For large businesses, artificial intelligence promises to serve as a path to increased productivity and bigger profits. But C-suites are often pushing for transformational AI implementation without agreeing on who owns the process or what success looks like.
That turns the process on its head. Instead of a straight line, enterprise AI can become a winding road with potholes and switchbacks. The moment the C-suite’s AI conversation shifts from “let’s do it” to “let’s actually make this happen,” things get complicated.
Rolling AI out across a large enterprise means stitching the technology into dozens of workflows owned by different teams, each with its own systems, incentives, risk tolerance levels and definitions of “good.” What might look like a single initiative from the executive suite quickly becomes a massive coordination challenge across business units.
Not just data governance and security, legal and compliance, and product marketing teams are swept in. Employees whose teams and functions are touched by the technology have to be trained. Disparate tech flows have to talk to each other. Getting business units, let alone the entire company, to line up in formation becomes a giant heave.
“For most large enterprises, organizational readiness is still the bigger barrier than cost,” Ben Schein, chief analytics officer, SVP of product at Domo, told PYMNTS.
New research from PYMNTS Intelligence’s “The Enterprise AI Benchmark Report” reveals that over 7 in 10 executives (71%) at companies with $1 billion or more in annual revenue believe that organizational readiness is the primary limitation on AI performance. Meanwhile, just 11% think that AI technology itself is the main barrier.
In other words, nearly all executives surveyed agreed AI can add value to their company. But most also feel that internal bottlenecks are holding things back. “What’s really holding it back is that most AI tools don’t learn and don’t integrate well into workflows,” a report from MIT report said last July.
The Readiness Challenge
The biggest limitation enterprises face is rarely AI technology itself — it’s usually the company’s ability to harness it.
The gap between AI’s theoretical capabilities and real-world impact is increasingly tangible across business units at large enterprises. Finance wants more accurate forecasting and streamlined quarterly reporting, but the data is spread across dashboards and databases with mismatched formats. Sales wants AI to draft proposals, but customized CRM modules often require manual inputs. Customer support wants automation, but the source of its policies lives in PDFs and email threads. Only 5% of enterprises have AI tools integrated in workflows at scale, the MIT report found. Seven out of 9 sectors — the exceptions are technology and media — show no real structural change.
In each case, AI exposes bottlenecks that are organizational. That in turn can focus C-suite attention on the areas where the technology can have the greatest impact. Most companies are seeing benefits from using AI in compliance, risk management and quality control, according to a report by ISG. As for fueling growth and reducing costs? Only around 1 in 4 AI initiatives is meeting the C-suite’s expectations for revenue impact.
“The real issue is not just whether a company can use AI. It’s whether they know where AI should be used, and where it shouldn’t,” Schein said. He added that the companies that will get the most value aren’t applying AI in a spray-and-pray fashion. “They’re using it where it creates real leverage and avoiding it where a simpler, cheaper, or more deterministic approach already works,” he said.
It’s Not Just Cost
To dig deeper, we asked executives which specific barriers they think are limiting AI performance at their firm. Perhaps unsurprisingly, issues related to company data quality, availability or fragmentation came up the most often, cited by 63% of executives. This reflects the challenges of using AI across databases and other sources of information that are not always interoperable or in compatible formats.
Budget, time and other resource constraints ranked second, cited by 49% of executives as limiting AI performance at their firms. But interestingly, about the same share named governance, risk and approval processes (48%) and lack of clear ownership and accountability (46%).
Just 30% of organizations are redesigning key processes around AI and fewer than 4 in 10 report using the technology “at a surface level, with little or no change to underlying business processes,” a report from Deloitte found in January.
Company decisions often come down to cost and return on investment targets. But it’s significant that executives are equally focused on institutional pieces of the AI equation.
Nearly half (45%) of the executives surveyed by PYMNTS cited systems and workflows as a barrier for AI performance. Almost the same share pointed to internal skills and talent (42%) and to lack of leadership or alignment or sponsorship (40%) as holding back the positive impact of AI.
Zooming out, our data makes clear that the enterprise AI story has shifted from “What can the technology do?” to “How can organizations roll out and use AI most effectively?” The companies that get disproportionate value won’t be the ones trying to put AI everywhere, but the ones that treat data, governance and ownership as the real AI stack — and build the muscle to deploy it consistently.
Read the April report: The Enterprise AI Benchmark Report
At PYMNTS Intelligence, we work with businesses to uncover insights that fuel intelligent, data-driven discussions on changing customer expectations, a more connected economy and the strategic shifts necessary to achieve outcomes. With rigorous research methodologies and unwavering commitment to objective quality, we offer trusted data to grow your business. As our partner, you’ll have access to our diverse team of PhDs, researchers, data analysts, number crunchers, subject matter veterans and editorial experts.
