I set up EBITDA tracking in SimpleKPI for a small business: the six metrics EBITDA is made from, a monthly budget, KPIs that compare the two, and a dashboard. The built-in AI did the building from a few sentences, and two years of monthly figures went in from a spreadsheet. The business is invented, a US commercial services company with a busy summer and a quiet winter. The dashboard is real.
The calculation turned out to be the easy part. The decisions around it are where EBITDA dashboards quietly go wrong, in any tool, spreadsheets included: how you compare with budget, which signs go where, and how a percentage rolls up from months to a year. Get the last one wrong and a year that beat its target can be reported as a miss.
The short version: EBITDA itself is easy to calculate. Comparing it with a budget, and rolling it up from months to a quarter, is where small businesses get it wrong. This post walks through the setup and five mistakes worth avoiding, so you can skip them. If you only want the numbers, our free EBITDA calculator does all of it in your browser.
EBITDA in one paragraph
EBITDA is earnings before interest, taxes, depreciation and amortization. It takes net income and adds back the costs that say more about how a business is financed, taxed and equipped than about how well it runs. That makes it a fair way to compare one month with another, or your business with a competitor that borrowed more or bought its trucks instead of leasing them.
For a small business, three people care about it:
- Your lender. Many loan agreements set covenants on debt or interest cover measured against EBITDA. Finding out you breached one at year end is worse than watching it get closer each month.
- A buyer. Businesses with a management team are usually valued as a multiple of EBITDA. Smaller owner-run businesses are valued on seller's discretionary earnings, which starts from EBITDA too.
- You. It's the closest single number to "is the business itself making money", before the bank and the tax bill take their share.
EBITDA isn't defined under US GAAP, so there's no official version. That's fine, as long as you decide your definition once and stick to it every month.
The setup
You need five inputs and one calculation. I tracked them all monthly:
- Net income, the bottom line of the P&L
- Interest, which is better split into expense and income (more on that below)
- Income taxes
- Depreciation
- Amortization
EBITDA is a calculated KPI on top of them. Then, to see how the month went against plan, an EBITDA Budget KPI and a KPI comparing the two.
This is the part of my original prompt that set up the budget comparison:
Create an Adherence to EBITDA Budget calculated KPI, formatted as a percentage, to show the percentage variance between actual EBITDA and the budget, using current KPIs in the formula: (Net Profit + Interest + Taxes + Depreciation + Amortization) ÷ EBITDA Budget × 100.
The AI built exactly what I asked for. The trouble is in what I asked for.
Read that prompt again and you can probably spot the first mistake.
Mistake 1: calling attainment a variance
I asked for "the percentage variance" and then wrote the formula for something else. Actual ÷ budget × 100 is attainment: how much of the budget you reached. On budget is 100%. A variance is (actual − budget) ÷ budget × 100: how far you are from it. On budget is 0%.
Both are useful. The problem is the label. Call attainment a variance and anyone reading the dashboard sees 94% and thinks "94% off budget", and any red, amber and green thresholds get set the wrong way round. Pick one, name it for what it is, and set its colors to match. Or track both, named EBITDA vs Budget and EBITDA Variance, so nobody confuses them.
Mistake 2: dividing by a loss budget
Plenty of small businesses budget a loss in some months: a seasonal business in its quiet season, or a month with a big one-off cost. Say you budget an EBITDA loss of $20,000 and lose only $10,000. You beat the plan by half. Actual ÷ budget gives 50%, which reads as a miss. Lose $30,000 and it gives 150%, which reads as a win.
The fix is to measure variance against the budget's absolute value: (actual − budget) ÷ |budget| × 100. Beating a $20,000 loss budget by $10,000 is then +50%, and missing it by $10,000 is −50%. The calculator does this and warns you when the plain formula would point the wrong way.
On a dashboard, the simplest rule is this: if you budget a loss in any month, judge that month on the variance in dollars. A percentage of a negative number is hard to read whichever way you calculate it.
Mistake 3: averaging months into a quarter
This is the one that would have bitten hardest, because it only shows up when you change the date range.
A percentage KPI tracked monthly has to become one number when you look at a quarter or a year. There are three ways software can do that: add the monthly percentages, average them, or recalculate from the totals. Only the last one is right.
Here are six months for a seasonal business with a quiet winter. The numbers are invented, but the pattern is common:
| Month | EBITDA | Budget | vs budget |
|---|---|---|---|
| Jan | $4,000 | $2,000 | 200.0% |
| Feb | $6,500 | $5,000 | 130.0% |
| Mar | $21,000 | $24,000 | 87.5% |
| Apr | $19,500 | $25,000 | 78.0% |
| May | $23,000 | $26,000 | 88.5% |
| Jun | $24,000 | $26,000 | 92.3% |
| Total | $98,000 | $108,000 | 90.7% |
- Averaged, the six monthly percentages give 112.7%. The half-year beat its budget.
- Added up, they give 676.3%, which at least looks wrong.
- Recalculated from the totals, $98,000 ÷ $108,000 is 90.7%. The business missed its half-year budget by $10,000.
The average is the dangerous one, because it looks believable. The two quiet months beat small budgets by a wide margin, and an average gives them the same weight as the four busy months that all fell short.
You can see the difference on the dashboard we built. Over twelve months the business made $573,316 of EBITDA on $3,805,600 of revenue: a margin of 15.07%, just above its 15% target. Work the margin out month by month and average it, and the same year shows 14.85%, just below target. Same data, opposite verdict, because averaging lets the quiet winter months count as much as the busy summer.
That's why SimpleKPI lets you choose how each calculated KPI rolls up, with a Calculate by setting. Frequency works the formula out for each month and then rolls the results up. Total adds up each part of the formula for the whole period first, then does the division once. Frequency is the right choice for some KPIs, such as an average daily rate. For anything that's a ratio of totals (attainment, variance, margin), choose Total. For EBITDA itself it makes no difference, because it's all addition and subtraction.
Whatever you use, test it: set the date range to a quarter and check the percentage against total EBITDA ÷ total budget worked out by hand.
Mistake 4: getting the signs wrong
My original prompt had one box for interest and one for taxes. Most months that's fine. Then the business earns interest on a cash balance, or books a tax benefit, and someone types a positive number into a box that needed a negative one.
- Interest income is taken off, not added back. It wasn't earned by operating the business. A separate Interest Income metric makes this hard to get wrong, so split interest into two metrics.
- A tax benefit is negative. If the P&L shows an income tax benefit, adding it back as a positive number overstates EBITDA by twice the amount.
- Only income taxes count. Payroll taxes, sales tax and property tax are costs of running the business and stay in.
If your business is an S corp, a partnership or an LLC taxed as one, the income passes through to the owners. There's usually little or no income tax on the business's own P&L, so this line may be zero most months. The owners' personal taxes never go in.
Mistake 5: using tax paid instead of tax expense
This one is about which number you type in, not the formula. EBITDA adds back the income tax expense on the month's P&L. It's tempting to use what actually left the bank, because that's the number you know. But estimated tax payments go out four times a year and don't line up with the months the profit was earned. Use cash paid and EBITDA jumps in April, June, September and January for no operating reason.
The same goes for depreciation. Use the monthly charge from your books, not the cost of the equipment you bought that month.
The setup I'd use now
Putting the fixes together, this is the prompt I'd give SimpleKPI today. It's also on the calculator page with a copy button.
Create the monthly metrics Net Income, Interest Expense, Interest Income, Income Taxes, Depreciation and Amortization, in US dollars. Then create an EBITDA calculated KPI: Net Income + Interest Expense - Interest Income + Income Taxes + Depreciation + Amortization.
Create a monthly EBITDA Budget KPI in US dollars. Then create two calculated KPIs, and set Calculate by to Total on both, so a quarter or a year is worked out from the totals rather than by averaging the months:
- EBITDA vs Budget, as a percentage: EBITDA / EBITDA Budget x 100, where 100% means on budget.
- EBITDA Variance, as a percentage: (EBITDA - EBITDA Budget) / EBITDA Budget x 100, where 0% means on budget.
Finally, build an EBITDA dashboard showing the last six months: EBITDA against budget, the variance, and a trend for each of the metrics that make up EBITDA.
You can open this board live, with the full story behind its two years of numbers, on our EBITDA KPI dashboard example.
When EBITDA goes up and the bank balance goes down
EBITDA tells you whether the business is making money from operating. It doesn't tell you whether you have any. Here's another invented six months, for a business that's growing:
| Month | EBITDA | Change in cash | What happened |
|---|---|---|---|
| Jan | $18,000 | +$15,500 | A normal month |
| Feb | $19,500 | +$16,000 | A normal month |
| Mar | $21,000 | +$9,000 | The biggest customer moves to 60-day terms |
| Apr | $22,500 | −$31,000 | A $40,000 truck, bought outright |
| May | $24,000 | +$4,000 | Slow payers, plus an estimated tax payment |
| Jun | $26,000 | +$2,500 | Inventory built up for the busy season |
EBITDA rose every month, $131,000 in total. The bank balance went up by $16,000. Nothing in that table is a mistake. It's what growth does: customers owe you more, you buy equipment, you stock up. But a dashboard showing only EBITDA would have told the owner everything was fine right up to the month payroll got tight.
What to put next to EBITDA
EBITDA on its own can mislead. Put four other numbers next to it:
- EBITDA margin, EBITDA ÷ revenue. Revenue growth can hide a fall in profitability, and margin shows it.
- Cash from operations, for the reasons above. Divide it by EBITDA to get cash conversion. When that starts falling, look at why before it shows up in the bank.
- Days sales outstanding, how long customers take to pay. It explains most of the gap between EBITDA and cash in a growing service business.
- And one leading measure. Sales booked or proposals sent both work. EBITDA only gets confirmed weeks after the month closes, and this moves first.
I’d first look at monthly EBITDA against budget, alongside the EBITDA margin. That tells me whether we delivered the operating profit we expected and whether profitability is keeping pace with revenue. If there’s a gap, I’d break it down into revenue, gross margin and operating costs to understand what caused it.
For more financial KPIs and how they fit on one screen, see our financial KPI dashboard example.
Track the parts, not just the total. Record net income, interest expense, interest income, income taxes, depreciation and amortization each month, and calculate EBITDA from them. Add a monthly EBITDA budget, then a variance KPI that compares the two. Keep the definition the same every month, and make sure any percentage is recalculated from the totals when you look at a quarter or a year.
Variance in dollars is actual EBITDA minus budgeted EBITDA. As a percentage, divide that difference by the budget's absolute value and multiply by 100. A positive number means you beat budget. Using the absolute value keeps the sign right when the budget is a loss.
No. An average gives a small month the same weight as a big one, so a quarter can look ahead of budget when it's behind. In our six-month example the average of the monthly percentages was 112.7% while the true figure, total EBITDA divided by total budget, was 90.7%. Always recalculate a percentage from the period totals.
No. EBITDA adds back the income tax expense on the P&L for the month. Estimated tax payments go out four times a year and don't match the months the profit was earned, so using cash paid makes EBITDA jump around for no operating reason.
EBITDA margin, so growth in revenue doesn't hide a fall in profitability; cash from operations, because EBITDA isn't cash; and days sales outstanding, which shows when customers are paying more slowly. Add one leading measure, such as sales booked, which moves before EBITDA does.
Turn KPIs like these into KPI forecasts and anomaly detection, and pay one flat price: $79 a month for unlimited users and KPIs. There is a 14-day free trial, no card needed.
