Companies Are Spending Billions on AI. So Why Isn't Revenue Growing?

Artificial intelligence has accomplished something remarkable. It's making employees faster at nearly everything except proving that all this extra productivity is making their companies more profitable.

According to McKinsey's 2026 State of AI report, 80% of respondents say AI has improved their individual productivity. Yet only 37% report that AI has made a positive contribution to their organization's earnings before interest and taxes (EBIT).

Even more revealing, just 6% qualify as AI high performers, reporting that AI contributes at least 5% of EBIT and has delivered significant value.

Companies are generating more content, researching more accounts, producing more reports, and automating more administrative work. But the financial results haven't kept pace. For executives who have invested heavily in AI, that raises an important question.

If everyone is getting so much more done, where is the money?

‍

Key takeaways

• McKinsey's 2026 research found that 80% of respondents report personal productivity improvements from AI, but only 37% report a positive impact on enterprise earnings.
• Faster content production, prospecting, and reporting don't automatically create more revenue.
• Companies seeing the greatest returns are redesigning workflows rather than simply adding AI to existing processes.
• Measuring AI's impact requires connecting operational improvements to pipeline, conversion rates, costs, and revenue.

‍

The AI Productivity Boom Has a Revenue Problem

Consider a typical B2B marketing organization.

Before AI, the team published four articles a month, produced a handful of sales assets, and struggled to keep up with requests from the sales department.

Now, with AI, that same team can generate 20 articles, dozens of personalized emails, competitive intelligence reports, and a steady stream of social content.

On paper, productivity has skyrocketed.

But what if website traffic stays flat? What if the additional emails don't generate more meetings? What if the sales team continues struggling to convert leads into opportunities?

The organization has increased its output without necessarily improving its results.

This is the distinction many businesses overlook when evaluating AI investments.

Productivity measures how much work gets done. Business performance measures whether that work accomplishes anything valuable.

The two are related, but they aren't interchangeable.

A company can automate thousands of tasks and still miss its revenue target. It can reduce the time required to research a prospect without improving the quality of its outreach. It can generate an impressive volume of marketing content without attracting more qualified buyers.

AI has made producing work easier. Determining which work matters remains a business decision.

Why Faster Doesn't Always Mean Better

The first wave of enterprise AI adoption largely focused on individual tasks.

Marketing teams adopted AI writing assistants. Salespeople used prospecting tools. Revenue operations teams experimented with automated reporting. Customer success departments introduced AI-powered support capabilities.

Each department found opportunities to save time.

The trouble started when those improvements remained confined to individual functions.

Imagine a company with an AI-powered prospecting process that identifies hundreds of promising accounts.

The marketing team has another tool generating personalized campaigns. Sales uses a separate platform to manage outreach. RevOps maintains the CRM and monitors pipeline performance.

Every department has become more efficient.

But the systems don't necessarily share the same information, follow the same priorities, or respond to changes in customer behavior.

Marketing may continue targeting accounts that sales has already disqualified. Sales may spend time researching information that another team has already collected. Revenue operations may discover problems weeks after they began affecting conversion rates.

The company has automated the work without fixing the process.

In some cases, AI simply allows existing inefficiencies to operate at greater speed.

The Companies Getting Results Are Changing How Work Gets Done

McKinsey's research offers an important clue about what separates successful AI investments from disappointing ones.

Nearly three-quarters of AI high performers report fundamentally redesigning workflows because of AI, compared with approximately one-quarter of other respondents.

These organizations aren't merely asking employees to use AI tools.

They're reconsidering how work moves through the business.

For go-to-market teams, that distinction matters.

Traditional GTM operations often depend on a series of departmental handoffs. Marketing identifies prospects, sales development qualifies them, account executives manage opportunities, and RevOps monitors performance.

Each stage generates information that the next team needs.

When those transitions depend on manual updates, disconnected systems, or inconsistent processes, valuable context can disappear.

AI offers an opportunity to address those gaps.

For example, an automated GTM workflow could identify an account showing buying signals, enrich its profile, evaluate its fit against the ideal customer profile, prepare relevant outreach, and route the opportunity to the appropriate representative.

The process still requires reliable data, business rules, oversight, and human judgment. But the work no longer has to depend on several employees independently moving information between systems.

That's a more consequential use of AI than simply writing an email faster.

The Hidden Cost of Disconnected AI Investments

AI's operating costs are also becoming harder to ignore.

McKinsey found that approximately 20% of respondents say AI-related operating expenses, including token costs, have constrained their organizations' use of the technology.

That creates a second challenge for companies investing in multiple AI applications.

A business may pay for AI capabilities in its CRM, marketing platform, sales engagement software, analytics tools, and productivity applications.

Each purchase can appear reasonable on its own.

Together, however, those investments may create overlapping functionality, additional integration requirements, and greater operational complexity.

The issue isn't necessarily the number of tools. It's whether they contribute to a coordinated process with measurable results.

Consider two approaches to improving outbound sales.

In the first, a company purchases separate AI applications for account research, email personalization, lead scoring, and reporting. Employees coordinate the results manually.

In the second, the organization designs a connected workflow that handles those activities in sequence, using shared data and defined decision rules.

Both approaches use AI.

But the second makes it easier to understand how the individual activities contribute to an overall business outcome.

It also creates opportunities to identify where the process breaks down and make improvements.

The objective should be to reduce unnecessary work, not simply add more technology to the existing workload.

Five Metrics That Reveal Whether AI Is Actually Paying Off

Executives looking to measure AI's financial impact need to move beyond adoption rates, hours saved, and the number of automated tasks.

Those measures can be useful, but they don't tell the entire story.

A more meaningful evaluation should include five areas.

1. Pipeline generated per dollar invested

Are AI-powered marketing and prospecting workflows producing more qualified pipeline relative to their total cost?

This calculation should account for software, implementation, maintenance, and the human effort required to supervise the process.

2. Lead-to-opportunity conversion

If AI helps a company research and contact more prospects, does a greater percentage become qualified opportunities?

An increase in outreach volume accompanied by declining conversion rates may indicate that automation is creating activity rather than demand.

3. Sales cycle length

Can AI reduce delays associated with account research, lead qualification, follow-up, and internal handoffs?

Shorter sales cycles can improve capital efficiency, although companies should also monitor deal quality and win rates.

4. Cost per qualified opportunity

How much does the organization spend to generate an opportunity that meets its qualification standards?

This metric can reveal whether automated workflows are improving efficiency or simply increasing operating expenses.

5. Revenue per employee

Can the organization support more revenue without increasing staffing at the same rate?

Revenue per employee is influenced by many factors beyond AI, so it shouldn't be treated as proof of AI's contribution on its own. Combined with controlled comparisons and other performance measures, however, it can help executives evaluate whether workflow improvements are translating into greater organizational capacity.

These metrics provide a more useful foundation for AI investment decisions than counting how many employees have adopted a particular tool.

From AI Tools to Revenue-Producing Workflows

The next stage of enterprise AI adoption will require companies to examine how work connects across departments.

That's particularly important for go-to-market organizations, where revenue depends on the coordination of marketing, sales, customer success, and operations.

Platforms such as Copy.ai approach this challenge through AI-powered GTM workflows that connect activities such as account research, prospecting, content development, and sales execution.

Rather than treating each AI-generated output as an isolated accomplishment, the goal is to organize those activities around a repeatable business process.

An account research workflow, for example, becomes more valuable when its findings inform qualification, personalization, and sales follow-up.

Likewise, marketing content becomes more useful when it supports specific buyer needs and can be incorporated into the sales process.

The technology creates opportunities for efficiency. The workflow determines whether those efficiencies contribute to revenue.

The Next AI Investment Should Start With a Business Problem

Nearly nine in ten organizations surveyed by McKinsey report regularly using AI in at least one business function. The more important question is what companies are doing differently because of it.

Are they reaching better prospects? Converting more opportunities? Reducing customer acquisition costs? Improving sales execution? Removing the operational delays that prevent revenue teams from performing?

Or are they simply generating more work, faster?

The companies that ultimately benefit most from AI may not be those with the largest collections of tools or the highest automation counts.

They'll be the organizations that can connect AI investments to business outcomes and demonstrate where those investments create value.

Because nobody on the executive team wants to hear that the company produced 10,000 more emails this quarter.

They want to know whether any of those emails helped close a deal.

Explore how Copy.ai's GTM AI platform can help connect disconnected tasks into coordinated revenue workflows.

‍