WaterfallBridge algorithm + AI assist — one sentence to insight.

Explain Any KPI Change —
Faster Analysis, Smarter Strategy

Break down value, % and rate KPIs — revenue, margin %, conversion, retention, yield — by product, region, channel, segment, campaign or any dimension, and reveal the root cause behind every move, across finance, sales, marketing, operations, and more.

Try Demo → See Use Cases → Watch Tutorial →
✨ AI assist available — one sentence, full analysis
✨ Two ways to use WaterfallBridge — pick the experience that fits you best.
🌐

Use Online — Web app

Smart-generate Key Findings, Insights, and analysis reports.

Use OnlineWeb app · No install required
  • No installation — works in any browser
  • Cross-platform: Windows, Mac, mobile
  • Demo available instantly — no account needed
  • No user data is stored
  • ✨ AI-assisted — just say it in one sentence, the AI Assistant runs the analysis for you
📊

Use with Excel

Install for Windows Excel. Faster for large datasets.

Download on Microsoft Store
  • Requires Windows + Microsoft Excel desktop
  • Fast local processing for large files
  • 🔒 100% local data processing — no data ever leaves your device
  • If Microsoft Store is disabled, download & install directly
🎬 See WaterfallBridge in action — a quick demo of the full workflow.
▶ WaterfallBridge Demo — YouTube
✨ Two ways to use WaterfallBridge — pick the experience that fits you best.
Use Online — Web app screenshot
Use Online — Web app screenshot
Use With Excel — Desktop screenshot
Use With Excel — Desktop screenshot
How It Works

What Makes WaterfallBridge Different

One philosophy — pinpoint the real drivers behind every KPI change — delivered through five concrete breakthroughs that generic AI tools and traditional variance/waterfall workflows simply cannot match. All built in by default. No scripting, no spreadsheet gymnastics, no black-box guessing.

The Black Box Generic AI

Ask a generic AI "why did profit drop 7%?" and you get a plausible-sounding narrative — impossible to audit, impossible to reconcile to the source numbers.

Generic AI
The Black Box
  • price data
  • cost data
  • volume
  • product mix?
  • Black Box
  • ≈ −7% probabilistic guess
Internal logic opaque — cannot reconcile, cannot audit. A best-guess narrative from a probabilistic model. Not reproducible. Not auditable. Cannot tie back to the source numbers.
WaterfallBridge
The Glass Box
  • price data
  • cost data
  • volume
  • product mix
  • Δ = ∑ wᵢ·Δxᵢ — closed-form decomposition
Price −3% + Mix −2% + Cost −2% = −7% ✓
Every driver derived from a closed-form formula. Auditable, reproducible, traceable to the last decimal point.

Beyond Absolute Values Percentages and Rates Too

Every traditional waterfall tool can only decompose absolute amounts — Revenue, Cost, Profit. WaterfallBridge handles percentage and rate KPIs with the same one-click workflow.

Traditional Tools
Values Only ($, €, ¥)

Profit Bridge ($)
$200 Start → $190 End
+$60 UnitPrice −$40 UnitCost −$30 SalesQty

Traditional tools can decompose dollar KPIs — but what about margin %, conversion rate, or retention?
WaterfallBridge
Values + Percentages + Rates

Gross Margin % Bridge
36.6% Start → 34.6% End
+8.4 pt UnitPrice −7.5 pt UnitCost −2.9 pt SalesQty

Margin %, Conversion Rate, Retention, Yield — all decomposed automatically, in points.
Core Breakthroughs

Five ways WaterfallBridge is different

One philosophy — pinpoint the real drivers behind every KPI change — delivered through five concrete breakthroughs that generic AI tools and traditional variance/waterfall workflows simply cannot match. All built in by default. No scripting, no spreadsheet gymnastics, no black-box guessing.

1

Reject Black-Box Decisions — A 100% Transparent White-Box Engine

Ask a generic AI "why did profit drop 7%?" and you get a plausible-sounding probabilistic narrative — impossible to audit, impossible to reconcile to the source numbers. WaterfallBridge is a 100% transparent white-box engine: every basis point of the change is decomposed by deterministic, closed-form math into named drivers that sum back exactly. In financial audit and core business attribution, "approximately right" is the same as wrong.

Generic AI
The Black Box
price data · cost data · volume · product mix?
Black Box → ≈ −7% probabilistic guess
Internal logic opaque — cannot reconcile, cannot audit. A best-guess narrative from a probabilistic model. Not reproducible. Not auditable. Cannot tie back to the source numbers.
WaterfallBridge
The Glass Box
Profit Bridge: −7% — decomposed exactly
price data · cost data · volume · product mix
Δ = ∑ wᵢ·Δxᵢ — closed-form decomposition
Price −3% + Mix −2% + Cost −2% = −7% ✓
2

Beyond Absolute Values — Percentages and Rates Too

Every traditional waterfall tool can only decompose absolute amounts — Revenue, Cost, Profit. WaterfallBridge handles percentage and rate KPIs with the same one-click workflow: Gross Margin %, Conversion Rate, Retention, Loss Ratio, SLA, Yield, and more.

Traditional Tools
Values Only ($, €, ¥)
Profit Bridge ($)
$200 Start → $190 End
+$60 UnitPrice −$40 UnitCost −$30 SalesQty
Traditional tools can decompose dollar KPIs — but what about margin %, conversion rate, or retention?
WaterfallBridge
Values + Percentages + Rates
Gross Margin % Bridge
36.6% Start → 34.6% End
+8.4 pt UnitPrice −7.5 pt UnitCost −2.9 pt SalesQty
Margin %, Conversion Rate, Retention, Yield — all decomposed automatically, in points.
3

Drivers That Add Up — Exactly

When analysts manually split a percentage change (e.g. Margin %, Conversion Rate), the driver contributions almost never sum back to the headline number — leaving an "unexplained gap" that managers have to handwave. WaterfallBridge uses a mathematically reconciled algorithm: every driver's contribution sums precisely to the total change, with no residual to explain away.

Manual Decomposition
Doesn't Reconcile
Margin %: 36.6% → 34.6% (Δ −2.0 pt)
36.6% Start +6.0 pt Price −5.0 pt Cost −2.0 pt Qty −1.0 pt ? 34.6% End
Drivers sum to −1.0 pt, but Margin % actually moved −2.0 pt. 1.0 pt unexplained — the analysis no longer ties.
WaterfallBridge
Reconciled to the Last Decimal
Margin %: 36.6% → 34.6% (Δ −2.0 pt)
36.6% Start +8.4 pt Price −7.5 pt Cost −2.9 pt Qty 34.6% End
+8.4 − 7.5 − 2.9 = −2.0 pt ✓
Drivers reconcile exactly to the headline change. Nothing to handwave.
4

Drill Any % Change Down to the SKU

Manual decomposition typically stops at one level — you see that Unit Cost hurt Margin %, but not which product, channel, or customer drove it. WaterfallBridge can drill any driver in a percentage or rate bridge down to any finer granularity, attributing the contribution to every single SKU automatically.

Manual / Traditional
Single-Level Decomposition
Margin % Bridge (top level only)
36.6% Start +8.4 pt Price −7.5 pt Cost −2.9 pt Qty 34.6% End
Stops here. Which SKU?
You learn the categories — but cannot tell which product, channel, or customer caused each driver.
WaterfallBridge
Drill Into Any Driver, Down to the SKU
Product / SKU Price (pt) Cost (pt) Qty (pt)
SkyView X1+1.4−1.1−1.4
SkyView Lite+1.4−1.1−0.5
Gaming Beast+0.8−0.6−0.1
UltraSound Pro+0.7−1.2+0.1
Grand Total+8.4−7.5−2.9
5

Automatic Formula Decomposition — Any KPI, Any Dimension

Just type the KPI formula (e.g. Revenue / Visits, Claims / Premium, GMV / Orders) and WaterfallBridge parses and decomposes it automatically. Switch dimensions — product, region, channel, customer, campaign — with a single click. Works for any industry, any KPI. What used to take an analyst hours — or, for complex KPIs with many dimensions, even several days — per scenario is now an instant pivot.

Manual / Spreadsheets
Rebuild Every Time
  1. Re-pivot the source data for the new dimension ~30 min
  2. Hand-derive the variance formula for the new KPI ~45 min
  3. Build the contribution table row by row ~60 min
  4. Reconcile against the headline — chase the residual ~30 min
  5. Redraw the bridge chart by hand ~20 min
≈ 3 hours per dimension, per KPI — and for complex KPIs across many dimensions, often several days. Every new KPI, every new dimension, every refresh — the analyst starts from scratch.
WaterfallBridge
Type the Formula, Click a Dimension
  1. Type the KPI formula (e.g. Margin = (Price − Cost) × Qty / Revenue) ~30 s
  2. Pick Baseline & Comparison periods ~10 s
  3. Choose a dimension — Product, Region, Channel, Customer… ~5 s
  4. Reconciled bridge + per-item contribution table render automatically instant
  5. Switch dimension to re-cut the analysis — one click instant
< 1 minute — works for any industry, any KPI. Finance, marketing, e-commerce, insurance, operations, SaaS — one tool, every KPI, every dimension.
✨ New · AI-Assisted Analysis

Just Say It in One Sentence — AI Runs the Analysis

Not an analyst? No time to learn the workflow? You no longer have to click through anything. Describe your question in plain language and an AI assistant prepares your data, runs the Bridge, and hands you the key drivers and insights — automatically.

Platform note: The AI Assistant is currently available for Windows. Support for macOS and other systems is coming soon.

AI is the operator, not the brain.

Every number you see is still produced by the deterministic WaterfallBridge engine — the same transparent, auditable, closed-form math described above. The AI never invents results or guesses; it simply drives the app for you and explains what the engine computed. The algorithm remains 100% WaterfallBridge.

"Analyze sales-2024.xlsx in my Downloads and tell me why profit dropped this quarter."
→
  1. Prepare — Reshapes your file into Bridge format
  2. Run — Builds the Bridge with the WaterfallBridge engine
  3. Read — Pulls the chart & contribution table
  4. Explain — Writes the drivers & insights in plain words

Bring your own AI access. WaterfallBridge does not resell or bundle AI usage — the engine itself is free of AI cost, and the language model is supplied by you.

⬇ Download AI Agent (Windows)

Required for both the "Ask AI" panel and Claude Code / Cursor / Codex. Free, Windows-only for now.

💬 1 · The "Ask AI" panel in the web app

Open WaterfallBridge in your browser and click Ask AI on the right. Chat in plain language and watch it prepare data, run the analysis, and summarize the results — no separate tool to install in front of you.

  • Built right into the app — nothing extra to open
  • Requires your own AI provider API key — any compatible provider, e.g. OpenAI / Anthropic (stays on your machine; you pay the provider for usage)
  • Great for non-technical users who just want answers
See the step-by-step setup →

⌨️ 2 · From Claude Code / Cursor / Codex

Already living in an AI coding tool? WaterfallBridge plugs in as an MCP server. Ask your assistant to analyze a file and it will operate WaterfallBridge for you — right where you already work.

  • Uses your own existing subscription/plan for that coding assistant.

Claude Code · Cursor · Codex · Claude Desktop

See the step-by-step setup →
Read the full AI Assistant guide →
Use Cases

Do you often encounter similar situations like this?

Typical Scenario
Boss:
How has our profit margin been doing over the past few quarters?
You:
Sales increased by 10%, but overall profit margin decreased by 7pt.
Boss:
That's not good. Can you give me a breakdown and analysis report of the 7pt profit margin decline? I need to know what specific products and reasons caused this decline. It would be best if you could do a comprehensive analysis by region, product type, channel, or team across different dimensions. You can compare with different time periods like last year, the year before, or our budget. This way we can identify problems promptly and solve them immediately.
You:
(thinking to yourself): If it were just revenue analysis, that would be simple enough. But breaking down the 7pt profit margin decline to specific 1000 SKUs and knowing which factors — price, discount rate, cost, etc. — contributed how many pts to the overall decline, that's not an easy task. If we add regional, product type, and channel dimensions, it becomes even more complex. Then comparing with last year, budget, and other data versions — it's almost impossible...
Do you frequently encounter similar needs in your work?
Do you spend a full day, or even days or longer (for multi-dimensional analysis) to reach an analytical conclusion? Does this analysis contain a lot of manual estimation, and the final aggregation of all impact factors doesn't accurately equal the 7pt profit margin decline, requiring manual adjustments? Do you always harbor doubts about whether the calculation process is 100% correct?

Please use WaterfallBridge to complete similar analysis in just minutes, getting accurate analytical results through algorithms and decomposing revenue, profit margin changes into the impact factors, different dimensions, and scenario comparisons you desire. This helps you pinpoint problems precisely, take timely action, and improve enterprise competitiveness.

At the same time, WaterfallBridge can also calculate contribution at the lowest level of detail within each factor — for example, at the SKU level — so users can trace not only which high-level driver changed the result, but also which underlying items contributed to that change.

Who Can Use WaterfallBridge

WaterfallBridge is not just for finance. Any team that needs to explain why a KPI changed — especially percentage and rate metrics — can use it.

Industry / Domain Typical Roles Common Analysis Questions Core Factors WaterfallBridge Decomposes
Finance / FP&A FP&A Manager, Finance Analyst, CFO Office Why did Revenue / Profit / Margin % change vs budget or last year? Price, Volume, Mix, Cost, FX, Structure
Banking / Financial Institutions FP&A, Treasury, ALM, Finance Manager Why did NIM or NII change? Which product, deposit type, or branch drove the shift? Rate, Volume, Mix, Balance Sheet Structure, Product, Maturity, Currency, Branch
Insurance Finance Analyst, Underwriting Analyst, Actuarial Analyst Why did combined ratio, loss ratio, or underwriting margin change? Loss Rate, Expense Rate, Premium Volume, Product Mix, Region, Channel
Marketing / Advertising Performance Marketing Analyst, Growth Analyst Why did CTR, CVR, ROAS, or CAC change? Which campaign or keyword drove the shift? Traffic, Channel Mix, Device Mix, Geography, Campaign, Keyword, Offer
E-commerce / Retail E-commerce Analyst, Category Manager, Commercial Analyst Why did sales, gross margin %, discount rate, or conversion change? Price, Units, Discount, Product Mix, Channel Mix, Freight, Cost
Procurement / Supply Chain Procurement Analyst, Supply Chain Analyst, Sourcing Manager Why did spend or unit purchase cost rise? Which supplier / category / plant drove it? Unit Price, Volume, Supplier Mix, Category Mix, FX, Logistics
Manufacturing / Operations Operations Analyst, Plant Controller, Manufacturing Finance Why did unit cost, yield, scrap rate, or gross margin change? Volume, Yield, Scrap, Labor, Overhead, Material Cost, Product Mix
SaaS / Subscriptions RevOps, BizOps, Growth Ops, Customer Success Ops Why did trial-to-paid, retention, churn, ARPU, or NRR change? Customer Mix, Plan Mix, Price, Volume, Cohort, Region, Channel
Sales / Commercial Ops Sales Ops, Business Analyst, Commercial Excellence Why did win rate, average selling price, customer margin, or regional sales shift? Price, Volume, Customer Mix, Product Mix, Region, Sales Team
Product / Data Analytics Product Analyst, BI Analyst, Data Analyst Why did engagement, activation, conversion, or retention change? User Mix, Channel Mix, Version Mix, Region, Device, Feature Adoption
Healthcare / Education / Public Sector Operations Analyst, Planning Analyst, Program Manager Why did operational or performance KPIs shift? Volume, Mix, Resource Allocation, Region, Service Type
Gaming / Mobile Apps Game Analyst, Live Ops, Growth Manager, Monetization Analyst Why did D1 / D7 retention, payer conversion, ARPPU, or engagement rate change? User Mix, Version Mix, Channel Mix, Country, Device, Cohort, Content / Event Mix
Call Center / Contact Center Call Center Manager, WFM Analyst, Service Ops Analyst Why did abandonment rate, service level, handle time, or first-contact resolution change? Call Volume, Queue Mix, Team Mix, Shift, Language, Region, Issue Type, Staffing
DTC / Small Business / Creator Commerce Shopify Store Owner, Creator Manager, Growth Operator Why did store conversion, refund rate, AOV, or product margin change? Traffic Mix, Product Mix, Price, Discount, Channel, Creator / Campaign, Geography
📚 Video Tutorial

Step-by-step: How to Use WaterfallBridge

Follow along with our detailed walkthrough and master key driver analysis from day one.

🌐 Use Online

No installation required. Upload your file in the browser and analyze instantly. Works on Windows, Mac, Linux, and mobile.

📊 Use with Excel

Install the desktop add-in for Windows. Requires Windows + Excel 2016 or higher. 100% local data processing — no data ever leaves your device.

Initial Panel Operations

The initial panel displays two buttons: "New Bridge File" and "Open Existing Bridge File"

1.1 Click "New Bridge File" button — Generates a new Excel file containing a new "Data" worksheet. Please enter the corresponding data in the "Data" worksheet for further analysis.

1.2 Click "Open Existing Bridge File" button — Select a previously used Excel file. This file should contain the relevant data and have Data Types configured.

Data Worksheet

In the Data worksheet starting from column B, enter data types in the first row, field names in the second row, and data in all rows from the third row onwards.

2.1 Data Type Selection — In the first row, use the dropdown menu to select data types: Dimension, Key, SumY, SumN, Result. Each field must be defined with one data type.

2.1.1 Dimension — Represents data dimensions such as date, region, and other facets. These fields do not participate in calculations and are only used to filter data.

2.1.2 Key — Key is a type of Dimension, and only one field can be marked as Key. The field marked as Key will serve as the minimum granularity for analysis. You can change other dimensions to Key fields for analysis at different granularities.

2.1.3 SumY — Data that can be added together and the sum is meaningful (e.g., Quantity of different products).

2.1.4 SumN — Data that cannot be added together and the sum would be meaningless (e.g., Unit Price of different products).

2.1.5 Result — Only one field can be selected as Result. The Result column must contain formulas that calculate the Result through other variables. Formulas can only use Excel's basic arithmetic operators: +, −, *, /. Result formulas should contain at least one SumY field.

2.2 Field Name Setup — Enter field names in the second row starting from column B. Field names cannot be duplicated.

2.3 Data Input — Enter specific data in the Data worksheet starting from the third row in column B.

Bridge Button Operations

After clicking "New Bridge File" or "Open Existing Bridge File", blue buttons "New Bridge" and "Refresh Bridge" will appear on the panel.

3.1 New Bridge — Generates new dropdown menus and Bridge worksheets and other analysis frameworks based on the Data worksheet data. (Note: Related worksheets such as Bridge worksheets and corresponding data in the original workbook will be deleted.)

3.2 Refresh Bridge — After selecting options in the Baseline and Comparison dropdown menus, click to generate the required Bridge charts and Contribution Analysis.

Dropdown Menu

The dropdown menu is divided into two areas: Baseline and Comparison. Users can filter which options serve as Baseline data and which serve as Comparison data.

4.1 Baseline dropdown — Select the baseline data for comparison analysis. For example, select 2023.

4.2 Comparison dropdown — Select the Comparison data. For example, select 2024 as the year of interest to compare with 2023.

Bridge Worksheet

The Bridge worksheet displays the final analysis results, including Bridge charts and Contribution Analysis tables.

5.1 Bridge Data Source — Starting from cell B3 in the upper left corner, field names and data will appear. This is the data used to generate Bridge charts.

5.2 Bridge Chart — The blue bar on the left is Baseline, the middle shows the impact contributed by each variable, and the right bar is Comparison.

5.3 Contribution Analysis Table — Directly below the Bridge chart, a Contribution Analysis table will be displayed, showing the contribution of each Item in the Key field to overall changes under different factors.

Through the Bridge Chart and Contribution Analysis Table, you can quickly identify the main Items (SKUs) affecting changes and the factors causing changes (price? cost? or quantity mix?).

SumY and SumN Calculation Rules

Formulas in the Data worksheet follow these type inference rules:

Addition (Subtraction): SumY + SumY = SumY · SumN + SumN = SumN
Multiplication: SumY * SumY = SumY · SumY * SumN = SumY · SumN * SumN = SumN
Division: SumY / SumY = SumN · SumY / SumN = SumY · SumN / SumY = SumN · SumN / SumN = SumN

Formula Usage: Each formula can only use other data from the same row. Formulas can only use basic arithmetic operators. The program will automatically copy the formula from the fourth row to the entire column after removing fixed cell reference symbols "$".

Nested Formulas: If the Result column formula references other cells that also contain formulas, the analysis will merge all nested formulas into one formula for analysis. The program will automatically reorder variables and apply the distributive property for better data analysis.

Added Data and Removed Data

In real business scenarios, the items being compared between two periods are rarely identical. New products are launched, old products are discontinued, customers come and go, regions open and close. Bridge Analysis handles these structural changes through two dedicated concepts: Added Data and Removed Data.

7.1 Definition: Added Data — Records that exist in the Comparison period but did NOT exist in the Baseline period (e.g., newly launched product line). Removed Data — Records that existed in the Baseline period but NO LONGER exist in the Comparison period (e.g., discontinued product).

7.2 Why They Must Be Separated: Because there is no "before" value for Added Data and no "after" value for Removed Data, their impact cannot be attributed to any single factor. They must be isolated and reported on their own bars in the Bridge chart.

7.3 Typical Business Examples: NPI (New Product Introduction) and EOL (End of Life) product lifecycle · Retail / Store Network · Customer / Subscription business · Geographic expansion · M&A Activity · Headcount analysis.

7.4 How to Mark: No manual tagging required — simply ensure your Key columns (such as Product Code) are consistent across both periods, and Bridge Analysis will handle NPI / EOL and similar structural changes correctly.

Data Security & Transparent Analysis

Data Security:
Desktop Version (Use with Excel): All data exists only in local Excel files. The program runs entirely on your machine with no network communication during the analysis process.
Online Version (Use Online): Your spreadsheet data stays in your browser. Only lightweight algorithm API requests are made to perform the Bridge calculation. No user data is stored on our servers.

Transparent & Explainable Analysis: Reject the ambiguous conclusions and inexplicability of AI analysis. The analysis calculation process is completely transparent and explainable.

Process Worksheet Transparency: Contains the logic and algorithms for generating Bridge charts and data. Every calculation step can be traced and verified.