Stop Starting Work By Accident

Work that starts by accident finishes by accident. Ask three people on your team when a case enters the prep stage and you will get three answers, all of them reasonable, none of them the same. That is not a training gap. That is a missing field.

Firms obsess over what happens inside a stage. The checklist, the template, the assignment. Almost nobody defines the moment the stage begins. So the work starts whenever someone happens to notice it should — which is to say, whenever the owner mentions it.

01The moment nobody defined

An entry trigger is the specific, observable event that moves a case into a stage. Not a date on a calendar. Not a feeling that things are ready. An event: the retainer clears, the filing is confirmed, the records arrive, the consultation ends.

Look at your case board and you will find stage names and probably owners. What you will almost never find is a written statement of what causes a case to land in each one. That absence is why the same case type takes eleven days at one desk and three at another. Nobody disagreed about the work. They disagreed about when it starts.

Every stage has a start. If you didn’t design it, you didn’t remove it — you just handed it to whoever happens to be paying attention.

Law Firm Architects · Process Design

02What drift actually costs

Undefined starts do not announce themselves. They show up as symptoms that get blamed on people instead of design. A case sits for four days because everyone assumed someone else had picked it up. A paralegal starts prep too early and redoes half of it. The owner becomes the human start button, because the owner is the only one certain enough to press it.

• Cases idle between stages while everyone waits for a signal that never comes.
• Two staff members start the same stage at two different moments, so cycle time is noise.
• Work begins before its inputs exist, and the rework gets counted as normal.
• The owner announces starts out loud, which is delegation in reverse.

None of that reads as a design failure in the moment. It reads as a busy week. But measure the same case type across three matters and the variance is not effort, it is the start line moving. You cannot improve a cycle time you cannot locate the beginning of, which is why most firms have never measured one honestly.

03What a real entry trigger looks like

A real trigger is an event, it is observable in a system you already use, and it leaves no room for interpretation. “Signed retainer uploaded and initial payment cleared” is a trigger. “Once intake is wrapped up” is a wish.

The test is blunt: could someone who is not you look at the matter and know, without asking, whether this stage has started? If the answer requires your judgment, you have not written a trigger. You have written another reason to be interrupted.

The best triggers are already sitting in your systems, unlabeled. A payment posts. A signature returns. A court portal updates. Your firm treats these as information. Design them as instructions and the same events that used to generate a status question start generating the next stage instead.

Notice what this pairs with. An exit condition tells your team when a stage is finished. An entry trigger tells them when it begins. Together they turn a label on a board into a stage that runs on its own.

04Default vs. Designed

One sentence per stage. That is the entire intervention, and it changes who the firm depends on to get moving.

05Write one per stage this week

Take your highest-volume case type. Walk it stage by stage and write the single event that puts a case into each one. Make it observable. Make it something a system can see, because a trigger a platform can watch is a trigger a platform can fire.

That is the sequence. Define the trigger first, then automate it. Automating a start nobody defined just makes the guessing faster. Once the event is written down, the stage no longer needs your attention to begin — and the firm stops waiting on you to say go.