Most plants are funding the AI model and treating integration as a project cost.
The biggest buyers in the market spent the last 12 weeks doing the opposite. Accenture, Honeywell, and IFS, all paying up for integration talent instead.
Before we get into what changed, quick question:
Where does most of your AI budget go?
👋🏻 I'm Leonardo Ubbiali. This week we're looking at what the buyers paid up for, and what that says about where your own AI budget should sit.
Accenture and IndX announced the deal from Rome and Chicago on June 17.
IndX has 650+ professionals spread across Italy, the US, India, Germany, Mexico, and other European countries. The work is unglamorous. Wiring PLM, digital twins, SCADA, and industrial edge into a single stack.
Tracey Countryman, Global Supply Chain and Engineering Lead at Accenture, said manufacturers are investing in software, data, and AI to make their operations more flexible and connected, but most companies struggle to integrate those technologies across products, factories, and supply chains.
Tony Hemmelgarn, President and CEO of Siemens Digital Industries Software, called the deal a milestone for the Accenture Siemens Business Group. The group was formed in 2025. Two new Centers of Excellence are planned in Italy and India.
Accenture paid for the people who can connect models, not for the models themselves.
What's Driving The Bids
Three deals in twelve weeks.
Accenture bought 650 Siemens-certified integration specialists. Honeywell, at its investor day, called out a $2 to $4 billion automation deal as its next target, and IFS picked up Softeon for warehouse integration.
The buyers are all paying for the layer that connects a model to a 20-year-old historian, an MES, an ERP, and a SCADA stack.
That layer is hard to hire for, and the supply is shrinking.
Per BCG's research on AI value capture in manufacturing, only 10% of the value comes from the AI and algorithms themselves, 20% comes from the IT and OT infrastructure underneath, and 70% comes from people, processes, and the change management around them.
The 70% is where the money is, and almost nobody is funding it.
I have looked at how most US and UK plants are allocating their AI spend in 2026.
The biggest line item is usually the model, platform, and vendor seats, while integration sits below that, buried as a project cost, often outsourced to a local SI.
That allocation is upside down.
Grant Thornton's 2026 AI Impact Survey of 950 leaders found that only 10% of manufacturers have fully integrated AI into operations.
The biggest buyers in the market just paid up for the integration layer.
The plants getting this right are flipping the ratio.
They are funding integration first and the model second. The data foundation, historian connections, and the OT-to-IT wiring are getting their own budget line, not buried inside the AI project.
In-house integration bench is growing because the consolidation is closing the door on outsourced talent, and every new AI vendor contract has a clause that lets the integration partner of the plant's choice access the system.
Anyone budgeting 2027 AI work as a model-and-platform line will spend the next 18 months learning what their integrator costs.
Five Things You Can Do This Quarter
The problem: You need to know whether your AI budget reflects where the work happens.
What you need: A breakdown of your current AI spend by category (model, platform, integration, data, vendor seats), your in-house integration headcount, and the SI partners doing integration work today.
The Prompt (copy this):
I'm a [YOUR ROLE] at a [FACILITY TYPE] manufacturer. Our AI budget this year is roughly [AMOUNT], split across [LIST CATEGORIES AND AMOUNTS]. Our in-house integration team is [HEADCOUNT AND SKILLS]. We outsource integration work to [LIST SI PARTNERS AND ANNUAL SPEND].
Tell me: Is my budget split where the value lives, or where it has always lived? What share should be going to integration and data based on what BCG and the biggest buyers are saying? Which categories should I cut to fund integration? What does my in-house bench need to look like to stay independent of any one consultancy?
What you'll see:
A reallocation map of your AI budget, a target split between model, platform, integration, and data, a hiring plan for the in-house bench, and a renegotiation checklist for your SI partners.
McKinsey Global Tech Agenda 2026
The McKinsey survey of 632 C-level executives shows AI has surpassed cybersecurity and infrastructure modernization as the top investment priority for 2026. The section on how top performers invest in AI gives the cleanest read on where 2027 spend is shifting.
Time to value: 25 minutes
Accenture paid for 650 integration specialists instead of a model. Where is your AI budget paying for the model and not the people who can connect it?
Hit reply. I read every email.
Leo