From the team
Joust field notes.
Notes from Ron and the Joust team on technology decisions, client work, and lessons from building systems.
Featured
AI Snake Oil or the Real Thing?
Most AI snake oil starts with something real; the claims just get ahead of it. Ten questions that make an AI product's promise specific enough to evaluate, from a $20 tool to a company-wide system.
Ron Davis
Sep 10, 2026
When AI training doesn't stick
The session goes well, and a month later little has changed. Why AI training fades, the four things that make it stick, and what to check in the weeks after the workshop.
Ron Davis
Aug 12, 2026
The Survival Test
A price-sensitivity exercise for AI investments: consider what changes if the technology becomes much more expensive or much cheaper.
Ron Davis
Jul 15, 2026
A Better Mental Model for Understanding AI
Prediction is a useful starting point for understanding language models, their tool use, and why plausible answers still need checking.
Ron Davis
Jun 26, 2026
One resignation away
When a process depends on one person’s judgment, useful documentation starts with the exceptions they handle and the decisions they make.
Ron Davis
Jun 24, 2026
When recurring requests take over the day
Repeated status questions, approvals, and lookups can point to a process that needs attention. Start by looking at how the work arrives.
Ron Davis
Jun 17, 2026
When an AI pilot stalls
Three areas to examine when a pilot is not becoming useful daily work: the task, its ownership, and the way results are measured.
Ron Davis
Jun 3, 2026
What an Operating Review Looks Like, in Three Weeks
A composite walkthrough drawn from Joust distributor engagements: three workflows with $1.84M in estimated annual operating cost, and the recommendations behind the numbers.
Ron Davis
May 27, 2026
Mapping a workflow before an AI pilot
A practical starting point for an AI pilot: follow one piece of work, include the exceptions, and agree on who will use the result.
Ron Davis
May 20, 2026
The cost of recurring manual work
How to estimate the time spent on a workflow and decide whether changing it is worth a closer look.
Ron Davis
May 13, 2026
Tools, not gods.
A practitioner’s view of AI: useful applications, the work of running them, and the questions that matter before adopting a system.
Ron Davis
May 11, 2026
Getting a recommendation into production
A recommendation needs an owner, a workable scope, and a place in the team’s day. Here is what we look for before implementation.
Ron Davis
Apr 12, 2026
Tools second
Before comparing tools, agree on the task, the people using the result, and what would count as an improvement.
Explaining a difficult recommendation
How we approach recommendations that challenge an existing plan: show the evidence, explain the consequences, and work through the options.
Ron Davis
Feb 20, 2026
How to use the automation capture rate
The calculator’s capture rate is an assumption about repetitive work that could be automated. Here is how to test that assumption.
Build vs. buy: decide one workflow at a time
A useful way to compare software options: separate standard business needs from the work that depends on how your company operates.