The PtahCast Blog

Cost & Margin used to price a project by applying today's rate to every week left — right on day one, quietly wrong the moment the rate or the timeline changed. It now banks the real cost of each period as it happens, so the number stops lying the moment a rate changes.

Jira tracks work for engineering teams. Trello is the simplest kanban board there is. Neither one forecasts a delivery date. Here's what each is actually built for, and where PtahCast is a different kind of tool entirely.

One timeline, every epic on the board, bands generated automatically in value order — each one starting only once the epic ahead of it is expected to land. Switch between Aggressive, Commercial, and Safe and the whole thing recomputes, not just the top band.

Clients can now file a support ticket from the same portal as their delivery forecast, reply on it, and see it through to resolved. The ticket list carries a live Monte Carlo forecast for the whole queue, and each ticket gets its own plain-language ETA based on where it sits in line.

Score a risk Low/Medium/High on probability and impact, add a mitigation note, and it stays agency-private by default. If it comes true, materialize it straight into a Blocker with one click — pre-filled, linked back, no re-typing it into a second log.

Override the remaining scope or apply a throughput multiplier and PtahCast runs the same Monte Carlo engine against it — with your last real forecast rendered behind it as a dashed gray "ghost" curve for comparison. Nothing is saved; refresh the page and it's gone.

Every blocker now gets tagged Client, Agency, Third-party, or Other the moment it's logged, and PtahCast rolls the totals up on the board and in the client portal — so "the forecast moved" becomes a number you can actually hand a client.

Building in public isn't just a changelog. Here's a real bug PtahCast's own test suite caught before it ever shipped — a column-delete rule that would have silently destroyed archived ticket history — and why the number of tests matters less than what they're actually for.

The curve is more honest than a single date ever was, but it isn't self-explanatory. What to say before the chart goes up, what to skip, and how to keep a client meeting from turning into a statistics lecture.

The two get used interchangeably constantly, and most explanations skip exactly where each clock starts. Here's the precise difference, where the confusion actually comes from, and why a forecast built on the wrong one quietly lies to you.

Connect a PtahCast board directly to a Jira Cloud project or a Trello board, map their statuses onto your columns, and pull your tickets straight in — no re-typing a backlog by hand. One-directional, configured per board, and synced only when you click "Sync now."

A genuine, from-scratch explanation for anyone who's never encountered the idea — starting with a weather forecast and building up to why a single delivery date is missing exactly what a probability distribution provides.

"Monte Carlo forecasting tool" isn't one category, it's four — Jira marketplace plugins, standalone flow-metrics analytics platforms, enterprise PPM tools, and agency-focused client-facing forecasting. A practical map of where each one actually fits, and how to tell which one solves your problem.

Jira has no native Monte Carlo forecasting — that requires a marketplace plugin, or a different kind of tool entirely. What Jira gives you out of the box, what a plugin adds on top of it, and where PtahCast's client-facing approach is genuinely different.

Start here. Sprints, story points, kanban boards, cycle time, Monte Carlo simulation, P50/P85/P95 — how they all fit together into one honest answer to "when will this be done," in a single guide that links out to every post below for the deeper dive on each piece.

Most PM tools price by headcount. PtahCast's plans scale with client engagements instead — 3, 10, or 25. Here's the reasoning behind that unusual choice, and why per-seat pricing quietly discourages the collaboration a delivery tool should encourage.

Every post on this blog so far has been about the agency-facing board and forecast. Here's the other half — what a client sees on Agency and Scale plans, and why what's left out matters as much as what's shown.

Jira, Trello, Linear, and Asana are excellent at showing what's where. None of them tell you when it'll actually be done, or how confident to be. Here's the gap between a well-run board and an actual forecast.

A string of missed deadlines doesn't make the next one "due" to land on time. Why teams reason about estimates the way gamblers reason about a roulette wheel, and what to do instead.

A Gantt chart and a Monte Carlo forecast can describe the same project and imply completely different things about how certain the date is. Here's what each one actually assumes, and where each one belongs.

The cycle time scatter is the flow metric with the most nuance left on the table. Here's how to tell a genuine outlier from normal variance — and why deleting outliers before forecasting is usually a mistake.

Story points are the most common unit of estimation in agile teams, and one of the most misunderstood. Here's why they were never built to predict a calendar date — and what actually is.

Every PtahCast forecast carries a confidence label, not just a date. Here's what actually moves a forecast from Low to Medium to High — and why the label matters as much as the date itself.

Most quotes set a price and a delivery date off the same sloppy estimate — which means when that estimate is wrong, both promises fail together. Here's why that's worse than it looks, and how decoupling cost estimation from a real forecast fixes it.

Most SOWs promise a bare date, and that single line is where a disproportionate share of client conflict starts. A practical three-part structure — plus starting-point language — for a delivery clause built around a confidence level instead.

A Monte Carlo forecast hands you three dates, not one. Understanding what they mean is easy — deciding which one actually goes in the email to a client is a separate question. Here's a framework for making that call.

Teams watch velocity, deadlines, and backlog size. The number that actually predicts whether a delivery date holds is the one almost nobody limits on purpose: how many tickets are in progress at once. Here's the math behind why, and what it has to do with your forecast.

WIP over time, a cumulative flow diagram, a cycle time scatter, a throughput run chart — four charts, and two of them aren't just background data, they're the exact raw material your Monte Carlo forecast is built from. Here's how to actually read all four.

A brand-new client engagement has zero completed tickets — which means zero throughput history to forecast from. Here's how PtahCast seeds a synthetic Poisson distribution to solve that honestly, and two real bugs that showed why the lifecycle around it matters more than the math.

Every agency has given a client a date like "we'll have it done by the 15th" that was really just a guess wearing a suit. Monte Carlo forecasting replaces that guess with a distribution built from your team's real cycle time and throughput — thousands of simulated futures instead of one confident-sounding number.

A forecast is only useful if someone can check it later. An accountability log is a permanent, dated record of every forecast an agency has made, and whether it held up — the thing that turns a Monte Carlo forecast from a nice chart into something a client can actually trust.