Scatter Graph Method in Cost Accounting — Mixed Cost Estimation Made Simple

The scatter graph method (also called the visual fit method) is a cost accounting technique for splitting a mixed cost into fixed and variable components. You plot activity level (units produced, machine hours, etc.) on the X axis and total cost on the Y axis. The line of best fit reveals both numbers at once: the slope is your variable cost per unit, and the Y-intercept is your total fixed cost.

Traditionally, accountants drew this line by eye on graph paper. The free scatter plot tool calculates the regression line mathematically — more precise than visual estimation and faster than the high-low method. Here is how to use it for mixed cost analysis.

The Scatter Graph Method in Cost Accounting

A mixed cost has both fixed and variable elements. For example, a factory's monthly utility bill includes a fixed connection fee (same every month) plus a variable usage charge (scales with production). To forecast or control this cost, you need both numbers separated.

The scatter graph method works because of the cost equation:

Total Cost = Fixed Cost + (Variable Cost per Unit × Activity Level)

Or in math notation: y = mx + b, where:

• y = total cost (what you pay)
• x = activity level (units produced, hours used)
• m = slope = variable cost per unit
• b = Y-intercept = total fixed cost

If you plot historical data and fit a line, the line's slope and intercept give you both numbers directly. No two-point calculation, no averaging. The scatter plot tool shows both in the stats bar below the chart.

Worked Example: Splitting a Factory Utility Bill

Scenario: A manufacturer wants to separate fixed and variable components of monthly electricity costs based on units produced.

12 months of data (units produced, total electricity cost in dollars):

1000, 3200
1200, 3450
1500, 3800
1800, 4150
2000, 4400
2200, 4650
2500, 5000
2800, 5350
3000, 5600
3200, 5850
3500, 6200
3800, 6550

Paste into the scatter plot tool. The regression equation that comes back will look approximately like:

y = 1.195x + 2003

Interpretation:

• Variable cost per unit = $1.195. For every additional unit produced, electricity cost increases by about $1.20.
• Fixed monthly cost = $2,003. Even at zero production, the base electricity cost is about $2,000.

R-squared above 0.99 confirms the cost behavior is highly linear — the fixed plus variable model fits the data well.

Sell Custom Apparel — We Handle Printing & Free Shipping

Scatter Graph Method vs. High-Low Method

Cost accounting textbooks present three main methods for splitting mixed costs:

Method	How It Works	Strength	Weakness
High-Low	Uses only the highest and lowest activity months	Simple math	Ignores all other data; heavily influenced by extremes
Scatter Graph (visual)	Draw line of best fit by eye	Uses all data	Subjective — different people draw different lines
Least Squares Regression	Mathematical line of best fit	Most accurate; uses all data; reproducible	Requires calculation or tool

The free scatter plot tool uses least squares regression, so you get the accuracy of the third method with the speed of the first. No manual drawing, no two-point limitation.

High-low is fine for a quick estimate. Scatter graph with regression is the right choice when the analysis will inform budgeting or pricing decisions, because it uses all your data instead of discarding 10 out of 12 observations.

Using the Results for Budgeting and Forecasting

Once you have fixed cost (b) and variable cost per unit (m), you can forecast costs at any activity level:

• Next month forecast — if you plan to produce 4,000 units: Total cost = $2,003 + ($1.195 × 4,000) = $6,783.
• Budget planning — if sales projections range from 2,500 to 3,500 units, your electricity budget range is $4,991 to $6,186.
• Break-even analysis — combine fixed and variable costs with pricing to find break-even quantity.
• Make-or-buy decisions — compare variable cost per unit against third-party quotes to decide whether to outsource.

A warning: these forecasts assume costs continue behaving linearly at new activity levels. If you project 10,000 units when your data only goes up to 3,800, you are extrapolating. The linear relationship may break down at higher volumes due to overtime, capacity constraints, or volume discounts on inputs. Note this uncertainty in your analysis.

Frequently Asked Questions

What is the formula for the scatter graph method?

The cost equation is y = mx + b, where y = total cost, x = activity level, m = variable cost per unit, and b = fixed cost. The free tool calculates m and b automatically via linear regression on your data.

How many months of data do I need?

At least 12 months is common practice in cost accounting. Fewer data points (6-8 months) can work but are more susceptible to outlier months (holiday shutdowns, plant maintenance, one-time events).

What if R-squared is low?

Low R-squared (below 0.70) suggests the cost does not behave linearly with the activity driver you chose. Consider: is there a better cost driver? Are there step costs (fixed costs that change at certain activity thresholds)? Is the data contaminated by one-time events you should exclude?

Can I use this for multi-variable cost analysis?

The scatter plot tool handles single-variable (X and Y) analysis. For multi-variable cost analysis (where cost depends on multiple drivers), you need multiple regression in a statistical tool like Excel's data analysis toolpak, R, or Python.