A distributor whose first AI project is pricing should give each kind of data an AI pricing engine reads one business owner and one monthly count, starting with inventory levels, replacement costs, competitor prices and customer buying history, before it names anyone to run the AI program.
Pricing leads distributors' AI priorities. The National Association of Wholesaler-Distributors published a guide on July 31, 2026 drawing on a study by the NAW MDM research team of over 400 senior leaders in distribution, a report four software companies sponsored. In that guide 27% of distributors ranked pricing as their number one AI priority.
The same guide describes AI pricing engines that ingest vast amounts of data, including current inventory levels, replacement costs, regional competitor pricing, and historical customer behavior. Those four are where the owners come from. The guide's word is including, so a distributor whose engine reads other inputs, its own pricing history for instance, gives each of them an owner too.
On where to start, in a session on data governance at Gartner's Data and Analytics Summit in London, summarized on Gartner's newsroom on May 11, 2026, a Gartner analyst said that D&A leaders, meaning data and analytics leaders, should start by identifying business outcomes first, and should focus effort on prioritized business outcomes to accelerate ROI and minimize the required effort. For the counts, the UK government's Data Quality Framework, published on 3 December 2020 with its ask to adopt directed at central government, draws heavily on the Data Management Body of Knowledge and DAMA UK's Data Quality Dimensions white paper, and says many of its concepts and approaches are broadly applicable. It defines uniqueness as the degree to which there is no duplication in records, and completeness as the degree to which records are present.
At a distributor the head of operations can own inventory levels and count the share of items at the month's cycle count whose system quantity did not match the shelf, while purchasing owns replacement costs and counts items whose cost predates the supplier's latest price notice. The pricing manager holds competitor prices; that count is completeness, the top-selling items with no competitor price recorded in the quarter. Sales owns customer buying history and counts uniqueness, the customers carrying more than one account number, because a duplicate account splits one customer's buying history across two records.
In the first month the CEO names the four owners, each writes down its count and the CFO's team takes the first reading. In the second, owners correct errors where they are entered, in the ERP or the price file, and the counts are taken again. By the third the four counts sit on the monthly operating review, and the CEO names the person who will run the AI program, who inherits four owners and three months of numbers. The owners' time comes out of budgets they already hold, and none of these sources prices it. An operating partner can ask every distributor in a portfolio for at least the same four names and four counts.
Data the engine does not read, such as delivery records, waits for a use case that reads it.
In Gartner's survey of 504 data and analytics executive leaders worldwide, run from September through November 2024 and published May 12, 2025, 70% of chief data and analytics officers had the primary responsibility for building the AI strategy and operating model.
