The Bottleneck Heatmap Guide: Four Ways to See Where Your Process Actually Hurts

The heatmap exists for a specific reader: a small-and-mid-sized business owner, or the consultant working with one, who needs to know where the money is leaking in a process before committing to a transformation plan. Not every task deserves attention and spreadsheets hide the priority. Colour on the diagram surfaces it in under a minute.

A heatmap on a BPMN diagram is a colour overlay on every task node. The colour carries exactly one piece of information: how this node ranks, relative to the others in the diagram, on a metric you pick. The interesting part is that a diagram can tell you wildly different stories depending on which metric you colour by. The same process that looks calm when you colour by duration can scream when you colour by cost; the same process where the biggest cost is painfully obvious can hide a frequency problem that only shows up when you colour by impact.

LucidFlow's heatmap has four modes, not one, because a single view does not answer a single process question. Normal turns the heatmap off so the diagram is readable as a diagram. Cost colours by the dollar amount each task spends per execution. Duration colours by time-on-task. Impact folds cost, duration, and monthly frequency together to answer "which tasks drive the most total value through this process, in any dimension". You switch between modes in one click.

Normal

The heatmap is off. Every task node is rendered in its neutral theme colour; the diagram reads as a diagram. Normal is the view you want when you are editing: adding nodes, refining labels, rearranging layout: because the task colours carry no extra meaning and do not distract from structure. Most sessions start in Normal and only switch to a metric view once the structure is stable.

Cost

Tasks are coloured by their estimated cost per execution. The formula is straightforward: each task's cost value (usually labour cost for human tasks, subscription cost for automated ones) is normalised against the maximum cost on the diagram using min-max normalisation. The highest-cost task renders deep red; the lowest-cost task renders blue; everything else sits on the gradient in between. Swimlane parent nodes aggregate the sum of their children, so a lane whose members collectively cost more renders more intensely than a lane where cost is spread thin.

Duration

Same math, different input. Tasks are coloured by estimated duration in minutes. This is the view that matters when your constraint is clock time rather than budget, you are trying to shorten cycle time, reduce customer waiting, fit a process inside an SLA. Duration-heavy tasks often turn out not to be duration-heavy in wall time but in waiting time: messages awaiting replies, approvals sitting in queues. Those show up red under Duration and often do not show up at all under Cost.

Impact

Impact is the composite view: cost × duration × monthly frequency, normalised over the diagram. The mode answers the question "where is the value actually concentrated, given everything we know". A task that costs $2 per execution looks trivial under Cost, but if it runs 5,000 times per month it lights up under Impact. A task that takes five minutes per execution looks trivial under Duration, but at 10,000 executions per month it becomes one of your top three budget drivers. Impact catches patterns the single-dimension views miss.

A good heatmap reading takes less than two minutes. Pass 1: look for the red nodes. They are the candidates for investigation. Pass 2: check whether the red nodes are on the critical path: LucidFlow highlights the critical path separately, so a red node that is also on the critical path is the highest-priority finding. Pass 3: switch modes. If a red Cost node is also red under Duration, that is a genuinely bad task; if it is blue under Duration, you are looking at a cost-heavy but fast step: an automation candidate rather than a time-saving one.

1. Start in Normal, confirm the diagram is structurally correct before colouring anything.
2. Switch to Cost. Note the two or three reddest nodes. Click each to see the raw number.
3. Switch to Duration. Look for overlap with Cost reds (structural bottlenecks) vs. disjoint reds (time-only pain).
4. Switch to Impact. Watch for surprises: tasks that were blue or amber on single-dimension modes but red on composite.
5. Go back to Normal to edit, or open the cost dashboard for the full financial view.

The heatmap toggle sits in the canvas toolbar and emits a single mode value; the canvas re-renders tasks with the new colours in under 100 ms. The computation happens in the browser over the current diagram state, not on the server, so switching modes has no latency penalty and no privacy concern: your KPI numbers never leave your browser when you switch views.

Swimlane parents aggregate the sum of their children's KPIs automatically. This matters for "who is the bottleneck" analysis: if your Compliance lane sums to 40% of total cost but no individual Compliance task is a top-three red, the lane colour tells you the problem is distribution rather than a single hot spot. The cost dashboard reads from the same KPIs and formalises the financial version of the same finding.

The heatmap is the diagnostic layer, not the answer. For an SMB or the consultant advising one, the red tasks are the starting points of the conversation, not the end of it. The real value is the AI transformation plan that follows: which of those red tasks should be eliminated, simplified, standardised, integrated, or automated, and in what order. The heatmap tells you where to look; the transformation plan tells you what to do about it.