Getting mobile notary KPIs right the first time

A monthly review that takes two hours will be skipped by March. This one takes half an hour, fits on a page, and every line on it has a decision attached.

Eight measures. Each with a definition tight enough that two people would compute it the same way, a data source you already have, and a stated action if it moves the wrong way.

What to take away

The page

Eight rows, half an hour. Anything you add beyond this is what stops the review happening.

What each one triggers

Measure 1 alone triggers nothing. It is context for the others and it is the one people mistake for a result.

Measure 2 rising means the intake is failing. Look at the name check and the confirmation of who must be present before you look at anything else, because those two account for most of it.

Measure 3 rising for one service type means either that type has changed or your bands are stale. Re-time the drives before adjusting anything.

Measure 4 is the real profitability number. Falling for one service type means the travel line for that type is behind your costs, or the type is absorbing waiting time nobody priced.

Measure 5 falling means the calendar, not the driving. Windows are sized wrongly or the day is sequenced across the area rather than around it.

Measure 6 is the only direct evidence of quality you get, and any movement in it deserves a file review the same week rather than a note in a report.

Measure 7 widening means a cash problem forming, and it usually shows in one client type rather than across the board. Chase that type specifically.

Measure 8 shifting toward intermediaries means your own client base is shrinking behind a full calendar, which is a risk that only becomes visible when a contract ends.

The measures interact, and reading them singly produces bad decisions.

A rising completed count with falling contribution per hour means you are buying volume with distance. A falling failure rate with falling window reliability usually means you are accepting jobs further out. And a widening payment gap with a shifting source mix means your calendar is quietly becoming somebody else's.

That last combination is the one to watch, because it is the case for looking hard at geography and client base rather than at process, which is the subject of what expansion actually involves.

Setting thresholds honestly

Take three months of your own data and use it as the baseline. Then set a trigger as a movement rather than a level: any measure moving by a meaningful amount against last quarter gets looked at.

For example, if failure rate held at 7 percent across the baseline quarter and this month reads 12 percent, the move is five points against last quarter. That crosses the trigger, so the intake steps get checked before any other measure is read. A move of one point would not.

Resist the urge to import a target from somewhere. There is no reliable published figure for failure rates, door-to-door times or contribution in this trade, and a number that arrived from a forum will be defended long after it stops describing anything.

Where the numbers come from

Everything on the page comes from two habits: noting departure and return times for every job, and noting every booking that produced no completed act with the reason.

Keep those as ordinary business records, in the way the IRS describes in its guidance on what a business should keep. Keep them entirely apart from your notarial journal, which is a statutory record with its own contents and retention rules and is not a place to store analysis.

Protect job logs like the client data they partly are. They contain names, addresses and appointment details. A spreadsheet on an unencrypted laptop that lives in a car is a real exposure.

The National Institute of Standards and Technology's quick-start guides for small businesses and the Cybersecurity and Infrastructure Security Agency's resources for small and medium businesses set out controls at a scale that fits a one-person practice.

Running the monthly review

Once a month, at a fixed time, fill the page and answer three questions in writing.

The third question is the useful one. It catches the case where nothing moved because you stopped recording something, which happens more often than any genuine trend.

Sample filled page for one month. The figures are an illustration, not a benchmark.

Written answers: intake failed on two signings where the signer's ID was not confirmed at booking. Payment days rose with one title client, and arrival times went unlogged in the last week.

When there is more than one person

Compute measures 2, 5 and 6 per person as well as overall, because those three are the ones that identify a training gap rather than a market shift. Compute measure 4 only overall, since attributing revenue per person in a shared calendar produces arguments rather than insight.

Handle measure 6 carefully. A document query may point to a notarial defect, a compliance matter for that person's own commission, not a performance score. The distinction is in what a second commissioned notary changes.

Whether to add a person at all is a question this page answers directly, using measures 1 and 3 together. The selection protocol is in the four-stage way to test a candidate.

For the systems behind the page, see what a mobile notary's systems actually have to hold.

Common questions

Is eight measures too few?

It is the number that gets reviewed. A larger set that is computed twice and abandoned is worth nothing.

Should I automate this?

Only after doing it by hand for three months. Automating a measurement you do not yet trust produces a confident wrong number.

What if a measure will not move?

Then either the lever you are pulling is not connected to it, or the definition is wrong. Check the definition first; it is usually the definition.

Do these work for a part-time practice?

Yes, with a longer window. Compute quarterly rather than monthly so the numbers are not dominated by one unusual week.