WaterfallBridge is a KPI change explanation and variance bridge analysis tool. It explains why a metric changed by attributing the change to each dimension and key driver — for value metrics and for percentage / rate metrics.
WaterfallBridge is a KPI change explanation and variance bridge analysis tool.
It explains changes in value metrics and percentage / rate metrics by dimension and key drivers. It can explain revenue, margin %, GOP %, conversion rate, retention rate, yield, efficiency and other KPI movements. It is useful for finance teams, consulting teams, marketing analysts, operations teams and business leaders. WaterfallBridge runs as a web app, as a Microsoft Excel add-in, and as an AI-callable engine through MCP.
Every business reviews KPIs that change from one period, scenario or segment to another: revenue went up, margin % went down, conversion improved, retention slipped. The hard question is always the same — why? Answering it normally means building fragile Excel formulas by hand, and percentage / rate metrics are especially difficult because percentages do not add up linearly.
WaterfallBridge automates this. You give it a baseline scenario and a comparison scenario, and it produces a bridge that attributes the total change to each input variable, broken down by the dimension you choose (product, region, channel, segment, campaign, or any other). The result is a clear, auditable, root-cause explanation.
Explain revenue, cost, margin % and GOP % variance for monthly, quarterly and budget-vs-actual reviews.
Produce defensible driver-based decompositions for client deliverables in minutes, not days.
Decompose conversion rate, ROI and CAC changes by campaign, channel and traffic source.
Break down yield, efficiency, occupancy and throughput movements by line, site or product.
Get a plain-language root cause for any KPI move without waiting for an analyst.
WaterfallBridge handles both additive value metrics and non-additive rate metrics:
| Capability | WaterfallBridge | Excel (manual) | Power BI | Waterfall chart | Traditional variance analysis |
|---|---|---|---|---|---|
| Automatically computes driver attribution | Yes | Manual formulas | Visualization only | No — display only | Manual |
| Decomposes percentage / rate metrics correctly | Yes | Very hard | No | No | Error-prone |
| Breaks down by any dimension & key | Yes | Manual pivots | Partial | No | Manual |
| Handles added / removed data | Yes | Manual | No | No | Often ignored |
| Auditable, deterministic, closed-form math | Yes | Depends | N/A | N/A | Depends |
| Callable by AI agents (MCP) | Yes | No | No | No | No |
In short: a waterfall chart shows a decomposition; WaterfallBridge computes it — including for rate metrics — and can be driven by an AI agent.
A waterfall chart only visualizes numbers you have already computed. WaterfallBridge computes the decomposition for you: it takes a baseline and a comparison scenario, attributes the KPI change to each input variable and dimension, and then produces the bridge. It also handles percentage and rate metrics, which a normal waterfall chart cannot break down correctly because percentages do not add up linearly.
Yes. WaterfallBridge is specifically built to decompose percentage and rate metrics such as margin %, gross operating profit %, conversion rate, retention rate and yield. It separates the effect of the numerator, the denominator, the mix between dimensions, and added or removed data, so you can see why a rate moved and not just that it moved.
Yes. Margin %, conversion rate and retention rate are exactly the kind of rate metrics WaterfallBridge specializes in. You define the numerator (SumY) and denominator (SumN), choose the dimension to break it down by, and WaterfallBridge attributes the rate change to each contributor.
Yes. WaterfallBridge exposes an MCP (Model Context Protocol) server, so AI agents in Claude, Cursor, Codex and ChatGPT-style tools can call it directly. You can simply say "Analyze this file and explain why Q4 margin dropped versus Q1" and the agent uploads the file, runs the bridge and explains the drivers. See the AI agents / MCP guide.
No user data is stored on WaterfallBridge servers. Your spreadsheet is processed locally and only lightweight algorithm requests reach the server to run the bridge math. The Excel add-in processes data 100% locally on your device. Your AI provider key, if you use the in-app Ask AI panel, stays encrypted on your machine.
Try the demo with sample data, or ask an AI agent to analyze your own file.
Try Demo with sample data → Use it with AI agents →