Recent research from McKinsey and BCG demonstrates that B2B growth leaders are more likely to increase their AI investments compared to other companies. This is due in part to their use of agentic workflows, which streamline processes and enhance efficiency.
This story was produced through MarketScale. See how Marketing Tech teams put it to work with AI Writing.
Two of the world’s largest strategy consultancies published research on the same day this week arriving at the same uncomfortable finding: most B2B companies are already doing something with AI, and most of them are not getting much back for it. The reports, from McKinsey and BCG, each released on July 16, 2026, describe a widening performance gap driven not by access to AI tools but by willingness to rebuild commercial workflows around them.
The 3x investment gap between growth leaders and the rest
McKinsey’s 2026 B2B Pulse Survey, which covers nearly 4,000 buyers and sellers across 13 countries, puts a precise number on the divide. High-growth B2B companies, those outpacing their peers on revenue expansion, are three times more likely to have increased AI investment by double digits this year: 71% of growth champions raised AI spend significantly in 2026 versus 25% of laggards, according to McKinsey.
The report is careful to separate investment volume from investment quality. Fragmented data, manual processes, disconnected sales and marketing teams, and thin change management continue to cap impact at most organizations. Layering AI on top of those constraints, McKinsey’s researchers argue, often just automates complexity rather than removing it.
Share of companies that increased AI investment by double digits in 202671High-growth companies25Laggard companiesMcKinsey 2026 B2B Pulse Survey · © MarketScaleDownload chart
What the growth leaders are doing differently is redesigning what McKinsey calls “impact journeys,” end-to-end commercial workflows that fuse data, decision logic, human judgment, and AI agents across the full arc of a deal, from opportunity identification and account planning through pricing, proposals, and post-sale expansion. The point is compounding: gains from individual AI use cases stack when the underlying workflow is continuous rather than siloed.
CMOs are behind on execution, not ambition
BCG’s concurrent research, drawn from an annual survey of nearly 300 global CMOs, finds a similar ambition-execution gap on the marketing side. Ninety-six percent of CMOs say AI is driving end-to-end transformation of their function, according to BCG. Only about 31% have made agentic execution a reality.
Ninety-six percent of CMOs say AI is transforming their function end to end. Only about a third have actually built the operating model to prove it.
That credibility deficit has external consequences. BCG cites Gartner data showing only 14% of CEOs and CFOs consider their CMO highly effective at driving market growth, a figure that puts marketing leaders on thin ice at precisely the moment their function is being asked to absorb the most organizational change.
BCG’s report frames the fix as a three-pillar rebuild: brand stewardship, market and customer intelligence, and customer experience. Each needs to be reimagined with AI before they can be integrated. The BCG team is explicit that today’s AI investment in marketing is still skewed toward efficiency, and that the larger prize, using AI as a growth engine, remains largely uncaptured.
Agentic AI introduces a second buyer at the table
BCG’s report raises a structural issue that goes beyond internal workflow redesign. As customers increasingly turn to AI agents to discover products, evaluate vendors, and complete transactions, B2B brands now face two decision makers on every deal: the human buyer and the AI model influencing their choices. BCG describes this shift as the beginning of agentic commerce.
The implication for brand and product teams is concrete. AI systems can instantly aggregate reviews, service records, complaints, and other publicly observable signals, making the gap between a brand’s marketing claims and actual customer experience far more visible than it was in a purely human-mediated buying process, according to BCG. Visibility to an AI agent is not the same as being recommended by one; the agent evaluates trustworthiness based on performance data, not messaging.
This also generates a new class of customer intelligence. AI-mediated interactions surface buyers’ underlying motivations, goals, and tradeoff preferences in a way that traditional search queries do not, BCG notes. For revenue operations and product teams, that signal has potential value well beyond the marketing function, feeding into pricing strategy, product roadmap, and partner decisions.
What operations and revenue leaders need to do now
Taken together, the McKinsey and BCG findings point to the same operational priority: the question is no longer whether to invest in AI but whether the organization is structured to let agentic AI compound across commercial workflows rather than run as isolated point tools. McKinsey’s research shows the companies doing this are already running away from the field on growth metrics.
McKinsey’s report identifies five specific impact journeys where agentic AI is delivering measurable lift when deployed end to end, building on gen AI use cases the firm documented in 2025 as delivering results across the B2B deal cycle. The current emphasis is on connecting those capabilities into continuous, self-reinforcing workflows rather than maintaining them as standalone features.
For CMOs, BCG’s data makes a strategic case that is also a political one: with CEO confidence in marketing leadership at 14%, demonstrating that agentic AI investments produce growth, not just cost reduction, is among the highest-priority moves available. The BCG research suggests that CMOs who align brand promise tightly with customer experience and integrate AI-generated intelligence into business decisions beyond the marketing function will be best positioned to close that credibility gap. The cohort that has already moved beyond pilots to full agentic execution, roughly a third of the 300 surveyed, is where the early proof points will come from.
