Conditions and branches
Where a workflow chooses. One arm runs, the other does not, and everything hanging off the arm that did not run simply never happened.
Sooner or later a workflow has to decide something. In this product a decision is a box like any other — which means it is visible on the canvas, and a rule you can see beats a rule buried inside a sentence somewhere.
A fork is a node
Open 3.1 · Call every new lead back within 5 minutes. Its second box asks one question about the enquiry that just arrived, and the two arrows leaving it are the two answers.
3.1 · Call every new lead back within 5 minutes
One question, two arms. A big enquiry goes to a branch head within thirty minutes; everything else is called back the same day.
Scroll for all 5 steps →
Three fields make the rule: a value to look at, a comparison, and something to compare it against. Here that is the size of the enquiry, >=, and 2875000 — a business decision written where anybody can read it and a client can argue with it.
Only one arm runs
This matters more than it sounds. The nodes on the arm that was not taken did not run slowly, or run and do nothing — they did not run.
| If the answer is Yes | If the answer is No |
|---|---|
| The high-priority task is created. | It is not. |
| Anything wired after that task also runs. | It does not, and nothing says so. |
| A later reference to that task resolves. | A later reference to it resolves to nothing. |
The high-priority task is created.
- If the answer is No
- It is not.
Anything wired after that task also runs.
- If the answer is No
- It does not, and nothing says so.
A later reference to that task resolves.
- If the answer is No
- A later reference to it resolves to nothing.
A real run makes the point better than the table does. On 18 August a ₹33.2 lakh enquiry went through this workflow, took the Yes arm, and the run record lists four nodes in a five-node workflow. The fifth is not marked skipped or failed. It is simply not there.
Read where the arms go, not just where they start
Both arms end in the same place. The task each one raises is different — priority against same-day — but they both run on into the note that records “enquiry received and routed automatically”. A fork does not have to stay forked: the two arms differ for exactly as long as they need to, then rejoin. That shared ending is why the note is written once rather than twice.
Tracing an arm to its end is a five-second habit worth having. What you are looking for is the difference between “these two arms do different things” and “these two arms do different things and then the same thing”. Read it wrong and you will tell a client something is missing when it is simply drawn further along.
There is also a many-armed version of this box for when two answers are not enough — a router with several named paths and a fallback. Your org uses one, and it comes up in Module 3 where there is room to do it properly.
Change one thing
Move the business rule and leave it moved — the next lesson depends on it.
- Open the condition box and change
2875000to10000000— one crore. - Save. The rule is now “only enquiries above a crore get a branch head”, which in this business means almost none of them.
- Do not run anything. Go to the next lesson and fire it for real.
Next: Draft and active: which workflows are actually running — before you make anything run for real, which of the nineteen are running at all.
Related lessons
Base rates — what a piece of evidence is actually worth
A face-recognition system that is 99.9% accurate and almost entirely wrong, and a number that sent an innocent woman to prison. Both are the same arithmetic, and it is the arithmetic that decides what any piece of evidence is worth.
ReadConfirmation and survivorship — what you never looked for
Two questions about evidence you did not go looking for. One is a rule you have to discover, and one is a pattern in five famous people — and in both, the thing that would have told you the truth is the thing nobody checks.
ReadLoss aversion, sunk cost and regression — what it costs you
Four questions you answer about yourself rather than about a scenario, and your own answers are the finding. Then the pattern that makes praise look useless and criticism look like it works, whatever you actually do.
Read