Accenture's newest enterprise AI survey found something investment alone can't explain: the share of companies reporting widespread, sustained business value from AI fell from 32% to 23% this year, even as 82% of C-suite leaders increased their AI budget (Accenture, 2026). **More AI spending isn't producing more AI value on its own — and two other major 2026 surveys point to the same specific fix: redesigning the actual job around AI, not just handing someone a new tool.**
Is AI Investment Actually Paying Off in 2026?
Not at the rate the spending would suggest. Accenture's Pulse of Change report surveyed 6,000 people — 3,000 C-suite leaders and 3,000 employees — across 20 countries and 19 industries at companies with at least $500 million in annual revenue, fielded between April and June 2026 (Accenture, 2026). Investment intent is close to universal: 82% of C-suite leaders are increasing AI spending this year. But the number that should worry them moved the wrong way. Only 23% of companies now report widespread, sustained business value from AI, down from 32% earlier in the same year (Accenture, 2026). Meanwhile the workforce is already shifting: 57% of employees say their role has already changed because of AI, and 78% of leaders expect more role changes within the next 12 months (Accenture, 2026). Jobs are moving faster than value is showing up — the opposite of what more AI spending is supposed to buy.
Why Doesn't More AI Spending Translate Into More Value?
Because most companies are training people to use a tool, not rebuilding the job around it. Deloitte's State of AI in the Enterprise survey — 3,235 business and IT leaders across 24 countries, fielded in August and September 2025 and published in January 2026 — asked companies what they're actually doing to get their workforce ready for AI. The single most common move, cited by 53%, is educating the broader workforce to raise general AI fluency. Only 30% are "combining or reimagining" their organization — restructuring roles, workflows, and career paths — based on how AI is actually being used (Deloitte, 2026). The same survey splits companies into three tiers by how far they've taken it: 37% are still using AI at a surface level, 30% have redesigned key processes around it, and only 34% are using it to deeply transform the business — building new products and services or reinventing how a core process works (Deloitte, 2026). Training is the easy move. Rebuilding the job is the one that actually shows up as value, and it's the one most companies skip.
What Happens When a Company Actually Redesigns the Workflow?
The payoff is large enough to show up across separate surveys independently. BCG's fourth annual AI at Work survey — close to 12,000 frontline employees, managers, and leaders across more than a dozen countries, published in June 2026 — found the share of organizations that have "graduated" to reshaping workflows end-to-end or inventing new business models around AI nearly doubled in a single year, from 22% to 42% (BCG, 2026). Among frontline employees who use AI regularly, 42% now report saving a full workday — eight hours a week — with the biggest gains concentrated in marketing (60%), IT (53%), and HR (50%) (BCG, 2026). But BCG found the catch sitting right next to that number: 66% of those same employees get limited or no guidance on what to actually do with the time they free up, and more than half say they aren't reinvesting it into higher-value work at all (BCG, 2026). Redesigning the workflow is what turns saved time into business value instead of just saved time.
Tool-Only AI Adoption vs. Workflow Redesign: What Actually Changes?
| Tool-only adoption | Workflow redesign | |
|---|---|---|
| Share of companies at this stage | 37% use AI only at the surface (Deloitte, 2026) | 34% have reached deep transformation of a process or business model (Deloitte, 2026) |
| Year-over-year movement | The dominant move: 53% of firms' top AI talent strategy is workforce-wide training (Deloitte, 2026) | Nearly doubled in a year — 22% to 42% (BCG, 2026) |
| What an employee actually gets | Access to a tool and training on how to run it | A rebuilt role, process, or decision point built around the tool |
| What happens to time the tool saves | 66% of regular AI users get little or no guidance on what to do with it (BCG, 2026) | Deliberately reinvested into higher-value work, per the minority BCG found doing it |
| Reported business value | Consistent with the 23% reporting sustained value overall (Accenture, 2026) | Concentrated among the smaller group that actually redesigned, across all three 2026 surveys |
The gap between those two columns is the same gap Accenture, Deloitte, and BCG each measured independently, from three different survey populations, in three different months of 2026.
When Is Just Adding an AI Tool — No Redesign — the Right Call?
Sometimes it genuinely is, and pretending otherwise isn't honest. A single-location business testing one narrow tool — an AI phone line, a scheduling assistant — doesn't need a formal workflow redesign on day one. Redesign carries its own cost: time spent mapping who does what, renegotiating who owns a decision, retraining a process that was working fine before. For a three-person shop running one tool, that overhead isn't worth paying yet. The data above describes what happens at scale — multiple tools, multiple locations, multiple departments — not what a business needs before its first AI project has even proven itself. The right order is adopt first, prove the tool works on one narrow job, then redesign the workflow around it once there's something worth rebuilding around.
What Should a Multi-Location Northwest Business Do With This?
The redesign question gets real the moment a business crosses from one location to several. A dental or veterinary group running separate offices in Missoula, Kalispell, and Spokane can hand each office manager an AI scheduling tool and call it adoption — that's the 37% surface-level tier Deloitte measured. Or it can redesign the intake workflow once, so a call to any location books into the same calendar, under the same rules, with the same follow-up sequence — which is what BCG's data says it actually is: the move that separates the companies seeing real value from the ones still waiting for their AI spending to show up anywhere. Multi-location service businesses across the Flathead Valley and the wider Northwest tend to hit this decision point first, because a second or third location is usually what turns "we added a tool" into "our tools don't agree with each other anymore."