Automatic vs. Passive vs. Timer-Based Legal Timekeeping

“Automatic,” “passive,” and “timer-based” are often presented as competing categories of legal timekeeping software. They are not quite parallel.

Timer-based and passive describe how evidence of work is captured. A timer depends on a person starting and stopping a clock. Passive capture observes supported work in the background without requiring that action. Automatic timekeeping is better understood as the larger workflow: turning evidence of work into a reviewable time entry with less manual reconstruction.

That distinction matters. Capturing six minutes is not the same as producing a correct entry for those six minutes. A usable entry may still need a client, matter, duration, narrative, task code, activity code, and an assessment against the client’s billing rules. It must then be reviewed by a person before it enters the billing system.

The useful comparison is therefore not “Which method has the most AI?” It is:

1. How does the system learn that work happened?
2. How much of the billable entry does it construct?
3. Who verifies and approves the result?

Timer-based timekeeping

A timer is an active capture method. The timekeeper starts the clock when work begins, stops it when the work ends, and switches timers when moving between matters.

Timers are well suited to long, bounded tasks on a known matter. If an attorney spends an uninterrupted hour drafting one motion, a running clock can provide a clear duration. The attorney knows why the timer is running and retains control over when the record begins and ends.

But elapsed time is only one part of a legal time entry. The attorney may still need to choose the matter and codes, write a narrative, separate non-billable activity, and check the client’s requirements. A timer can also be forgotten, left running during an interruption, or kept on the wrong matter after the attorney switches work.

Those weaknesses become more pronounced during fragmented days. A short email, a quick call, and a return to a document may each be legitimate pieces of work, yet each requires another start, stop, or switch. The method is only as complete as the timekeeper’s timer discipline.

Some systems improve this workflow with reminders or “smart timers.” Hourglass, for example, supports a timer mode that also captures the work performed while the timer runs and uses that evidence to draft the narrative. The clock remains an intentional signal from the timekeeper, but it does not have to remain an isolated duration with all the descriptive work left for later.

Passive timekeeping

Passive timekeeping captures evidence in the background while the attorney works in supported applications. It does not require the attorney to start and stop a timer for every task.

This approach is particularly useful for digital work that arrives in small pieces: reading and sending email, moving among documents, attending online meetings, researching in a browser, and handling calls through a connected phone provider. Instead of depending on memory, the system can assemble a record of what happened and when.

Passive capture does not, by itself, create a billable entry. An activity log that says “Microsoft Word was active for 18 minutes” does not establish the matter, the purpose of the work, whether all 18 minutes were active work, or how the task should be described to a client. The value depends on what the system can responsibly derive from the evidence and how easily a timekeeper can correct it.

It also creates evaluation questions that a timer does not raise to the same degree:

• Which applications and devices are supported?
• Does the system collect application names, metadata, or substantive work context?
• How does it distinguish idle time from work?
• How does it reconcile two sources describing the same event?
• Can a timekeeper pause or exclude capture?
• Who can see the underlying evidence, and how long is processing data kept?

More captured activity is not automatically better. The goal is not to place every minute of computer activity on an invoice. It is to create a stronger record from which a person can identify and describe legitimate work.

Automatic timekeeping

“Automatic” is the broadest and least consistent term. It can refer to an automatically started timer, passive background capture, AI-assisted matter suggestions, or complete draft generation.

For a useful evaluation, define automatic timekeeping by its output. A substantially automated workflow should turn captured evidence into a draft that may already contain the duration, client, matter, narrative, codes, and billable status needed for review. It should also explain what happens when the evidence is insufficient.

That last point is essential. A blank field can be safer than a confident guess. If the system cannot determine the matter or code, it should make the uncertainty visible and let the timekeeper complete the record. Automation should reduce reconstruction, not conceal it.

Automatic draft generation should also remain distinct from automatic submission. Under the ABA Model Rules, fees must be reasonable, and the person billing the work remains responsible for what the invoice represents. The ABA’s guidance on generative AI likewise emphasizes competent review of AI output and protection of client information. See Formal Opinion 512. Software can prepare a record; it should not become an autonomous billing authority.

Comparing the three approaches

These approaches can coexist. A litigation attorney might use passive capture throughout a day of email and document work, then deliberately start a smart timer for a focused drafting session. An in-person meeting may need to be added or completed manually. The test is whether those signals become one coherent record without duplication.

Consider a scheduled Zoom meeting. A calendar event and Zoom attendance may both describe the same meeting. Counting each source separately would inflate the duration. Hourglass correlates evidence describing the same event so the meeting is proposed once, with the resulting duration still subject to the timekeeper’s review. That is an example of why signal reconciliation matters as much as signal collection.

The same principle applies when related work spans applications. Reviewing a document and then drafting an email about it may form one task, while an unrelated interruption should not. A system should make its grouping editable: the timekeeper needs to be able to inspect, split, merge, reject, and add to the proposed entries.

The limitations a firm should test

No capture method observes every kind of legal work.

With Hourglass, ordinary calls placed through an iPhone or Android phone are not captured unless the call runs through a supported provider such as 8x8, RingCentral, Teams Phone, or Zoom Phone. Apple restricts access to ordinary phone logs, and Hourglass does not currently provide a native Android capture application. Paper review, travel, in-person conversations, and other work away from a supported source may also require the timekeeper to complete the record.

Passive evidence can be incomplete for less obvious reasons as well. The right document may be open while the attorney is speaking to someone about a different matter. A personal interruption may occur beside client work. Several applications may be active at once. Firms should test those cases during an evaluation rather than rely on a generic accuracy claim.

They should also ask to see what happens after a correction. Does a weak suggestion remain visibly incomplete? Can the timekeeper edit every field? Does the system retain the original evidence, suggestion, edits, approval, overrides, and export history? These behaviors show whether the workflow is designed for accountable review rather than frictionless-looking demos.

How Hourglass approaches the workflow

Hourglass timekeeping supports passive capture and smart timers rather than treating either as the only acceptable way to work. It gathers evidence from supported email, calendar, document, meeting, phone, and browser applications, then uses that record to propose the duration, client, matter, narrative, codes, and compliance result. If the evidence does not support a field, the field can remain blank for the user to complete.

The timekeeper retains full control of the draft. Entries can be inspected and edited before export, and nothing reaches the firm’s billing system without explicit human approval. Approved entries are then sent through the firm’s configured billing-system integration; the flow does not imply that Hourglass receives downstream deductions or collection data from that system.

For buyers, the meaningful unit of comparison is the entire record: the evidence captured, the duration calculated, the matter selected, the narrative and codes drafted, the compliance review performed, the human approval required, and the final delivery to the billing system.

A timer may be the right starting signal for focused work. Passive capture may be the better safety net for a fragmented day. Automatic drafting can reduce the work between either signal and an entry. The strongest fit is not a label; it is the combination that reflects how the firm actually works, exposes what the system does not know, and leaves the final billing decision with a person.