Why map before changing anything
Teams usually change a process because something hurts: a launch slipped, a report was late, a lead waited too long for a reply. The natural response is to fix the step where the pain was felt, and it often fails, because the place where work was noticed to be late is rarely the place where it was delayed. A registration page that goes live late may have been waiting for approved copy, which was waiting for a decision on the speaker, which nobody knew was theirs to make.
A map makes that chain visible before anyone spends effort on a fix. It shows the whole route on one surface, so the team argues about the same picture rather than about their memories of different parts of it. It separates the time spent working from the time spent waiting, which is usually where most of the elapsed time goes, as the process hub explains for cycle timeThe elapsed time from an accepted request to live work, read stage by stage to show where work waits; a process measure, not a target for individuals.. And it shows the handovers, where information is lost and ownership blurs, which written procedures tend to skip because each procedure describes one person's task.
The marketing operations hub suggests tracing the last few campaigns backwards as part of a first review. Mapping is the fuller version of that exercise, and it works for any recurring process, not only campaigns. The campaign workflow guide already draws the target route for campaigns with its stages and gates, so there is no need to redraw it; a map of how your campaigns actually travel can then be laid beside that route to see where practice departs from it.
| Process worth mapping | Where it usually starts and ends | Why it is a good candidate |
|---|---|---|
| A webinar or event | Idea agreed to follow-up sent | Many handovers between marketing, web, speakers and sales |
| The monthly report | Period closes to pack delivered | Waits on data from several systems; see the monthly reporting pack SOP |
| A product or landing page | Request accepted to page live and tracked | Copy, design, web and approval loops |
| Inbound lead handling | Form submitted to first contact | Crosses into sales; see the inbound lead handling playbook |
| A paid campaign | Brief signed to campaign live | Creative, trafficking and tracking dependencies |
Before you start
- Know the campaign route. Read the campaign workflow guide so that the stages, gates and handovers it names are familiar; the map uses the same vocabulary.
- Choose a facilitator. Someone who is not the owner of any step keeps the map honest. In a small team, the person mapping can be the marketing lead, provided they draw what they hear rather than what they expect.
- Gather the records. Work items with their stage history, the briefs, approval threads, launch checks and messages about the items you will trace. The work management guide explains which reports a work system can give you once stage changes are recorded.
- Find the people. Everyone who touched the items, including people outside marketing such as the web team, sales or a supplier. A process map drawn only by the marketing team stops at marketing's edge.
- Prepare a surface. A wall, a whiteboard or a shared drawing space with room for lanes. Avoid a polished template at this stage; the map will be redrawn several times.
The steps
Choose one process and fix where it starts and ends
Pick one process that recurs, crosses at least one handover and causes a visible problem: late launches, repeated rework, complaints from the people who request it. Then write down its start and end as events anyone could recognise, such as "the webinar topic is agreed" and "the follow-up email has been sent to attendees and no-shows". A process without fixed ends grows during mapping until it describes the whole department. If the team disagrees about where the process ends, that disagreement is often the first finding, because it usually means a step after the official end has no owner.
Pick a few recent items that went through it
Choose a few real items that went through the process recently, enough to show the usual route and some variation, and not so many that tracing them takes over the exercise. Include at least one that went well and one that went badly; a map built only from the bad case exaggerates its problems, and one built only from the good case hides them. Avoid items so old that nobody remembers them or so unusual that they will never recur. Write down which items you chose and why, so the map can later be checked against the same evidence.
Trace each item from start to end with the people involved
Follow each item through the records first: when it entered each stage, who it was assigned to, when it was returned and why. Then walk the route with the people who did the work, item by item, asking what they received, what they did, what they waited for and what they passed on. Ask about the steps that never appear in the work system: the message asking for a missing file, the copy pasted from an old email, the check someone does on their own because a previous launch went wrong. Those invisible steps are often where the time goes and where the knowledge sits.
Draw the steps in lanes by role
Draw one horizontal lane per role, not per person: requester, campaign owner, designer, web developer, approver, sales. Place each step in the lane of the role that does it, in the order it happens, and write it as a verb and an object, such as "write registration copy" or "build registration page". Keep each box to one piece of work done by one role without interruption. Draw what the items actually did, including steps the procedure does not mention, and leave out steps the procedure mentions but nobody performs; mark those separately, because they are findings too.
Mark every handover and what passes across it
Wherever work crosses from one lane to another, draw the arrow and label what passes: the signed Creative BriefA short document telling the people making creative work what it must achieve, for whom, with which message and within which limits., approved copy, a file in a stated format, a list of links with their UTMURL tags identifying traffic source. parameters. Then ask the receiver whether it usually arrives complete. Handovers where the receiver has to chase something are the most common source of waiting and rework, and the map should mark each one. Where nobody can say who receives the work, or two roles both believe the other owns the next step, record it as an ownership gap to settle later with a RACI matrix.
Mark waits, loops back and workarounds
Between steps, mark where the item waited and what it was waiting for. Read the waits from timestamps where the work system holds them; where it does not, ask the people involved and mark the answer as an estimate. A small number of items gives a picture, not a measurement, and reading small numbers explains why one long wait should not be treated as the usual one. Draw every loop back to an earlier step, with the reason it went back. Mark workarounds in a different colour: the spreadsheet kept on the side, the approval given in a corridor, the step done twice because the first output could not be trusted.
Check the draft with the people who do the work
Put the draft in front of everyone who appears on it, together, and ask what is wrong or missing. People correct a picture far more readily than they describe a process from scratch, and the conversation usually surfaces steps, waits and workarounds the tracing missed. Keep the tone factual: the map describes how the process behaves, not how well anyone performs in it. When two people describe the same step differently, draw both and note it; a step done in two ways is a finding, and often the reason the same item comes back for rework.
Find the step that limits the whole process
Look for the place where work queues longest or where most loops return. The theory of constraints explains why this step decides the pace of the whole process, and why improving any other step adds nothing until it is relieved. Check the candidate against the records rather than the loudest complaint. For a loop that keeps coming back to the same step, ask why it returns until the answer is something in the process, using the five whys; a returning loop is often caused upstream, at a handover where something arrived incomplete.
Choose the few changes worth making and keep the map
List the possible changes the map suggests, then choose only the few that act on the limiting step, a repeated loop or an ownership gap. Write each as a change to the process, with an owner and a way to see whether it worked: a field added at intake, a step moved earlier, a handover given a definition of complete, an owner named for a step that had none. Recurring tasks the map revealed with no written procedure become candidates for the SOP tutorial. Changes that compete with other operations work for time belong in one backlog with everything else operations wants to fix, ordered as the guide to prioritising marketing operations work describes. Date the map, store it with the process documents and redraw it after the changes have run for a while; the second map shows whether the constraint moved.
Conventions worth keeping
A map is easier to read, compare and update when every map the team draws follows the same few conventions. They do not need a formal notation; they need to be applied the same way each time.
| Element | Convention | Why it helps |
|---|---|---|
| Lanes | One per role, named as a role, never a person | The map survives staff changes and does not read as blame |
| Steps | A box with a verb and an object | Shows work done, not departments |
| Handovers | An arrow labelled with what passes across | Makes incomplete handovers visible |
| Waits | A marked gap with what the item waited for and the source of the timing | Separates waiting from working |
| Loops back | An arrow to the earlier step, labelled with the reason | Shows where rework starts and why |
| Workarounds | A different colour or marker | Flags steps that exist because something upstream fails |
| Decisions | A box naming who decides and on what | Shows where an approval or choice holds work |
| Evidence note | Which items were traced and when the map was drawn | Lets the next map be compared fairly |
Keep the current-state map and any proposed future-state map as separate drawings with the same lanes, so the difference is easy to see. When the team reports on the process later, measures such as cycle time and rework stay measures of the system, not a KPIA metric tied directly to an objective. for the people in its lanes.
A software company maps its webinar process (illustrative)
Mardoquin Software maps its webinar process
Mardoquin Software, an invented company selling HR software to mid-sized employers, ran a monthly webinar that regularly opened registration late and sent its follow-up email a week or more after the event. The marketing lead chose to map the process from "topic agreed" to "follow-up sent", and traced 4 recent webinars: 2 that had gone smoothly and 2 that had not.
The map had 6 lanes: product marketing, content, web, the webinar host, approvals and sales. Tracing showed that the web developer built the registration page from draft copy, product marketing then changed the copy at approval, and the page went back to the web lane for a rebuild. On the 2 late webinars this loop happened twice, and the page waited 9 and 11 working days in the web queue in total, against 3 days of building. The timestamps for the queue came from the work system; the building time was the developer's estimate, marked as such.
The second finding was at the end of the process. The host presented, then handed the attendee list to nobody in particular. On 3 of the 4 webinars the follow-up was eventually sent by whoever noticed it had not been; on one it was never sent. Nobody had owned that step since a content manager left the previous year.
The team chose 2 changes. Registration copy would be approved before the page was built, with the approved copy attached to the web request as its definition of complete. The follow-up email was given an owner in content marketing, with a handover from the host defined as the attendee and no-show lists delivered the day after the event. A redrawn map after the next 3 webinars showed one pass through the web lane each time and every follow-up sent within 2 working days.
Common mistakes
- Mapping the procedure instead of the process. Copying the documented route into boxes produces a tidy map of something nobody does. Trace real items.
- Leaving the ends open. Without a fixed start and end the map grows until it covers everything and helps with nothing.
- Lanes named after people. A lane called by someone's name turns every wait in it into a judgement on them. Name roles.
- Drawing only the work, not the waits. Most elapsed time sits between steps. A map without waits points at the wrong fix.
- Mapping without the people who do the work. The invisible steps, the workarounds and the private checks only appear when the people who perform them are in the room.
- Treating estimates as measurements. Waits recalled from memory, or read from a handful of items, are a starting point to check, not a figure to report.
- Fixing every finding at once. A map always shows more problems than a team can absorb. Choose the few that act on the limiting step and leave the rest on the list.
- Filing the map and never redrawing it. A map is a snapshot. Redraw it after the changes have run, or the team cannot tell whether they worked.
