Want to see how SaaS sales training can help teams simplify offers without sounding pushy?
Introduction to SaaS Gross Margin
SaaS businesses have traditionally benefited from an attractive economic model. Build the software once, sell subscriptions repeatedly and serve additional customers without the cost of delivery rising at the same rate as revenue. That model helped make SaaS gross margin one of the most closely watched measures of software profitability.
Artificial intelligence is beginning to change the equation. AI-powered products can require substantial computing power every time a customer asks a question, generates content, analyses information or asks an AI agent to complete a task. Instead of the cost being largely fixed, part of the cost can increase as customers use the product more.
That does not mean AI will automatically make SaaS companies less profitable. Falling inference costs, better models and more efficient infrastructure could improve the economics considerably. But it does mean software leaders may need to think differently about pricing, product design and the relationship between customer usage and the cost of providing the service.
What is SaaS gross margin?
SaaS gross margin measures how much revenue remains after subtracting the direct costs associated with delivering the software service. It is normally expressed as a percentage of revenue.
If a SaaS company generates £1 million in revenue and incurs £200,000 of direct costs to provide its service, its gross profit is £800,000. Its gross margin is therefore 80%.
Those direct costs can include cloud hosting, infrastructure, third-party software directly required to deliver the product and some customer support costs. Exactly what is included can vary between companies, which is why comparisons need to be made carefully.
The attraction of traditional SaaS was that serving the next customer could be relatively inexpensive. Once the product and infrastructure existed, revenue could grow much faster than the direct cost of providing the service. This is one reason businesses track a wider range of SaaS metrics alongside gross margin when assessing growth and financial performance.
Why has SaaS traditionally produced high gross margins?
Traditional software has very different economics from businesses selling physical products. A manufacturer must usually buy more materials to produce more units. A retailer must purchase more stock. A service company may need more employees as the number of customers increases.
A SaaS company can potentially add thousands of customers without increasing its delivery costs proportionately. Hosting and support costs increase, but the incremental cost of providing another software subscription can remain comparatively low.
Historic SaaS benchmarks illustrate why investors became accustomed to strong margins. OpenView reported gross margins around 70% to 80% across several SaaS revenue bands in its 2022 benchmark data.
This operating model helped businesses reinvest substantial amounts in product development, marketing and growth. Strong SaaS gross margin also gave successful companies room to absorb customer acquisition costs while building recurring revenue.
For businesses providing Sales Training for SaaS Companies, understanding these commercial economics matters because sales teams increasingly need to communicate value rather than simply sell access to software.
Why could AI put SaaS gross margin under pressure?
The challenge is simple. AI can introduce a meaningful variable cost into a business model that historically benefited from very low marginal delivery costs.
Every AI interaction may require inference. Models must process the customer’s input, perform computational work and generate an output. More complicated requests can require considerably more processing than simple ones.
An AI feature used occasionally may have little effect on SaaS gross margin. But the economics can look very different when thousands of customers use AI features repeatedly throughout the working day.
The problem becomes more significant with AI agents. A customer may see one completed task, while behind the scenes the system makes numerous model calls, searches databases, retrieves information, uses external tools and checks its own work.
Revenue might remain based on a predictable monthly subscription while the company’s underlying cost varies according to customer behaviour. That mismatch creates a margin risk that did not exist to the same extent in conventional SaaS.
This is also changing commercial conversations. Effective SaaS Sales Training Courses increasingly need to help salespeople explain the business outcome created by technology rather than relying on feature comparisons.
What costs does AI add to a SaaS business?
AI costs go beyond paying for access to a large language model. The complete cost depends on how the product has been designed and what customers ask it to do.
Inference is an obvious component. Companies using third-party models may pay according to the amount of input and output processed. Businesses operating their own models still face infrastructure, hardware and energy costs.
There can also be costs associated with vector databases, retrieval systems, data processing, storage, external APIs and specialist infrastructure. More advanced AI agents may interact with several services during a single workflow.
Human involvement can matter too. Some AI applications require people to review outputs, manage exceptions or maintain quality. If those activities are necessary for delivering the service, they can affect the economics behind SaaS gross margin.
The important number is therefore not simply the advertised price of a model. SaaS companies need to understand the complete cost of delivering the customer outcome.
Does more AI usage always mean lower margins?
No. Higher AI usage only becomes a problem when the additional value and revenue generated do not adequately compensate for the additional cost.
Imagine a SaaS company charging £100 per month. If normal software delivery costs £15 per customer, the economics look attractive. If intensive AI usage adds another £25 of direct cost, the SaaS gross margin changes substantially.
But the company may also be able to charge £150 because its AI functionality saves the customer hours of work or produces a measurable commercial result. In that situation, AI could create enough additional value to support higher pricing.
This is why pricing and positioning are becoming closely connected to AI economics. A skilled SaaS Sales Trainer may increasingly focus on helping teams establish the financial value of an outcome rather than presenting AI as another product feature.
The real question is not whether AI costs money. It is whether the company captures enough of the value AI creates to maintain healthy unit economics.
Could falling AI costs solve the problem?
Potentially, but SaaS leaders should be careful about assuming that cheaper models will automatically restore traditional software economics.
The cost of performing comparable AI tasks has been falling as models, chips and infrastructure become more efficient. Companies can also route simpler requests to smaller models, cache common responses and optimise prompts to reduce unnecessary processing.
These improvements could strengthen SaaS gross margin. A feature costing £1 to deliver today may cost considerably less in the future if the same result can be produced using cheaper and more efficient technology.
But falling unit costs can be accompanied by rapidly increasing consumption. When AI becomes cheaper and more capable, companies tend to give it more work. Agents may perform longer tasks, process larger amounts of information and make multiple model calls before producing an answer.
So the important measure is not simply cost per token. Businesses need to understand total cost per useful customer outcome.
This commercial shift can also affect B2B SaaS Sales Training, because customers may increasingly want to understand both the business result and how pricing changes as usage grows.
Why might SaaS pricing have to change?
The traditional per-user subscription works particularly well when the cost of serving an additional user is small and predictable. AI makes that assumption less reliable.
Two customers paying the same subscription could create dramatically different costs. One might use an AI assistant several times each month. Another might run thousands of queries or automate large workflows every day.
Unlimited AI inside a fixed subscription can therefore expose the provider to unpredictable costs and weaken SaaS gross margin.
This explains the growing interest in usage-based pricing, credits, tiered allowances, workflow pricing and outcome-based models. The wider choice between different SaaS pricing models is becoming more important as providers try to align what customers pay with usage, value and the cost of delivering AI functionality.
However, complicated pricing introduces another problem. Customers want certainty. If buyers cannot understand what the product will eventually cost, uncertainty can slow the buying decision.
That creates an important challenge for Corporate Sales Training for SaaS Companies. Sales teams need to explain variable pricing simply enough for buyers to understand the commercial risk as well as the potential return.
What should SaaS companies measure?
Revenue growth alone will not reveal whether an AI product has attractive economics. A company can grow rapidly while each additional customer creates significant additional delivery costs.
Businesses therefore need visibility into the cost of individual features, customers and workflows. That means understanding which models are being used, how many calls are required, how much data is processed and what third-party services are involved.
SaaS gross margin should also be considered alongside customer lifetime value, retention, expansion revenue and acquisition cost. Net Revenue Retention is particularly useful because it shows whether recurring revenue from an existing customer base is growing or shrinking after expansion, contraction and churn are taken into account.
A lower gross margin may still support a strong business if customers stay longer, spend more and receive substantially greater value. The cost of winning those customers matters too, making SaaS Customer Acquisition Cost another important measure when assessing whether growth is economically sustainable.
Product-level measurement is particularly useful. One AI feature may have excellent economics while another is heavily used but expensive to provide. Looking only at company-wide averages can hide that difference.
The objective is not simply to minimise AI usage. It is to remove low-value consumption while supporting the activities customers are willing to pay for.
How can SaaS businesses protect gross margin?
The first step is understanding where the money goes. AI expenditure should not disappear inside a general cloud bill. Teams need enough visibility to connect usage with individual products, customers and workflows.
Model selection can then make a major difference. Not every task needs the most capable or expensive model. Simpler requests can often be handled by smaller models, while more powerful models are reserved for work where their additional capability genuinely matters.
Caching, prompt optimisation, batching and better workflow design can reduce unnecessary processing. Companies can also introduce sensible usage limits or credits where unrestricted consumption would damage SaaS gross margin.
Pricing needs to evolve alongside those technical improvements. If a customer receives substantially more value because AI performs work previously completed manually, the commercial model should recognise that value.
And the sales team needs to understand the reasoning behind the pricing. In-House SaaS Sales Training can help teams move conversations away from comparing monthly licence prices and towards the operational or financial value the software creates.
Could AI actually improve SaaS profitability?
Yes. Focusing only on inference costs misses the other side of the equation.
AI can reduce the cost of developing software, supporting customers, producing documentation, testing code and operating internal processes. It may allow a SaaS company to serve substantially more customers without increasing headcount at the same rate.
AI could also create products that customers value more highly. Software that once helped someone complete a task might increasingly perform much of that task for them. If the economic value increases faster than the delivery cost, profitability can improve even if SaaS gross margin is lower than historical software benchmarks.
This distinction matters. A company with an 80% gross margin is not automatically a stronger business than one operating at 65%. Growth, retention, operating costs, capital requirements and customer value all affect the complete picture.
The strongest AI-enabled SaaS companies may therefore look different from the software businesses investors became accustomed to during the previous decade.
What does SaaS gross margin mean for sales teams?
Salespeople do not need to become accountants, but they do need to understand how the commercial model works.
If AI changes the economics of the product, sales teams may be asked to sell consumption plans, credits, workflows or outcomes instead of straightforward licences. Buyers will naturally ask how costs could change as usage increases.
A vague answer creates uncertainty. And uncertainty gives buyers a reason to delay. Understanding the SaaS Sales Cycle can help businesses identify where pricing, procurement, technical evaluation or commercial uncertainty is causing deals to slow down.
Salespeople therefore need to explain what customers are paying for, what drives usage and what business result that expenditure is designed to create. Good Sales Training for SaaS Teams should help commercial teams translate technical capability into clear business value.
That becomes particularly important when competitors appear cheaper. A lower subscription price means little if the products deliver different outcomes or use completely different pricing models.
Understanding SaaS gross margin therefore matters beyond the finance department. Product, finance and sales increasingly need a shared understanding of how AI creates value and what it costs to deliver.
Is AI making software less profitable?
It can, but that is not inevitable.
AI is introducing costs that traditional SaaS companies did not face at the same scale. Inference, compute, external models and increasingly complex agentic workflows can make the cost of serving customers more variable.
That could reduce SaaS gross margin, particularly where companies include expensive AI functionality inside fixed-price subscriptions without understanding how customers will use it.
But the technology is also becoming more efficient. Model costs can fall, workflows can be optimised and AI can create substantially more value for customers. Companies also have opportunities to change their pricing so revenue scales more closely with usage or outcomes.
Those economics are affected by what happens after the sale as well. Effective SaaS Customer Onboarding can help customers reach useful outcomes sooner, understand the product and use valuable features effectively rather than consuming resources without achieving the intended result.
The winners are unlikely to be the businesses that simply add AI to every feature. They will be the companies that understand which AI capabilities customers genuinely value, what those capabilities cost to provide and how to price them sustainably.
The old SaaS assumption was that software could scale while the marginal cost remained extremely low. AI complicates that assumption. The new challenge is to make sure the value created grows faster than the cost of delivering it.
Frequently asked questions about SaaS gross margin
What is a good SaaS gross margin?
A good SaaS gross margin depends on the type of software, business model, stage of growth and direct costs required to provide the service. Traditional SaaS businesses have often operated with gross margins around 70% to 80% or higher, although there is no single percentage that defines a successful SaaS company. AI-heavy products may operate at lower margins because inference, computing and other variable costs increase as customers use the product. Companies should therefore consider the level and direction of gross margin alongside growth, retention and overall unit economics.
How do you calculate SaaS gross margin?
SaaS gross margin is generally calculated by subtracting the direct cost of delivering the service from revenue, dividing the resulting gross profit by revenue and multiplying by 100. For example, if a SaaS company generates £1 million in revenue and incurs £250,000 in direct delivery costs, gross profit is £750,000 and SaaS gross margin is 75%. Companies should apply a consistent definition of direct costs when comparing performance over time.
Why does AI affect SaaS gross margin?
AI can affect SaaS gross margin because customer activity may create additional inference, computing, storage, data processing and third-party model costs. Traditional SaaS often has a relatively low incremental cost when another customer uses the product. Generative AI and AI agents can create greater variable costs because every query or automated workflow may require additional computing resources. The effect on margin depends on how heavily customers use AI and whether pricing captures enough of the value being created.
What is included in SaaS cost of goods sold?
SaaS cost of goods sold commonly includes the direct costs required to deliver the software service, such as cloud hosting, infrastructure and certain third-party services. Depending on the company’s accounting approach, relevant customer support or service delivery costs may also be included. AI SaaS businesses can have additional direct costs from model inference, GPUs, data processing, vector databases, external APIs and human review required to deliver customer outcomes.
Can AI SaaS companies achieve high gross margins?
Yes. AI does not prevent a SaaS company from achieving a strong gross margin. Businesses can improve SaaS gross margin through cheaper inference, efficient model routing, caching, prompt optimisation, better infrastructure and careful control of expensive workflows. Pricing also matters. If AI creates significant customer value, providers may be able to use consumption, tiered or outcome-related pricing to ensure revenue grows alongside the cost and value of increased usage.
Does usage-based pricing protect SaaS gross margin?
Usage-based pricing can help protect SaaS gross margin because customer revenue can increase as consumption increases. This can reduce the risk of heavy users generating substantially higher costs while paying the same fixed subscription as light users. However, the pricing unit needs to be understandable, predictable and connected to customer value. Poorly designed usage pricing can create uncertainty, make budgeting difficult and potentially slow SaaS purchasing decisions.
Will falling AI inference costs improve SaaS gross margin?
Falling AI inference costs could improve SaaS gross margin when companies can deliver the same customer outcome using cheaper models, hardware or infrastructure. But lower unit costs do not guarantee higher margins because customers and AI agents may consume substantially more computing as the technology becomes cheaper and more capable. SaaS companies therefore need to monitor total AI cost per customer, workflow and useful outcome rather than concentrating only on the cost of individual tokens or model calls.
Why is SaaS gross margin important to investors?
SaaS gross margin helps investors understand how efficiently a software company converts revenue into gross profit before sales, marketing, research, development and other operating expenses are considered. Strong margins can leave more revenue available to fund growth and product investment. For AI SaaS businesses, investors may also examine whether gross margins improve as customer numbers increase, infrastructure becomes more efficient and the company learns how to control AI delivery costs.
Can a SaaS company be successful with a lower gross margin?
Yes. A lower SaaS gross margin does not automatically mean a SaaS business is unsuccessful. A company may have lower gross margins while achieving strong growth, high customer retention, substantial expansion revenue or attractive cash generation. AI products may also deliver more valuable outcomes while carrying higher direct computing costs. Gross margin should therefore be assessed alongside customer acquisition cost, lifetime value, retention, growth, operating expenses and the wider economics of the business.
How can SaaS companies improve gross margin?
SaaS companies can improve SaaS gross margin by reducing unnecessary infrastructure expenditure, optimising AI usage, selecting appropriate models, improving model routing, caching repeated work and removing inefficient workflows. Businesses can also review expensive third-party services and understand which customers or features create disproportionate delivery costs. Pricing is equally important: the amount customers pay needs to reflect both the value they receive and the direct cost of providing the software service.
SaaS Sales Training That Improves Conversion
We provide SaaS sales training for teams who want clearer, more effective conversations. Our SaaS sales training includes sales coaching, corporate sales training, and practical workshop sessions built around real situations your team faces. We also deliver consultative selling training for SaaS businesses that want to simplify their message and close more of the right deals. Alongside our SaaS sales training, we support SaaS teams across the UK who want to communicate value better, avoid confusion, and win the right work without feeling pushy.
More sales training insights
- Why Buyers Don’t Feel Ready To Decide
- Sell Change To Comfortable Buyers
- Why Your SaaS Pricing Page Creates Doubt
- Too Many Small Customers In SaaS
- Why SaaS Buyers Forget Your Product
Ready to elevate your B2B sales techniques?
Whether you’re a B2B salesperson looking to enhance your sales skills or a leader aiming to sharpen your sales strategy in business-to-business selling, let’s work together to take your sales pitch to the next level
If you are comparing options, it helps to review focused SaaS sales training for SaaS companies that shows how clearer value leads to faster client decisions.
