Welcome to the documentation for Delivery Forecast by IMIRT, the AI-powered delivery intelligence app for monday.com. This guide covers installation, every feature, settings, dashboard widgets, and frequently asked questions.
1. Find Delivery Forecast by IMIRT in the monday.com marketplace and click Install.
2. Open any board, click the + button on the views bar, and select Delivery Forecast by IMIRT under Board Views.
3. The app will load and prompt you to configure your zone mapping.
Before the app can analyse your board, it needs to know how your statuses map to workflow zones. The Zone Mapping Wizard appears on first run and guides you through this:
Drag each of your board's statuses into the appropriate zone. The app uses this mapping to calculate throughput, forecast delivery, and power every other feature. You can change the mapping at any time from the Settings tab.
The five zones mirror how work flows through your team:
| Zone | What It Means | Analogy |
|---|---|---|
| Bench | Not started — sitting in the backlog | Players on the bench |
| Warm Up | Committed but not yet active | Warming up before the game |
| Doing | Actively being worked on | On the pitch, in play |
| Scoring Zone | Final review, testing, approval | In the box, about to score |
| Goal | Completed and delivered | Goal scored |
Every forecast in this app is powered by Monte Carlo simulation, a technique used in finance, engineering, and science to model uncertainty. Here's how it works in plain terms.
When someone asks "how long will this take?", the typical answer is a single number — "about three weeks." That number is almost always wrong. It doesn't account for interruptions, dependencies, sick days, scope changes, or any of the hundred things that routinely derail plans. Studies consistently show engineering estimates miss their targets by 30–50%.
Rather than guess a single answer, the app looks at what your team has actually delivered — how many items you completed each week over your recent history. Then it runs 1,000 simulated futures, each one randomly sampling from that real throughput data.
Some simulations pick high-throughput weeks (your team on a good run). Others pick low-throughput weeks (holidays, blockers, disruptions). The result is 1,000 plausible outcomes that reflect the full range of what could realistically happen — based on evidence, not optimism.
Instead of "three weeks", you get something far more useful:
The spread between these percentiles tells you how predictable your team is. A narrow spread means consistent delivery. A wide spread means high variability — and that's valuable to know before making commitments.
Runs 1,000 Monte Carlo simulations against your board's actual completion history to answer four key questions:
How Long — "We have 12 items. How long will they take?"
Enter the number of items and the app returns a range of likely durations at different confidence levels (P50, P85, P95).
When — "When will these 8 items be done?"
Enter a count and the app gives you calendar dates at each percentile band, so you can commit with the confidence level that suits you.
How Much — "How many items can we deliver in 4 weeks?"
Enter a time window and get a range of realistic throughput — how many items your team is likely to complete based on history, not aspiration.
How Likely — "Can we deliver 10 items by March 31st?"
Enter a target count and date, and the app returns a straight probability. No hedging — just the odds based on what your team has actually done.
Reading the results: Each forecast shows three colour-coded percentile cards: amber P50 (a coin flip — risky to commit), green P85 (the safe commitment), and blue P95 (conservative safety net). A coaching sentence between the cards and the histogram translates the numbers into plain-English guidance — what to commit to, where the risk sits, and what to watch.
The histogram below is colour-coded by delivery confidence: red bars are outcomes you'd be lucky to hit (<50% chance), amber are moderate risk (50–70%), green are the safe commitment range (70–85%), and blue are ultra-conservative (85%+). A legend below the chart explains the colours and ties them to the specific percentile values for your forecast.
The data quality badge at the top tells you how much history the forecast is based on: more history means tighter, more reliable bands.
Why it matters: Traditional estimates rely on gut feel and miss targets by 30–50%. Every forecast here is built from what your team has actually delivered — not industry benchmarks or optimistic guesses.
Shows how much your team can realistically deliver in a chosen time period, based entirely on historical throughput. Items are placed into three columns in priority order:
Items fill each bucket like water: once the green bucket is full, items overflow into amber, then red. There's no way to manually inflate the buckets — even adding people won't change them immediately, because new team members need ramp-up time and the data reflects what your team has actually delivered.
Drag-and-drop prioritisation: In board view, you can drag items between and within buckets to change their priority order. Grab the grip handle on any card and drag it to a new position. The bucket sizes stay fixed — they're determined by Monte Carlo simulation — but dragging re-slices the flat priority list, so promoting an item into the green bucket pushes the last green item into amber. When you're happy with the order, click Save Prioritisation Changes in the toolbar at the bottom. This writes an "IMIRT Priority" number to each item on your monday.com board and saves the order so it's restored next time you open Risk Buckets. Use Discard to revert unsaved changes. Items that changed bucket are highlighted so you can see the impact before saving.
How to use it: Share this view with sponsors and the business to agree on what must be delivered within the window. If the green bucket doesn't hold everything you need, the conversation shifts to prioritisation — drag the most important items into the green bucket and discuss what moves out. This makes trade-offs visible and concrete: promoting one item means demoting another.
Borrowed from basketball: when the shot clock runs out, you lose possession. Every active item has a progress meter per workflow state, measured against a time limit derived from your board's own averages.
The view breaks down by state so you can see exactly where work is ageing — Doing, Reviewing, Blocked — not just that something is old.
How to use it: Expand each state group to see which items are closest to their limit. Red items have overstayed — ask why. Is there a blocker? Does someone need help? Should it be split? The coaching message on each overdue item gives you a starting point.
Configuration: Choose between Auto mode (board average multiplied by a configurable factor) or Manual mode (set explicit time limits per state). Configure these in the Settings tab under Shot Clock.
Inspired by Expected Goals (xG) in football analytics, eXpected Delivery (xD) tracks your actual delivery performance against a predicted trajectory derived from your own historical throughput.
The chart shows:
How to use it: Check weekly. If the actual line is tracking inside the confidence band, you're on course. If it drops below, dig into what changed — new blockers, scope creep, team disruption. If it jumps above, ask why — sustainable improvement or quality risk?
Drift detection: The app automatically flags consecutive weeks of under-delivery with a drift alert, so you catch problems early rather than at end of quarter.
Analyses your workflow states — not individual items — to find where your process is breaking down. Two lenses are available:
Acute View — What's getting worse right now? Each state is scored on:
States are colour-coded red, amber, or green so you can see at a glance where the pain is.
Structural View — Where does cycle time actually go? This reveals states that consume a disproportionate share of every item's journey. If 50% of your cycle time is spent in Reviewing, that's your highest-leverage improvement target regardless of whether it's acutely overloaded today.
How to use it: Start with Acute to see what's deteriorating now — a red state with rising queue growth needs immediate attention. Then check Structural to understand the deeper pattern and prioritise long-term process improvements.
Builds a dependency graph from your board and calculates the actual probability of each item being delivered. Every dependency is a failure point: an item with three dependencies doesn't just have three risks — it has only one path to success and multiple paths to failure.
Two models:
Critical Chain View — Highlights your longest dependency path, which is your true delivery risk. Think of it like cyclomatic complexity for delivery: the more routes through the system, the more ways things can go wrong.
How to use it: Look at the blockers first — items blocking two or more others are your critical path. Unblocking them improves odds across the board. If an item shows 20% odds, trace its dependency chain to find the stuck link. The goal is fewer, shorter chains.
A visual, interactive view of where all your work sits right now — laid out on a themed playing field with workflow zones, item cards, and dependency lines.
What you see:
Themes: Choose from Soccer, GAA, or Generic layouts in Settings. Each theme changes the zone labels and visual style while keeping the same workflow logic.
Drag and drop: Drag items between zones to update their monday.com status directly. The board syncs immediately.
How to use it: Use it in standups and team discussions to see the whole picture at a glance. Congestion in the execution zone means too much WIP — stop starting, start finishing. Long dependency lines crossing multiple zones highlight unmanaged risks. Items with high days-in-zone badges are ageing — investigate before they become blockers.
Your IMIRT Game Manager — an AI coach that reads your board the way a performance analyst reads the pitch. It synthesises signals from every tab — throughput, risk buckets, shot clock, expected delivery, bottlenecks, dependency odds, and scoring zone congestion — into a single coaching statement.
What you get:
How to use it: Read the coaching narrative first for the big picture. Then explore the detail cards. Hit Refresh to regenerate coaching after board changes. Share the coaching report in standups and planning sessions to ground conversations in data, not assumptions.
Access settings from the Settings tab (gear icon). All settings are saved per board, so different boards can have different configurations.
IMIRT provides dashboard widgets that aggregate data across multiple boards — ideal for portfolio-level views and executive dashboards.
1. Open a monday.com Dashboard (or create one).
2. Click + Add Widget and search for "IMIRT" or "Delivery Forecast".
3. Select the widget type you want.
4. Choose which boards to include. The widget aggregates data from all selected boards.
Widgets combine throughput data from all selected boards to produce portfolio-level forecasts and analyses. Each board's zone mapping is respected independently, so boards with different status labels work correctly together.
What data does the app access?
Only the boards where you've added the app as a board view. We access item names, statuses, activity logs (status change history), and board structure. We never access boards outside the installed scope. See our Privacy Policy for full details.
How accurate are the forecasts?
Forecasts are based on your team's actual completion history using Monte Carlo simulation — the same technique used in financial modelling and engineering. Accuracy improves with more history: 8+ weeks of data gives reliable results; 12+ weeks is ideal.
Does it work with any board?
Yes, as long as your board has a status column. The zone mapping wizard lets you map any set of statuses to the five workflow zones.
What happens if I have very little history?
The app will still work, but forecasts will have wider confidence bands. The data quality badge shows you how much history is available. With fewer than 4 weeks of data, forecasts should be treated as rough estimates.
Can I use it across multiple boards?
Yes. Add the board view to each board individually, and use dashboard widgets to aggregate data across boards for portfolio-level views.
What are the subscription tiers?
Does adding more people change the forecasts?
Not immediately. Forecasts are based on historical throughput. New team members need ramp-up time before their contributions appear in the data. You can use the Parallel Work Streams setting to model planned capacity changes.
Is my data stored outside monday.com?
Forecasting calculations are performed on our secure servers and results are returned in real time. We do not permanently store your board data. Settings are stored in monday.com's own storage. See our Privacy Policy for details.
We aim to respond to all support requests within one business day.