SMEs could be wasting thousands each year on unnecessary SaaS tools, as bloated software stacks and rising subscription costs continue to spiral
Software-as-a-Service, commonly known as SaaS, has become the backbone of modern business operations. But for many, growing teams and rising subscription costs are quickly draining internal IT budgets.
But with the rise of Artificial Intelligence (AI), and consequently, AI agents, we are seeing the reliance on software being flipped on its head.
AI agents are autonomous or semi-autonomous systems capable of acting on their own accord, working across platforms and automating workflows. When deployed strategically, AI agents can significantly reduce the need for SaaS and subscriptions, reducing costs, and potentially improving efficiency.
New insights from Fasthosts suggest that SMEs could be overspending on SaaS tools by thousands each year, as growing software stacks and rising subscription costs continue to put pressure on IT budgets.
The SaaS cost problem
Many SMEs use anywhere between 10 to 20 SaaS tools across marketing, sales, and operations, making software one of the fastest-growing operational costs in business. By the end of 2025, the global software-as-a-service (SaaS) market was estimated to be worth 299 billion U.S. dollars.
A report from Medium found that on average SaaS prices are increasing 8-25% annually, with SaaS currently running at 5x general market inflation. In the same report, it was noted that businesses across the globe now spend an average of $7,900 per employee annually on SaaS tools, a 27% increase over the previous two years.
Additionally, in 2024 evidence 40% of organisations had at least 1-2 redundant SaaS tools in their repertoire. And despite high tool dependency, last year, Workspace 365 found that almost four in ten (39%) don’t use the programmes their companies have invested in.
Let’s do the maths
Let’s take a hypothetical SME in the UK, a small digital marketing agency with around 15 employees. It might rely on tools like Slack for communication, Asana for project management, HubSpot for CRM, SEMrush for SEO, Hootsuite for scheduling, Xero for accounting among others.
We could estimate a business like this relies on 10-15 SaaS subscriptions for its daily operations, before even factoring in additional tools for reporting, automation, integrations, or niche requirements.
If each tool costs an average of £10–£50 per user per month, a small agency could easily be spending £150 – £300 per employee monthly on software alone, amounting to £27,000 to £54,000 per year for a team of 15. Removing just two tools could result in savings of up to £10,000 per year.
With overlapping functionality, underused licences, and multiple tools required to complete a single workflow, it’s easy to see how SaaS costs can spiral. And this is exactly where AI agents are beginning to shift this model by consolidating tasks and reducing the need for multiple tools, while most crucially, cutting unnecessary spend.
How AI and Automation can reduce costs
Search interest in AI agents has surged over the past year, with terms like “what is an AI agent” and “agentic AI” seeing exponential growth. Alongside this, tools like Claude, Zapier, and Lindy AI are accelerating the shift towards more autonomous workflow solutions.
AI Agents have exemplified the latest SaaS trend of low-code-no-code products, lowering entry barriers to software traditionally requiring IT teams or complex technical setups. AI-agents and AI-powered automation allows organisations to reduce and consolidate workflows by combining tasks that would normally require several tools to work together. Similar to SaaS, they are cloud-operated and can run on a simple virtual private server (VPS), making them accessible and scalable for SMEs without significant infrastructure investment.
However, where they differ is in how work is executed. Rather than relying on multiple disconnected platforms, AI agents act as a single orchestration layer. They can analyse data, trigger actions, and move between systems without manual or human input. This means tasks such as lead management, customer communication, reporting, and scheduling can be handled by a singular agent end-to-end within one unified workflow.
As a result, businesses are beginning to shift away from stacking multiple SaaS subscriptions, and rather towards deploying fewer, more capable systems. By consolidating tools into agent-led workflows, organisations can reduce software sprawl, minimise duplication, and ultimately lower operational costs while maintaining, or even improving, overall business efficiency.
AI Agents are changing SaaS pricing structures
For years, “per-seat” or per-user pricing has been the foundation of SaaS pricing. But AI is fundamentally disrupting this model. When a single AI agent can now perform the work of multiple employees, per-user pricing no longer really reflects value.
As a result, SaaS providers are shifting towards usage- and outcome-based pricing models, where in lieu of paying for access, businesses are increasingly being charged for actions such as conversations, resolutions, or tasks completed instead.
Many SaaS platforms are also now beginning to introduce AI agents and features as premium add-ons within “pro” or “advanced” tiers, layering additional costs on top of existing subscriptions. For example, Salesforce Agentforce charges around $2 per conversation, whereas Zendesk AI charges $1.50–$2.00 per automated resolution, alongside additional per-agent fees.
So while it is estimated that by 2026, around 40% of enterprise SaaS will include outcome-based pricing elements, this pricing shift presents a paradox: while SaaS companies are using AI to justify higher prices, AI agents themselves are reducing the need for multiple tools altogether.
Final words
Ultimately, AI agents could be seen as a double threat; They won’t just reduce SaaS costs, but instead will continue to shape how businesses consider their work processes all together. Instead of paying for access to dozens of fragmented tools, we might see businesses take the upper hand, reducing their reliance on uncoordinated software subscriptions, and move towards a more streamlined working model.
While concerns about AI replacing jobs continue to dominate headlines, these systems also present a clear opportunity: to handle workflows more efficiently, reduce costs, and enable businesses to focus on higher-value work. The question for businesses now is no longer how many tools they need, but rather how few they can actually operate with.