Individuals are gaining
80% of respondents say AI has improved their own productivity, and about half say it helps them develop new skills and make better decisions. The pattern holds from the C-suite to individual contributors.
Pointing AI at growth
AI’s returns follow the decisions it is pointed at.
McKinsey’s The state of AI in 2026 finds that eight in ten respondents say AI has improved their own productivity. Yet the share reporting any contribution to EBIT is unchanged from last year at 37%, and the group McKinsey calls AI high performers is still about 6% of respondents.
The survey points to a difference in what AI is asked to do. Almost every organisation uses it for efficiency. Those reporting the most value also point it at growth and innovation: where to compete, what to build and which relationships to develop. Those are the decisions to which Fractionova brings AI-native Opportunity Intelligence and principal-led judgement.
The state of AI in 2026: On the road to ROI is published by McKinsey & Company and QuantumBlack, AI by McKinsey. Released in August 2026, it draws on an online survey of 1,719 participants in 97 countries, conducted from 4 May to 8 June 2026. Figures are McKinsey’s and reflect what respondents report; the interpretation is Fractionova’s. Fractionova is not affiliated with McKinsey & Company.
What do organisations ask AI to achieve?
AI high performersAll other respondents
Efficiency
78%
82%
Growth
74%
47%
Innovation
65%
49%
Efficiency is close to universal. Growth and innovation are where the organisations reporting most value differ.
The gap between productivity and profit
Adoption keeps rising. Nearly nine in ten respondents report regular use of AI, 44% say it is scaling across their enterprise, up from 38% a year earlier, and 56% use it in three or more functions. Four findings show where that use is reaching results, and where it is not yet.
Enterprise results are flat
37% attribute at least some EBIT impact to AI, about the same share as a year ago. AI high performers, who attribute 5% or more of EBIT to AI and describe its value as significant, remain about 6% of respondents.
Agents are scaling, mostly in large companies
At organisations with revenues of $1 billion or more, the share scaling AI agents rose from 27% to 40% in a year. At smaller organisations it was essentially flat.
$1bn+ revenue, 202527%
$1bn+ revenue, 202640%
Under $1bn, 202521%
Under $1bn, 202622%
At least scaling AI agents · % of respondents
Forecasts outran outcomes
A year ago, 32% of respondents expected AI to reduce their organisation’s head count over the following year. This year, 14% report that it did. Now 39% expect reductions over the coming year. An expectation is a hypothesis to test, not a basis for commitment.
Expected, 2025 survey32%
Reported, 2026 survey14%
Expected, 2026 survey39%
AI-related decline in total head count, next or past year · % of respondents
Where the value is reported
McKinsey compares its AI high performers with all other respondents. Five differences stand out, and each has a direct parallel in how a growth decision is reached.
Objectives
Growth and innovation alongside efficiency
About 80% of both groups pursue efficiency. 74% of high performers also pursue growth and 65% innovation, against 47% and 49% of others. High performers are 3.3 times as likely to intend AI to transform their business within three years.
For a growth decision
Which market, capability or relationship should AI help us evaluate, not only which cost it should cut?
Workflow
Redesigning the work, not inserting a tool
Nearly three-quarters of high performers have fundamentally redesigned workflows because of AI, up from 55% a year earlier. About a quarter of other respondents have done so.
For a growth decision
Is our next commitment reached through occasional desk research, or through connected and continuing investigation?
Judgement
Deciding where human judgement sits
High performers are more likely to have determined when model outputs need human validation, and twice as likely to say their senior leaders demonstrate ownership of AI initiatives.
For a growth decision
Who frames the question, challenges the findings and stands behind the recommendation?
Measurement
Defining impact before the work begins
High performers are twice as likely to have defined processes to quantify the impact of their AI initiatives.
For a growth decision
What evidence would justify the commitment, and what would change our view?
Cost
Treating intelligence as a cost to manage
High performers report cost constraints on software coding agents three times as often as others, 18% against 6%, and are more likely to actively manage token, compute and storage costs.
For a growth decision
How much analysis does this decision deserve, and how continuous does it need to be?
These are associations in survey data, and the high-performer group is 92 respondents. They describe what the organisations reporting most value do; they do not show that any single practice produces the result.
Two shifts that redraw commercial maps
Beyond adoption, the survey records two changes that create new growth questions for suppliers, investors and the businesses that buy from them.
Build or buy has become a market question
32% of respondents say their organisations decided against buying at least one software product or feature because they could build it with AI coding tools. Among high performers, nearly half say so.
Technology41%
Healthcare payers & providers39%
Professional services38%
Energy & materials38%
Financial institutions36%
All respondents32%
Public & social sector17%
Selected industries · % of respondents
For a business that sells software, digital features or engineering services, the customer’s own coding agents are now a competitor. The question becomes where the capability still competes, and for whom.
AI is not too cheap to meter
20%say operating costs, including tokens, have constrained AI use
60%expect AI investment to increase next year
28%spend over 10% of their ICT budget on AI
McKinsey’s Michael Chui observes that per-token costs have fallen, but the number of tokens consumed and generated has grown faster, especially in agentic software development and complex reasoning.
That rising demand for computation runs through power, cooling and data-centre capacity, one of Fractionova’s focus areas.
Decisions this page is written for
Bring the question in the form your board would ask it.
Our customers can now build what we sell. Where does our capability still compete?
Which market, partner or entry model deserves our next commitment?
We are investing in AI. Which growth decision should it inform first?
How does an AI engine investigate a decision, and where does human judgement come in?
The same practices, applied to one decision
Fractionova’s way of working reflects the practices McKinsey associates with high performers, applied to a single consequential decision rather than an enterprise programme.
| What high performers do | In a growth decision | Fractionova’s approach |
|---|---|---|
| Set growth and innovation objectives, not only efficiency. | The question is where to compete, enter, partner or invest. | Each mandate starts from the decision, not from a technology or a dataset. |
| Fundamentally redesign workflows. | Technology, demand, economics, counterparties, policy and timing are investigated together. | Coordinated specialist AI agents investigate the six commercial dimensions as one connected case. |
| Determine when model outputs need human validation. | Someone accountable tests what the findings justify. | Principal-led judgement frames the mandate, challenges analytical findings and guides the development of recommendations. |
| Define processes to quantify impact. | The criteria for commitment are explicit before the conclusion. | A recommended direction supported by evidence, alternatives and explicit decision criteria. |
| Actively manage AI operating costs. | The depth of analysis matches what is at stake. | Scope and fees reflect the analytical depth, market coverage, counterparty work and continuity required. |
AI provides the analytical breadth, coordination and continuity. Principal-led judgement establishes what the evidence supports. This describes how Fractionova designs its work; it is not a claim about any organisation’s financial results.
When customers start building what you sell
This is a hypothetical illustration, not a client engagement or a claim of achieved outcomes.
A specialist engineering-software company sells validation and test-automation tools to automotive suppliers in India and Europe. Two long-standing customers have said they will build similar tools internally with AI coding agents. Its leadership needs to decide where the next phase of growth should come from.
Fractionova’s AI engine would connect customer build-or-buy signals with segment economics, integration and data requirements, partner capability and access to new markets. Principal-led review would establish which route supports the next commitment and what evidence would justify it.
A
Move up to integration and validated dataCustomers may write their own code but still value certified data, integration and support. The questions are which customers, what they would pay and what capability that requires.
B
Concentrate on customers that still buyMid-sized suppliers without engineering capacity may remain buyers. The questions are the size of that segment, the cost to serve it and how long it stays a buyer.
C
Take the capability to a new marketMobility and industrial programmes in the Gulf could need the same tools. The questions are demand evidence, certification requirements and the local partner who can deliver access.
Three ways to commission the intelligence
You do not need to choose a format first. The starting point is the decision your business needs to make.
Decision Sprint
Concentrated intelligence for a consequential choice.
A focused investigation around a defined decision, culminating in a decision brief with the recommended direction, unresolved questions and conditions for commitment.
For example: is there still a defensible market for our tool when customers can build their own?
Market & Partner Mission
A coordinated mandate to develop market access and commercial relationships.
An integrated investigation into market priorities, entry structures and counterparties, with prioritised partners and preparation for engagement.
For example: qualify the segments and partners for a new route to market.
Fractionova Intelligence Office
A retained intelligence capability around your strategic agenda.
Recurring investigation and monitoring, supported by agreed business context and structured principal-led reviews.
For example: an agreed watch on the technology and build-or-buy shifts that affect your customers.
Each engagement is commissioned against a defined business objective. Scope and fees reflect the analytical depth, market coverage, counterparty work and continuity required.
Principal-led, AI-native
Founder & Principal Prabha Shankar leads the technical-commercial judgement behind Fractionova’s work. An engineer and international business specialist, he brings nearly three decades of experience across automotive engineering management, advanced mobility, new energy, international market development and complex programme execution.
At Fractionova, that experience shapes the questions put to the AI engine, the assumptions challenged during investigation and the commercial significance assigned to its findings.
ExperienceIndia, Europe and the Gulf, with automotive companies, institutions and government stakeholders
MotorsportSecretary of the Event for three editions of the Formula One Indian Grand Prix
EducationElectrical & Electronics Engineering; postgraduate study at Stevens Institute of Technology
Questions this page answers
What does McKinsey’s State of AI 2026 report find?
McKinsey’s survey of 1,719 participants, published in August 2026, finds that AI use is deepening: 44% of respondents say AI is scaling across their enterprise and 80% say it has improved their own productivity. The share attributing any EBIT impact to AI is unchanged at 37%, and AI high performers remain about 6% of respondents.
Why has enterprise impact not kept pace with AI adoption?
The report finds that individual productivity gains have not yet translated into enterprise financial results for most organisations. Those reporting the most value differ in how they deploy AI: they pursue growth and innovation as well as efficiency, redesign workflows, decide where human judgement applies and define how impact is measured.
What do AI high performers do differently?
74% of high performers pursue growth with AI, against 47% of other respondents. Nearly three-quarters have fundamentally redesigned workflows, against about a quarter of others. They are twice as likely to report senior-leadership ownership of AI initiatives and defined processes to quantify impact, and more than twice as likely to spend over 15% of their ICT budget on AI.
What does the build-or-buy finding mean for technology suppliers?
32% of respondents say their organisations decided against buying at least one software product or feature because they could build it with AI coding tools. Suppliers of software, digital features and engineering services need to establish where their capability still competes: in integration, data and support, in segments that continue to buy, or in new markets.
What is Fractionova?
Fractionova is an AI-native Opportunity Intelligence company, based in Chennai with a strategic presence in Dubai. It identifies, develops and qualifies commercial opportunities at the intersection of technology, markets, capital and geopolitics, combining agentic AI investigation with principal-led judgement from Founder & Principal Prabha Shankar.
More perspectives
- Mobility
Automotive’s Next Growth Map
PwC finds automotive executives turning to South Asia for growth and Indian automakers buying capability. What that means for licensing, partners and market entry.
- Trade & market access
Selling into AI-Ready Markets
Microsoft finds AI users in India and the UAE well ahead of the global average. What buyer readiness changes, and does not, for suppliers entering.
- Trade & market access
Where AI Is Built
Deloitte finds 77% of companies weigh where an AI solution comes from. What sovereign AI changes for cross-border sales, local partners and market entry.
Bring the decision that deserves deeper intelligence.
A market to enter. A capability to reposition. A partner to qualify. An investment direction to establish.
Start with what you are considering, what is at stake and what you need to determine.
Sources and notes. Figures are from McKinsey & Company, The state of AI in 2026: On the road to ROI (August 2026): an online survey of 1,719 participants in 97 countries, conducted from 4 May to 8 June 2026 and weighted by each country’s contribution to global GDP. Figures are respondents’ own reports. AI high performers are respondents who attribute 5% or more of their organisation’s EBIT to AI and report significant value from it; the group numbers 92. Fractionova is not affiliated with McKinsey & Company.
Fractionova equips clients with intelligence and considered recommendations to inform their decisions. Investment and implementation decisions remain with the client; execution and outcomes depend on the organisations involved and prevailing conditions.