AI
Why Businesses Struggle to Put AI to Work: The Barrier Isn't the Technology
AI is no longer a distant idea for most businesses. But there is still a wide gap between trying an AI tool and running part of the business on it, and most companies are stuck somewhere in between.
In a 2024 survey of 1,000 executives across 59 countries, Boston Consulting Group (BCG) found that 74% of companies have yet to show tangible value from AI. Only 26% have the capabilities to move beyond proofs of concept, and just 4% have built cutting-edge AI capability across functions.
The reason is the interesting part. BCG attributes roughly 70% of AI implementation challenges to people and process, 20% to technology, and only 10% to the algorithms. In other words, the biggest barrier is not the technology. It is how the business applies it.
Key Takeaways
- 74% of companies have yet to see tangible value from AI; only 26% get beyond proofs of concept (BCG, 2024).
- Around 70% of AI implementation challenges come from people and process, not technology.
- 78% of US enterprises struggle to connect AI tools with their existing systems (Zapier, 2025).
- SMEs should start with one specific process, clean up the data behind it, and measure results before scaling.
Source: BCG, AI Adoption in 2024: 74% of Companies Struggle to Achieve and Scale Value, published 24 October 2024.
Australian SMEs are no exception
In Australia, the National AI Centre's SME AI Pulse found that 44% of small and medium businesses reported some level of AI adoption in February 2026. The most common uses were content generation and data analytics (54% each), followed by cybersecurity (48%).
Most of these sit alongside existing processes rather than inside them. The step from “we use AI” to “AI creates value for us” is where most businesses get stuck.
Barrier 1: Scattered, inconsistent data
AI only performs well when the data going into it is clean and consistent. In many businesses, customer data lives in the sales system, quotes live in email, and operating figures live in separate spreadsheets kept by each team.
When the same customer is recorded three different ways, whatever the AI produces will be unreliable or unusable. Cleaning up the data is usually the most time-consuming part of the work, and it is also what decides whether an AI project succeeds.
Barrier 2: Existing systems that do not connect
A Zapier survey published in October 2025 found that 78% of US enterprises struggle to connect AI tools with the systems they already run, and 53% rate integration as moderately to extremely difficult. Data quality (29%) and AI skill gaps in the team (35%) were also among the leading barriers.
Smaller businesses rarely have an IT team to handle this. The result is AI that lives in a separate chat window while staff still copy its output into the accounting system or CRM by hand.
See our AI & Workflow Automation service
Barrier 3: Starting with the tool instead of the problem
Many AI projects start with “which tool should we use?” when the better question is “which process costs us the most time and the most errors?”. Without a specific problem, projects tend to stall at the demo, or spread thinly across the business without delivering a clear result anywhere.
BCG also found that 62% of AI's value sits in core business processes rather than in peripheral support tasks.
Barrier 4: People and processes are not ready
AI leaders split their resources 10-20-70: 10% on algorithms, 20% on technology and data, 70% on people and processes (BCG). AI is not software you install once. It changes how work gets done: who checks the output, who handles the exceptions, and who is accountable when something goes wrong.
In Australia, about half of current AI users already check AI output before it reaches a customer (National AI Centre). That discipline is what makes AI use safe and trustworthy.
Where should a small or medium business start?
You do not need a thirty-page AI strategy. These four steps are enough to find out whether AI genuinely fits your business.
1. Pick one specific process
Favour work that repeats often, follows clear rules and eats time: entering invoices, preparing quotes, sorting customer enquiries.
2. Clean up the data behind it
You don't need to clean every record in the business. Just make sure the data that process relies on is recorded consistently, in one place.
3. Connect it to the systems you already use
AI should write its results straight into your accounting software, CRM or job system. If staff still copy and paste, the process has not really been automated.
4. Measure before and after
Record hours spent, error rates or turnaround times before you start, then compare after a few weeks of real use. Measured results are the only sound basis for scaling up.
When AI is not the right move yet
If a process is not stable yet and everyone does it differently, automating it only makes the mistakes happen faster. In that case, standardise the process first and consider AI after.
Likewise, for decisions that directly affect money, customers or legal obligations, AI should help prepare the information while a person makes the final call.
Common blockers we see in SME operations
Frequently Asked Questions
Do small businesses need a data team before using AI?
No. What matters is clean, consistent data for the one process you start with, and someone who owns that process. A data team comes much later, if at all.
Why do AI pilots stall after the demo?
Because the demo runs on tidy sample data. Real operations need live data, connections to existing systems, user permissions and a plan for exceptions. That work is usually larger than the pilot itself.
How do we know whether AI is paying off?
Measure before you start: hours spent, error rate or turnaround time for the chosen process. Then compare after a few weeks of real use. If you cannot measure it, it is not ready to automate.
The Short Version
Most businesses are not short of AI tools. What they lack is a specific problem, clean enough data and a clear enough process for AI to deliver value. Start with one process, measure the result, then scale.
Holdrix handles the tech, you focus on growth.
Want a second opinion on which process to put AI on first? Book a consultation.
Which process in your business should use AI first?
Talk to the Holdrix team for advice and find the right solution for your business.