AI & Technology

The Greatest Barrier to AI in Indirect Spend Isn’t Technology, It’s Adoption

By Ilija Ugrinic, Senior Vice President of Strategy & Marketing at Proactis.

AI has become the headline act in every procurement and finance conversation. It promises demand prediction, supplier recommendations, automated decisions and efficiency gains no team could match by hand. 

But there’s an uncomfortable truth: AI is only as intelligent as the data it’s given. In indirect spend management, the greatest barrier to AI’s potential isn’t the technology. It’s human behaviour. 

The Data Gap Behind the AI Hype 

Organisations have invested heavily in procurement platforms, purchasing systems and invoice automation for years. Despite this, indirect spend still depends on people and process as much as technology. Employees bypass approved catalogues. They ask suppliers to fulfil orders outside agreed processes, negotiate directly, or revert to email and spreadsheets because these feel quicker. Policies like “No PO, No Pay” help, but adoption remains patchy across parts of the process.   

Every manual shortcut creates more than inefficiency. It creates a data gap, and AI cannot learn from decisions it never captured.  

This is the paradox many organisations face. They want to explore AI while overlooking the foundation it depends on. The result: AI hunting for patterns in incomplete, inconsistent or biased data, and reinforcing poor behaviours rather than correcting them. 

An AI engine analysing 60% of purchasing activity isn’t learning how the business spends. It’s learning how the system is used – a very different thing. The conversation about AI in indirect spend management needs to shift. The first step isn’t artificial intelligence. It’s optimising adoption of the automation already in place.  

Automation as the Foundation for Intelligence 

Automation has long been seen as a way to reduce effort. Its greater value is creating complete, structured, reliable data. Every requisition raised through the right workflow, every supplier chosen from an approved catalogue, every approval completed digitally – each adds to the intelligence AI will depend on.

Automation isn’t just digitising work. It’s building the data foundation the future AI strategy depends on. Without broad adoption across the spend lifecycle, AI is solving a puzzle with pieces missing – a leadership challenge, and a real risk to AI delivering on its promise.   

Technology programmes have traditionally measured success by implementation: systems deployed on time, applications integrated, workflows configured. Increasingly, success will be measured by what people actually use. 

The organisations that get the most from AI won’t be those with the most advanced technology. They’ll be those with the strongest culture of digital adoption – shifting the goal from compliance to participation. Only then does the data become rich enough for AI to move from automation to genuine intelligence. 

With that foundation in place, AI delivers on its promise. It spots spending trends before they become budget problems. It recommends suppliers based on outcomes, not habit. It surfaces risks buried across thousands of transactions no person could realistically check. And it keeps improving with every interaction. 

The need for adoption has never been greater. Earlier technology automated manual tasks. AI augments decision-making itself, and augmentation without understanding is a real risk. An AI model trained on incomplete data doesn’t become intelligent. It becomes confidently wrong – and in organisations making hundreds of thousands of spend decisions a year, that can amplify poor decisions faster than any manual process ever could.  

Where the Real Competitive Advantage Lies 

The competitive advantage won’t belong to whoever adopts AI first. It will belong to whoever builds the strongest foundation of adoption and clean, accurate data – earning the right to move to the next stage of AI, and the benefits that come with it.  

The fundamentals are simple: implement process automation, then focus relentlessly on adoption. Get both right, and you build a complete, trusted, representative data foundation – and with it, an intelligence layer that keeps learning, predicting and improving. 

AI will unlock benefits in indirect spend management we haven’t yet imagined. But success won’t start with algorithms and features. It will start with people – the behaviours in the process, and the data those behaviours create.  

The organisations that recognise this today will be the ones that realise AI’s promise – not because they deployed smarter technology, but because they built the smarter conditions for it to learn.  

 

Related Articles

Back to top button