Decidr

My SaaS stack is messy, too cumbersome, driven by legacy and requiring evolution

Does this sound familiar? You’re not alone. The truth is, it doesn’t require evolution. It requires revolution.

As you explore AI opportunities within your organisation, you should be thinking horizontally. Ask yourself, “What platform can drive efficiency and productivity across my entire business?” If you adopt AI in the same way you built your SaaS stack i.e. piecemeal, problem by problem - you’ll end up with an expensive, fragmented AI ecosystem that replicates the very inefficiencies you were trying to fix.

For decades, businesses have been caught in a cycle of ever-expanding software dependence - one SaaS tool at a time. Each promising efficiency but instead creating a tangled web of complexity. This SaaS tax - the hidden cost of software stacks, redundant tools and siloed data - has drained budgets, slowed innovation and made businesses dependent on costly, fragmented ecosystems​.

“There must be a better way!” I hear you cry as you drop to your knees in Baker’s Delight. Well, AI is the better way, but let’s ensure we don’t solve today’s issues with a completely new set of problems.

Allow me to explain…

The hidden cost of the AI tax

For decades, businesses have been caught in a spin cycle of ever-expanding software dependence, one SaaS tool at a time. Each promising efficiency but instead creating complexity and redundancy. This “SaaS tax” is the hidden cost of tangled software stacks, siloed data and overlapping tools. It drains budgets, slows innovation, and locks businesses into fragmented ecosystems.

AI has the potential to revolutionise enterprise operations, but most businesses I speak to are making the same mistake again. They are creating an AI tax, a costly accumulation of disconnected, vertical AI tools stacked on top of existing systems. This tax will drive up expenses, introduce inefficiencies and diminish returns.

AI investment is surging… But at what cost?

AI spending is growing 50% year-over-year, with AI-enabled applications projected to increase by 140%. Businesses are moving beyond the ‘AI Fear’ narrative to embrace the ‘AI Opportunity’. However, the challenge I hear most often is:

We know AI presents a massive opportunity, but we don’t know where to start.

But the real issue isn’t just where to start, it’s how. If you get the approach wrong, you’ll simply add an AI tax to your existing SaaS tax.

In recent weeks, I’ve spoken to multiple businesses on the path to overspending on disconnected AI tools without realising meaningful ROI. Take, for example:

• A retailer that has added two separate AI tools to solve two different efficiency problems, each requiring separate integration, management and costs.
• A logistics company currently running five active AI experiments, each from different providers, all operating in silos.

The problem isn’t AI itself, it’s how businesses are implementing it. AI is being treated as another layer of SaaS, rather than a new way to operate. To break free, businesses need to think agentic and horizontally.

The AI Tax: A Costly Burden

Businesses adopting AI without strategic integration risk:

• Fragmentation – Multiple disconnected solutions lead to redundant workflows and management headaches.
• Runaway Costs – AI tools requiring licensing across departments that may not fully utilise them.
• Maintenance Challenges – IT teams spend more time troubleshooting than driving innovation.
• Inefficient Management – AI tools operating in silos, preventing businesses from extracting full value from their data.

This is not the future of AI. AI shouldn’t follow SaaS’s mistakes, where more tools mean more costs, inefficiencies, and technical debt. The solution isn’t more AI… it’s better AI.

Be agentic: AI as the software, not just another tool

Instead of stacking AI on top of existing software, businesses must rethink their software architecture entirely. AI should not just augment tasks. It should replace fragmented, inefficient approaches.

Agentic AI is:

• Autonomous – It actively manages, optimises, and learns from business processes in real time.
• Integrated – It removes the need for middleware, enabling seamless workflows across the organisation.
• Adaptive – It evolves dynamically with business needs, reducing the need for constant human intervention.
• Goal-Driven – It doesn’t just execute predefined tasks. It identifies and optimises the best path to achieving business objectives.

Today, businesses are overpaying for inefficiency. Instead of investing in dozens of separate AI tools, the future belongs to one intelligent AI system that orchestrates finance, HR, marketing, customer service, and product development as a single entity.

One AI platform to rule them all

The biggest driver of the AI tax is vertical AI, where businesses deploy specialised AI models for different tasks, failing to unify them into a single, cohesive system.

This approach mirrors the SaaS bloat of the past and traps businesses in a cycle of inefficiency. Worse, it feels invisible at first, until suddenly, you have as many AI platforms as you once had SaaS platforms. This would be the part of the movie where ominous tones would be heard to signify impending danger.

I advise business leaders to pause and reflect: “Do you really want an AI stack that mirrors your bloated SaaS stack?” The answer is a resounding no.

The Alternative? Horizontal AI.

A horizontal AI platform functions as a single intelligence layer, seamlessly integrating across business functions, leveraging data from multiple sources instead of operating in silos. A horizontal AI approach means:

• Goals, not just tasks
• Productivity, not just efficiency
• A sustainable, scalable solution
• More collaboration opportunities
• Lower costs, higher value
• Better AI utilisation
• Less risk of overspending with little return
• Faster AI adoption and literacy
• A true revolution, not an evolution

Imagine a single AI brain within your organisation, rather than hundreds of disconnected AI tools. It would create a unified knowledge base, ensuring data flows freely across your business, rather than being trapped in silos.

It’s really rather exciting.

The AI Revolution Has Already Begun

The future of AI isn’t about adopting multiple tools, it’s about eliminating the need for them.

Businesses that embrace agentic, horizontal AI will be faster, more efficient and more resilient, while those clinging to fragmented AI models will struggle under the weight of inefficiency.

The AI revolution isn’t just an evolution of SaaS, it’s a complete departure from it. Businesses that recognise this shift now will be the ones leading tomorrow.

The question isn’t if AI will transform your business, but how.

So, will you layer AI on top of inefficiency, or will you replace inefficiency with intelligence?

The choice is yours.