Driver decomposition of change (LMDI)
Decomposes the change between two periods of a metric that equals a product of factors (revenue = traffic × conversion × units per receipt × price) into the contributions of each factor. The logarithmic mean method gives contributions that add up to the change with no remainder and does not depend on the order of factors. It shows the contribution in money and in shares of the change and ranks the factors.
At a glance
- What question does it answer?
- Which factor added or took how much from a metric between periods.
- What data is needed?
- Factor values in two periods.
- How to read the result?
- Start with the factor of the largest contribution in modulus. Example on the default values: The metric in the base period: RUB 18 million; The metric in the current period: RUB 19.7 million; Change: +RUB 1.69 million; Change in percent: +9.4%. The figures are illustrative: enter your own to see your result.
Why and how to reason
Revenue rose from 18.0 m RUB to 19.69 m RUB, +1.69 m RUB. The decomposition shows: traffic −0.97 m RUB, conversion +2.63 m RUB, units per receipt −1.08 m RUB, price +1.10 m RUB. Growth came from conversion and price, not from the flow of people.
When to use it
When explaining plan deviations for any product metrics.
How to read the result
Start with the factor of the largest modulus; the sum of contributions equals the change.
Common mistakes
Decomposing sums as products, zero values, reading a contribution as a cause.
What data is needed?
- Factor values in two periods.
Inputs
| Field | Unit | Value in the example | Meaning |
|---|---|---|---|
| The metric is in | money | Choose “money” if the product of factors gives rubles (revenue, profit), otherwise the contributions are shown as numbers. |
How it is calculated
| Step | Formula |
|---|---|
| The metric | Y = the product of the factors in each period |
| Logarithmic mean | Lw = (Y₁ − Y₀) ÷ (ln Y₁ − ln Y₀) |
| Factor contribution | Lw × ln(x₁ ÷ x₀) |
| Check | the sum of contributions equals Y₁ − Y₀ with no remainder |
Worked numbers
Default values from the app; in the interactive version you enter your own.
- The metric in the base period
- 18 m ₽
- The metric in the current period
- 19.7 m ₽
- Change
- +1.69 m ₽
- Change in percent
- +9.4 %
- The main contribution
- +2.63 m ₽Conversion, share
- Sum of contributions minus change
- 0 ₽It must be zero: the method has no remainder.
Data as a table
| Measure | Contribution |
|---|---|
| Traffic | −0.97 |
| Conversion, share | 2.63 |
| Units per receipt (UPT) | −1.08 |
| Unit price, RUB | 1.1 |
Contributions of factors
| Factor | Base | Current | Change, % | Contribution | Share of the change |
|---|---|---|---|---|---|
| Conversion, share | 0.04 | 0.046 | 15 % | 2.63 m ₽ | 156 % |
| Unit price, RUB | 2,500 | 2,650 | 6 % | 1.1 m ₽ | 65 % |
| Units per receipt (UPT) | 1.8 | 1.7 | −5.6 % | −1.08 m ₽ | −64 % |
| Traffic | 100,000 | 95,000 | −5 % | −965,896 ₽ | −57 % |
What to do with the result: explain a plan deviation by contributions: start with the factor of the largest modulus and ask whether it is controllable; for a sum (not a product) use the “Profit bridge” calculation.
A typical mistake: decomposing sums as products and entering zero values. The method is universal for any metrics of the form “factor × factor”; the result does not depend on the order of factors but depends on which factors you chose.
Nuances and limits
- Decomposing sums.
- Entering zeros.
- Reading a contribution as a cause.