Growth-tracking departments report strong AI returns at nearly 3x the rate of efficiency-only ones

The release says departments that measure AI against growth outcomes are "nearly three times as likely" to report strong ROI ^[3]. The ratio from its own figures is 2.7, which is 35% against 13% ^[15]. The remainder is the larger number: 65 of every 100 departments that track at least one growth measure still do not report strongly positive returns ^[16]. The bar for that group is also low, because a single measure qualifies, and the qualifying list includes customer satisfaction and output volume next to net-new revenue ^[4].

Whether the metric causes the gap is a separate question, and Wharton's own faculty describe a bundle. Prasanna Tambe, a WHAIR faculty co-director, said: "The companies seeing stronger returns are going further. They are redesigning workflows, putting clear ownership and governance in place, and measuring AI against outcomes that matter to growth." ^[14]

The release's other splits point the same way. Among managers and above, 58% at organizations with a newly designated Chief AI Officer report strongly positive ROI, against 8% where there is no such role or plan, a ratio of 7.25 ^[8][17]. At organizations standardized on one approved platform the figures are 50% and 9%, a ratio of 5.6 ^[9][18]. Both spreads are wider than the 2.7 on measurement. They are drawn from different bases (managers and above for the first, departments for the metric split), so they cannot be stacked.

We think the three splits are largely one variable: an organization mature enough to name an owner, pick a platform and count outputs, with the metric as a marker of that maturity. The counter-thesis is Tambe's, that choosing outcome metrics forces the redesign. The release text does not include group sizes or a before-and-after comparison of departments that changed metrics. We would be wrong if a cohort of departments added a growth measure and reported stronger returns a year later with owner, platform and workflow unchanged.

The allocation is where the gap shows. Departments scoring AI on hours saved outnumber those scoring it on net-new revenue by 13 points ^[19], and leaders who call cost reduction AI's main job outnumber those who say growth by 30 points ^[20]. Meanwhile 83% of leaders expect AI to transform or have a major impact on their industry within two to five years ^[10]. Tambe said: "Companies believe AI is going to transform their industries, but many are still asking it to prove itself by saving time." ^[13]