Trending Articles
Trending Topics
Trending
AI’s next proving ground for small and midsized businesses will not be a standalone chatbot. It will be ERP.
Sounds unglamorous?
Unfortunately, that is where the real business context lives: invoices, inventory, customer histories, service tickets, approvals, projects, margins and cash flow. For SMBs, the question is no longer whether AI can generate a useful answer. It is whether AI can help the business act faster, with cleaner data and fewer blind spots.
That is why Acumatica’s 2026 R1 release matters beyond the usual product-cycle noise. The update adds AI Studio, AI Assistant Preview, AI-powered anomaly detection, document autotagging, CRM activity summaries, real-time visibility features and a refreshed interface.
In Acumatica’s language, “ERP is moving from a system of record to a system of intelligence”. In plain English, the company wants AI embedded where work already happens.
This also builds on the message from Acumatica Summit 2026, where the company’s clearest theme was “impact now” rather than “potential later”. The more interesting point was not that Acumatica had AI tools, but that it framed them as human-centred, configurable and tied to operational outcomes rather than lab demos.
Gartner predicts organisations will abandon 60% of AI projects unsupported by AI-ready data through 2026. The warning is obvious: AI ambition without trusted data, workflow discipline and governance quickly becomes expensive theatre.
This is where ERP vendors now have to prove themselves. The battle is shifting from who can bolt on AI fastest to who can make AI useful inside finance, distribution, construction, retail, manufacturing and services workflows.
Africa adds another layer to the story, but it is not the whole story. Acumatica’s Africa event in Johannesburg showed that this contest has also moved into growth markets, where disconnected systems and “multiple versions of the truth” remain everyday business problems.
Karsten Group, the South African agribusiness highlighted by Acumatica, makes that point more tangible. The company moved from six disconnected systems to one, eliminated 2,500 printed invoices a month and consolidated 12 databases into three tenants. This clearly demonstrates the kind of operational clean-up businesses need before practical AI in ERP can be genuinely useful.
But as I argued late last year, Africa’s technology future also depends on deeper questions of power, capital and control.
The winning AI in SMB software may not look spectacular. It may simply help businesses close faster, spot exceptions earlier and trust their numbers more. That is the battleground that matters.
