The appointment was filed, the announcement went out, and for a few months two things are true at once. You have more permission to change how the company works than you will again, and you have less information about how it works than anyone who reports to you.
Over the trailing twelve months, SEC full-text search returns 36 filings announcing the appointment of a new chief operating officer. Those are filings rather than companies. What they share is the timing problem: a mandate that is widest at the start, and an understanding of the operation that is thinnest at the same moment.
The Window Is Real And It Is Short
The seasons of an executive's tenure (Hambrick & Fukutomi, Academy of Management Review, 1991) sets out distinct phases, with the earliest characterized by the widest information search and the greatest willingness to depart from inherited practice — and with commitment to a fixed approach hardening as tenure continues. The organization also grants more latitude early, because nothing currently broken is yet attributed to you.
Nothing in the job description tells you what to do first. The choice is yours, it is made on incomplete information, and the window in which it is cheap to make closes quickly.
The Operating Review Produces A List
The standard first move is an operating review, and it is the right move. Within six weeks it produces a set of findings that are almost identical across companies of this size:
- Some processes have no owner who can give their status without asking.
- Two people enter the same data into two systems, which disagree at period end.
- One critical spreadsheet is kept by someone who has never taken two weeks off.
- The enterprise system covers the standard path; everything unusual goes by email.
The findings are correct. The problem is what happens to them. They become an initiative list with owners and dates, and roughly a third require software that does not exist. Those items go to the information technology group, which owns the enterprise system, network and endpoints, and does not build new internal applications; that was never its function.
That outcome is not a failure of the review. It is a capacity gap the review cannot detect, because a list does not separate an item somebody can do and an item nobody can.
Choose The Workflow You Can Prove
The useful discipline in the first quarter is to reduce the list to one item that can be changed and shown to the board, and to apply four tests to each of the candidates on it.
| Test | A good candidate | A poor candidate |
|---|---|---|
| Ownership | One executive owns the outcome | Three functions share it |
| Boundary | A defined start and end, inside one quarter | A programme with phases |
| Position | It changes something — an approval, a handoff, a record | It reports on something |
| Visibility | Its current cost is felt weekly by people you can name | Its cost is a modelled estimate |
The third test is the one most often failed, and it is worth saying why. Reporting improvements are easy to buy and change nothing about how the work is performed. The work to choose is in the path where something is committed: the approval before the next step, the handoff between teams, the entry that becomes the record.
If you arrived after an acquisition, the candidate is usually obvious, because one process now has to span two entities while the systems are due to merge in a later year.
Make The Change Provable, Not Assertable
A change that people can feel is not the same as a change you can show the board. The difference is whether the new process produces its own record of the work.
This is a design decision, made before anything is built rather than assembled afterwards. A process that records each item, who handled it, when, against which version of the input, and what exceptions occurred, gives you three things: a weekly performance number, evidence for the auditor, and a baseline to measure the next change against.
It also removes the meeting you will otherwise inherit. Most status meetings in a company this size exist because nothing in the process reports its own state, and the meeting is the only place the current state is assembled. Removing it is visible in the first quarter.
First Steps
- Take the review findings and mark each item with who would build it. A name, not a department. Items with no name are not scheduled, whatever date is next to them.
- Apply the four tests to the unnamed items and keep exactly one. One workflow, executive-owned, bounded, in the path where something is committed.
- Write down what record that workflow should produce, before discussing any solution. Who did what, when, against which input, and with which exceptions. That one page is the specification, and it is short enough to write in a single afternoon.
One Workflow, Inside The Window
Act within the first quarter: the choice is still cheap, the operating review has just produced the evidence, and the organization expects that something will change.
A fifteen-minute call — nothing paid, nothing signed — takes one workflow from your list apart: what it does today, what it would do instead, what would have to be built, what it costs, and what it will be measured against afterwards. If the honest answer is that the workflow is fine and the problem is elsewhere, that is what we will tell you.
Where something has to be built, that work runs as a monthly engineering partnership, with the targets for each release agreed before it starts, so the commitment is checkable rather than merely promised. The work runs in a repository we own, and paid-for deliverables transfer to you monthly — full history and documentation. The engagement runs in TopDo, our own system, where you can watch the work as it happens daily and keep the whole workspace afterwards. After that, $10,000 a month puts one accountable person on keeping the system running, with a written monthly report of what ran and what changed.
References
- Hambrick, D. C., & Fukutomi, G. D. S. The Seasons of a CEO's Tenure. Academy of Management Review, 1991.
