Operations tools
Twelve working tools that answer one operating question each, with your own numbers, before anybody sells you anything. They exist because the same questions come up on almost every first call, and the arithmetic is more useful than the opinion.
In short
This page lists twelve tools for operating decisions: what a recurring task costs, whether to automate it, whether to build or buy, what an inquiry is worth, where value is leaking, and whether you would pass a diligence check. Each runs in the browser, takes one to five minutes, and returns a specific answer rather than a score.
What we take on
Numbers, not opinions
Each tool takes figures you already have and returns an answer you can put in front of a board. No modeling, no assumptions you cannot see, nothing that needs explaining twice.
Willing to say do nothing
Several of them return that a task is not worth automating, that a build never pays back, or that you do not need an outside firm this quarter. That is what makes the other answers worth reading.
One to five minutes each
The health check is the long one at about five minutes. Most of the rest are a minute, because they ask for four numbers rather than a discovery process.
A written read if you want one
Every tool ends with the option to send the result over. What comes back is a page of specifics about your answers within a business day, from a person.
Why a calculator beats a conversation, at the start
Almost every operating argument inside a company is really an argument about a missing number. Should we hire, should we automate, should we build it, is marketing working. All four are decidable, and all four get decided by seniority instead, because measuring feels slower than asserting.
It is not slower. Most of these take a minute. What takes months is running a project chosen from a meeting, discovering halfway through that the constraint was somewhere else, and then having the same argument again with more sunk cost attached to it.
The tools here are deliberately narrow. Each answers one question and refuses the others. A task cost calculator will not tell you whether to automate, because those are different questions with different answers, and a tool that claims to answer both usually answers neither well.
They are also built to disagree with you. The automation scorecard returns a verdict of not yet when a process is undocumented, because automating an unmapped process encodes the mess along with the work. The build calculator will tell you a build never pays back. The health check refuses to let a strong total hide a weak area.
Nothing here is a lead magnet in the usual sense. There is no gate, no drip sequence and no account. The result is yours at the end whether or not you ever speak to us, and it is written to be pasted into a leadership document as it stands.
If you are not sure which one applies, start with the business health check. It is the only one that looks at the whole operation, and its job is to tell you which of the others is worth your time.
Twelve tools, grouped by the question they answer
Every one runs in your browser, needs no account, and ends with the option to send the result over for a written read. Pick by the question you actually have rather than by the name.
Not sure what the problem is yet
Start here. It is the only one that covers the whole operation rather than one question.
What is this costing us?
Four ways to put a number on work that nobody has priced, using figures you already have.
Task cost calculator
Work out what a repeating task costs in hours a week, hours a year and money a year, and what half of it back is worth.
Open the tool →Cycle efficiency check
Touch time against elapsed time on one workflow, showing what share of the cycle is pure waiting and where to cut first.
Open the tool →Owner bottleneck check
How much of the week routes through one person, and what the queue waiting behind them costs across a year.
Open the tool →Tool overlap check
Prices what moving data between overlapping systems costs by hand, then returns consolidate, integrate, or neither yet.
Open the tool →Should we do this or that?
Decisions that get argued from opinion until somebody attaches arithmetic to them.
Automation scorecard
Six questions scoring one task on volume, rules, stability, cost of failure, handoffs and documentation, returning a verdict on whether automating it will pay back.
Open the tool →Build vs buy breakeven calculator
Four numbers returning the month a custom build overtakes a subscription, or that it never does, then which way the fit actually points.
Open the tool →AI fit check
Four questions on one task returning which shape of AI fits it, where it would go wrong, or that nothing fits yet.
Open the tool →Where are we weak?
Diagnostics that score what exists today rather than what is planned.
Value leak diagnostic
Nine questions scoring revenue, capacity and control, showing which of the three is leaking most and what to do about it first.
Open the tool →Reporting readiness test
Six questions on where your numbers come from and who can produce them, returning whether you have a reporting problem or a data problem.
Open the tool →Diligence readiness check
Five areas a buyer, lender or investor tests, scored on what you could actually produce tomorrow morning without a scramble.
Open the tool →What is growth actually worth?
The number that makes marketing spend arguable instead of a matter of taste.
Questions we get
Which tool should I start with?
If you can name the problem, use the tool that matches it. If you cannot, start with the business health check, because naming the constraint is the thing it exists to do.
The most common honest starting point is that something feels slow and expensive without anybody being able to say where. That is the health check.
Do I need to give an email to use them?
No. Everything computes in the browser and the result appears whether or not you ever type an address. The email field only exists at the end, for people who want the result read back by a person.
There is no account, no drip sequence and no list.
How accurate are the numbers?
They are as accurate as what you put in, and the usual error is entering a task time that is too low because the chasing and checking around it goes uncounted.
None of them are forecasting models. They do arithmetic you could do yourself, on the inputs that actually matter, and their value is that they force the inputs to be stated rather than assumed.
Can I use the results without hiring anyone?
Yes, and that is the intent. Each result names a specific next action, and several of them tell you to do nothing at all.
The method behind them is published on this site in full, including the parts that argue against buying software or adding headcount.
What comes back if I send a result in?
A page or so of specifics about your answers: what a company shaped like yours usually finds underneath that result, the order to fix things in, and what the first two weeks would need from your team.
It goes to one inbox and is read by Shaheer Shaikh. Reply within one business day.
Do these work for a company outside the United States?
The arithmetic does, and the questions assume an operating company rather than a jurisdiction. Currency is entered by you, so any currency works.
The defaults are set for a US company somewhere between one and fifty million in revenue, which is who most of the work here is for.
More in the guides and every answer in one place.
Shaheer Shaikh, operations lead at LARVOL, a San Francisco AI company working with global pharma on clinical trial data and model benchmarking. Six Sigma on the process side, Anthropic certified on the Model Context Protocol, ten years across eight industries. More about the firm