Worker blog
Notes on running and growing a business
Practical ideas for growing your business, leading a team and putting AI to work, from John Roberts, Rene Bystron and the Worker team.
Latest
The work that starts when the meeting ends
Find the hidden work between leadership meetings and make it part of the routine, so the next meeting needs less explanation from the owner.
Could the business run for five days without you?
Use a planned week away to find the work that still waits for you, agree when the team should interrupt, and fix one dependency at a time.
Keep quarterly goals connected to business results
Check whether a finished quarterly goal produced the result it promised, then use the answer to decide what to continue, change or stop.
Would ten more customers make your business better?
Before you take on new customers, check whether your team and your own hours can deliver the work, then pick the fix that matches the real limit.
Have your weekly numbers ready before the meeting
Make weekly numbers dependable before the meeting: settle each one’s source, period and owner, then ask what decision it informs.
Did AI save time? Count the checking too.
Measure whether AI saves time by counting all the human work around it: checking, fixing, setup, upkeep and whose hours went up.
More leads will not fix a slow start
A signed customer is not always ready to start. Set realistic start dates and remove the delays before the first useful result.
Keep meeting decisions connected to the work
Keep the conditions behind a meeting decision, such as a needed approval, in the task record so the next person acts on what was agreed.
The customer who pays well and consumes your week
See whether a well-paying customer’s workload matches the service you sold, and agree a clearer arrangement before resentment sets in.
Give recurring problems a history
Keep the decision, fix and result with each recurring problem, so the team knows whether it needs a decision, an action or a different fix.
A two-week AI trial needs an end date and a pass mark
Set the scope, test cases and pass marks for a two-week AI trial before it starts, then end it with a clear decision to keep, revise or stop.
Keep processes useful after they are written
Find where a written process stops helping, agree a change, make the new version the only one in use, and check whether it worked.
Your first manager needs decisions they can make without you
Hand a first manager real decisions: clear limits, the facts behind your judgment and room for choices you would not make.
Growth should come with a stop-doing list
Make room for a new opportunity by choosing what to stop, reduce or pause, then check that the time actually comes free.
See where teams need help without attending every meeting
See what a useful team update says beyond “on track,” so you know where teams need your help without sitting in every meeting.
The risk hiding in “only Pat knows”
Turn “only Pat knows” into real coverage: pick one job, teach the judgment behind it and test the backup on a real run.
Make operating routines work across teams
Adapt your meeting routine to fit a branch or other team, and hand off what it cannot solve alone without the COO carrying every update.
Before AI sends a message, decide who can make the promise
Before AI helps answer customers, decide which promises a message can make, what allows each one and who approves the rest.
Bring the whole company into the system
See why access for the whole company matters, and how to check that an employee can raise a problem, see who owns it and hear back.
When faster drafts create slower decisions
Faster AI drafts can bury an owner in decisions. Learn to limit what enters review and what each proposal must answer first.
Your growth plan needs a life plan
Put one personal boundary next to your revenue target, then work out what the business must change so growth does not cost you the life you want.
Choose the recurring work worth delegating
Use four checks to pick the recurring work worth handing off, with weekly meeting preparation as a worked example.
Stop treating every customer request as urgent
Sort customer requests by what is at risk, promise updates your team can keep, and offer rush options that do not break other promises.
Give your business coach a clearer view
Give your coach a short briefing: what changed, what happened, where the team disagrees and which decision you need help making.
Your AI needs current answers, not more documents
Learn to pick the source that answers each customer question, mark which documents are current, and test whether your AI follows them.
What would make changing systems worthwhile
Before you switch systems, name the job that must improve, test it through an ordinary bad week and check who ends up doing more work.
Independent should not mean doing everything yourself
Owning a business should not mean doing all the groundwork yourself. Get help preparing decisions and keep control of what you build.
A small business does not need to be everywhere
Pick channels by what customers do there, count the weekly work each one takes and test a new one with clear limits before you commit.
Goals your team can finish
Five habits that help a team finish its quarterly goals: one owner, a finish line you can check, a link to the year, a short list and a weekly check.
Why a full calendar can hide a slow business
Measure the waiting between steps, not just the work, and test whether earlier handoffs get orders out sooner without overtime.
Build your scorecard from numbers you already have
Pick a few weekly numbers you already track, give each an owner, a target and a definition, and decide which misses need action now.
What happens when your AI gets it wrong?
When an AI workflow sends customers wrong details: stop it, find who was affected, fix their situation and prove the repair before restarting.
Solve the cause, not the symptom
Name the problem, check the suspected cause against the work, give the fix an owner and a date, and confirm the improvement holds.
The next customer should not require another app
Before adding an app for one customer, follow a single job through it to see what work it removes, what it adds and what to test first.
AI can give you capacity. What will you do with it?
When AI frees up hours, decide what they are for. See how one owner splits six saved hours and handles a request that needs more.
No posts match that search.