Three executives gave three answers when a buyer asked who owned the pricing model

The reason the pattern holds is that every answer in that room is true from where the person sits, and no document contradicts any of them. Hawkinson wrote that ambiguity about who owns a decision is survivable internally, because people route around it, and that under outside scrutiny it becomes a finding, and by then it cannot be fixed ^[8]. What failed in the cases he reviewed was the authority structure around the models, and nobody could answer a question from outside the company on a deadline ^[7].

His test for who the accountable person is rules out the two candidates a governance document usually lists. "Not the system owner or the team that built the model, but the executive accountable for what the output causes," he wrote ^[14].

The figures he offers for what naming an owner changes come from companies he does not name. At a $2.5 billion global manufacturer, CRM adoption sat at 6% for years. The standard explanation was that people needed more training. Employees knew how to use the system and did not want to, and sales managers did not enforce it because no one could explain why it mattered ^[15]. Adoption reached 89% after named owners by region, documented decision rights over the forecast and a platform those owners could run themselves, once the forecast leadership relied on had to originate in that system ^[16]. That is a gain of 83 percentage points ^[21].

The same absence showed up as a metric dispute at a private-equity-backed global manufacturer. Roughly a quarter of the data domains had no business owner, because nobody had assigned the responsibility ^[10]. Operations reported 94% on-time delivery on a plant calculation and commercial reported 68% on a customer calculation, a spread of 26 points, and nobody owned which number was right ^[11][22]. Naming an owner for that calculation ended the argument ^[12].

At the $5 billion industrial manufacturer, the environmental health and safety model was built to flag conditions that could lead to life-altering injuries. It produced results none of the team expected. Because accountability was already assigned, the short project comparing assumptions against the data took a phone call. Hawkinson's estimate of the alternative is a month of meetings about who was allowed to stop the model ^[18].

What the record does not contain is a named company or an independent count. The pricing engine is a composite. Hawkinson says he saw the same structure inside a $5 billion industrial manufacturer and a $6 billion global food company before he saw it raised in a diligence room ^[6]. He describes his own vantage point as building data and AI functions at companies between $1.5 billion and $6 billion, and sitting across from boards as a chief digital and AI officer ^[20]. The part that travels beyond his casework is the timing. He wrote that override and escalation questions read as administrative until a regulator, a litigant or a buyer asks them on a deadline ^[19]. In his example, the buyer asked eighteen months after the model started producing daily recommendations ^[4].