in, tick “Clarified” and enter the outcome. Open questions are also listed under Measures in the analysis, in the report and in the shared view, so everyone can see what is still missing before the next session.
Questions to ask at every step
These questions collect the data that the analysis needs later:
| Question | What it gives you |
|---|---|
| How often does this happen per month? | Volume, and shares at decisions |
| How long do you actually work on it? | Processing time |
| How long does it lie before you start? | Waiting time |
| What do you need to start, and what if it is missing? | Queries, query rate and the delay they cause |
| Which system, form or list do you use? | IT system, artefacts, media breaks |
| Who else can do this? | Number of people per role and their availability |
Estimating times as ranges
Nobody in an office measures their processing times with a stopwatch, and nobody needs to. Ask for three values instead: how long does it take in the best case, typically, and in the worst case. People find this easier than a single number, and the range shows how uncertain the estimate is.
For example, a back-office employee entering an invoice into the ERP might say “usually six minutes, four if everything is clean, twelve if I have to look up the supplier”. The waiting time before that step might be “half a day, sometimes two days”. Record minutes of working time; a working day is the number of working hours per day.
The calculation then runs three times – with all best, all typical and all worst values – which gives a range for the lead time as well. How the times combine into a lead time is explained in lead time calculation.
Typical traps
- Describing the future process: people tell you how it should work. Ask about the last real case instead.
- Ignoring exceptions: note them, estimate how often they occur, and decide later whether they need their own path. See case types and exceptions.
- Too much detail too early: a step is something one role does in one go. “Open the email, read it, save the attachment” is one step.
- No volumes: without the number of cases there is no cost per year and no utilization.
Step by step: from an empty canvas to the metrics
The “Purchasing via purchase requisition” template in four stages, the way a process takes shape in the editor. Drawn with the app’s export from the model it calculates with.
1. Empty: just the lanes, i.e. who takes part in the process. 2. Steps: tasks, decisions and connections in the order of the flow, still without times. 3. Times: processing and waiting time per step. From these the app already calculates the lead time; without case volume and headcount there is no utilization yet. 4. Metrics: with case volume (500 / month), headcount and cost rates, cost, staffing and utilization are added: lead time 12.9 working days, cost per case €38.15, staffing 2.4 FTE, highest utilization Goods receipt at 76%.
Try it in the editor
Current-state capture
The empty template starts with a trigger, a result and two lanes, so you can map the process step by step during the conversation.
Open the example in the editorFrequently asked questions
Should I map in a workshop or in individual interviews?
Ideally both: short interviews to collect details and workarounds, then one workshop to put the whole path together and resolve contradictions.
What if people give very different times?
Record a range from best to worst case and ask why the values differ. Often the difference points to an exception that deserves its own path, or to a query that should be recorded separately.
How detailed should the current process be?
Detailed enough that each step has one role, a processing time and a waiting time. Going below that level rarely changes the result of the analysis.
Can I start with an empty diagram?
Yes. The empty template contains only a trigger, a result and two lanes; you add steps as the conversation progresses.
Self-test
Keep learning
Would you like to map or improve a process together with us? Contact