
Gross Profit Margin
The gross profit margin shows what percentage of revenue a company keeps after subtracting the direct costs of the product sold. It is one of the most important metrics for comparing software, hardware, and AI business models.
A company earns money through sales. This sum is called revenue. However, in order to deliver the goods or services at all, direct costs arise: materials, components, manufacturing, and for digital services also the computing power in the data center. If you subtract these direct costs from revenue, what remains is the gross profit. The gross profit margin indicates how large this remainder is in relation to revenue, expressed as a percentage. A company with 100 million euros in revenue and 30 million euros in direct costs therefore has a gross profit margin of 70 percent.
What the margin reveals about a business model
Revenue alone says little about whether a business is doing well. Two companies can both earn a billion euros and yet be in completely different positions. For one, 800 million remains after direct costs; for the other, only 150 million. The first can put a lot of money into research, advertising, or new products. The second has to stretch every euro three times.
Typical ranges help with classification. Classic software companies often achieve 75 to 85 percent, because an additional copy of a program costs almost nothing. A car manufacturer tends to be around 15 to 25 percent, because every car consists of expensive sheet metal, electronics, and labor time. Supermarkets are even lower. Comparing margins is therefore only truly meaningful within the same industry.
It is also interesting whether the margin rises or falls over the years. A falling margin can mean that competitors are pushing down prices or that purchase prices are rising. At companies working with artificial intelligence, this figure is watched especially closely, because operating the models continuously consumes electricity and expensive chips.
The calculation behind the percentage figure
The formula is simple: gross profit divided by revenue, times 100. Gross profit in turn is revenue minus the direct costs of what was sold. These direct costs are often called cost of goods sold in financial reports, or Cost of Revenue in English. This includes everything directly related to the service delivered.
What matters is what is not included. Salaries of the development department, marketing budgets, office rents, and taxes are left out. These items are only deducted later and then lead to other metrics, such as net margin. The gross profit margin thus describes only the first step of the calculation, not the actual profit at the end.
A common mistake is therefore to equate a high gross profit margin with profitability. A start-up can have an 80 percent gross margin and still post deeply negative results, because it spends enormous sums on personnel and advertising. The margin shows the potential of the business model, not the outcome.
Where the figure appears in quarterly reports
Publicly traded companies release figures every three months. The gross profit margin almost always appears near the top and is immediately compared by journalists and analysts to the previous year. If it deviates by even a few percentage points, the stock price can react significantly. In reports you then read sentences like: The gross margin fell from 74 to 71 percent.
This metric is currently especially prominent among AI providers. Anyone operating chatbots or image generators pays for computing time for every single request. These costs push the margin below the level of classic software. That is precisely why providers are working to make their models run more efficiently. Every computing step saved directly improves the margin.
You also encounter this principle in everyday life, without anyone using the term. A drink at the cinema costs a multiple of its purchase price, while a TV at the electronics store costs barely more than its procurement cost. Both stores live off different margins and must build their business accordingly differently.