The Procurement Function Retail Has Been Ignoring Is About to Become a Competitive Weapon

Retail has invested heavily in AI for customer-facing applications: personalization, demand forecasting, dynamic pricing, inventory optimization. The back office has largely waited its turn, and the wait is creating a widening performance gap between retailers who have started building procurement intelligence capability and those who have not. The next meaningful wave of competitive separation will run through the CPO’s office.

Karoline Dygas, VP and Chief Procurement Officer at Nordstrom, offered a candid assessment at the Manifest 2026 conference in Las Vegas this February. Her team is using AI heavily within its procurement spend analytics platform, and the results are less about transformation than about time. Supplier research that once took hours now takes minutes. Category strategy work that required manual data aggregation now gets a running start from AI-generated synthesis. The gains are real, but Dygas was clear-eyed about the gap between where Nordstrom is and where she wants to go.

The distance between AI as a time-saver and AI as a proactive strategic advisor is the defining tension in procurement technology right now. How retail leaders navigate it over the next 18 months will have lasting implications for cost structure, sourcing resilience, and organizational design.

The Spend Visibility Problem Retail Has Never Fully Solved

Before getting to AI’s ceiling, it helps to be honest about the floor most retail procurement organizations are starting from. Indirect spend, meaning everything a retailer buys that does not end up on a shelf, is equivalent to 10 to 15 percent of sales on average, according to McKinsey research specific to the retail sector. Most of it is poorly managed. Fragmented ownership across departments, inconsistent vendor approval processes, and an almost total absence of category-level visibility mean that a significant share of that spend escapes procurement’s influence entirely.

The challenges are not new. Retailers have long contended with limited transparency into procurement workflows, siloed not-for-resale category management, and few incentives at the business unit level to reduce indirect costs. What has changed is the scale of opportunity being left on the table. Retailers that treat indirect costs as a business transformation lever rather than a procurement matter alone can boost return on sales by as much as two percent, McKinsey found, and the potential holds even at retailers that have been working on cost reduction for years.

AI-powered spend analytics tools address the visibility deficit directly. They classify transactions, normalize supplier names across disparate systems, flag anomalies, and surface consolidation opportunities that previously required armies of analysts or simply went undone. Deloitte’s 2025 Global CPO Survey, which drew on responses from 265 CPOs across 40 countries, found that procurement digital leaders, defined as those pairing advanced technologies with meaningful investment in talent, hit or exceeded their cost savings targets 96 percent of the time, compared to 80 percent for followers. That gap is not marginal. It compounds across every category a procurement team manages.

Speed Is the Entry Price, Not the Prize

Dygas described the current value of AI in Nordstrom’s procurement operation in terms of speed and time saved, which is a fair characterization of where most organizations stand today. The question serious CPOs are asking is what comes next.

The EY Global CPO Survey 2025 found that 80 percent of global CPOs plan to deploy generative AI in some capacity over the next three years, with near-term focus on spend analytics and contract management. Yet only 36 percent of procurement organizations currently have meaningful AI implementations in place. The gap between intention and deployment matters because the competitive advantages from AI in procurement are not evenly distributed across time. Organizations building better data foundations now will have a structural advantage when more sophisticated capabilities mature.

The progression Dygas outlined at Manifest maps onto a maturity curve the broader procurement community is navigating. Descriptive AI, covering what happened and where money went, is where most organizations have started. Predictive AI, which uses historical patterns to forecast demand and likely cost shifts, is within reach for retailers with reasonably clean data. Prescriptive AI, which builds on those forecasts to recommend specific sourcing actions, is where the real ambition lies. Dygas put it plainly: she wants AI to surface what she needs to know, rather than requiring her team to formulate the right questions in advance.

Query-driven tools extend human capacity without changing how procurement operates. Proactively surfaced intelligence changes how teams are structured, what analysts are hired to do, and how procurement earns its seat at the strategy table. An efficiency gain and a functional transformation are not the same investment, and the CPOs who treat them as interchangeable will find out the hard way.

The Data Governance Prerequisite Is Not Optional

Dygas was equally direct about the threshold any AI vendor must clear before Nordstrom engages: strong data governance, meaning accurate outputs, airtight security, and no hallucinations. That standard is not overcautious. Procurement decisions involve real financial commitments and supplier relationships with long-term consequences. A hallucinated benchmark price or a fabricated supplier risk flag does not just waste time; it distorts strategy and erodes the credibility procurement leaders have spent years building with finance and legal counterparts.

The risk is well documented. A 2025 EY study found that nearly all large companies deploying AI reported some risk-related financial loss, from compliance failures to flawed outputs, and that organizations with stronger responsible AI frameworks consistently outperformed on cost savings and operational outcomes. Governance, by that evidence, is not a constraint on AI investment. It is a condition for getting a return on it.

For retail specifically, the data challenge is compounded by organizational structure. Procurement spans multiple categories, including store operations, marketing services, technology, logistics, and facilities, each with its own legacy systems and data standards. Building a trusted, unified data source across all of them is painstaking work. It is also the work that determines whether AI procurement tools eventually deliver strategic value or remain expensive time-savers for a small team. According to Efficio’s 2026 procurement outlook research, the main bottleneck to scaling AI is no longer the technology itself but the fragmented and inconsistent data foundations that undermine model accuracy and organizational trust in the outputs.

For CPOs evaluating vendors, data governance is a selection criterion, not a feature to negotiate in the second conversation.

The Orchestration Gap

One of the more consequential observations Dygas made at Manifest was about the fundamental architecture mismatch between how procurement actually operates and how most software is built to support it. Procurement is not linear. A sourcing engagement involves parallel tracks, covering supplier qualification, commercial negotiation, internal stakeholder alignment, legal review, and compliance checks, all of which intersect, loop back, and get reprioritized based on what surfaces along the way. Most procurement tools are built for sequential workflows because sequential workflows are easier to design. The reality of the work is considerably messier.

Deloitte’s CPO survey identified this directly, noting a growing need among procurement leaders to build automated guidance through process orchestration, shifting procurement from a reactive function that responds to late-stage stakeholder requests toward one that actively routes the organization toward preferred sourcing paths and suppliers before situations become urgent.

Procurement process orchestration tools are attempting to close this gap, with the ambition of managing the non-linear reality of complex sourcing engagements: routing the right information to the right people at the right stage, flagging when a track is falling behind, and surfacing relevant supplier data when a negotiation pivots. Whether available tools are mature enough yet to deliver consistently is a legitimate open question. The framing, however, is correct, and retail CPOs managing complex category portfolios should be tracking this space closely rather than waiting for a finished solution to appear.

What the Performance Gap Looks Like in Practice

The Hackett Group’s 2025 Digital World Class Procurement research, drawn from global benchmark studies across hundreds of companies, quantifies what separates leading procurement organizations from the rest. Top-performing teams deliver 2.6 times greater ROI than peers, operate with 31 percent fewer full-time employees, and execute sourcing cycles 24 percent faster. They also generate twice the cost savings as a percentage of spend and lose 60 percent less in savings to maverick buying and contract noncompliance.

These organizations spend 1.8 times more on procurement technology, and their analysts spend 26 percent more time on actual data analysis rather than manual data collection. The investment logic is not complicated: better tools free skilled people to do more consequential work. The difficulty is that the payoff requires getting the data infrastructure right before the more sophisticated capabilities can function as intended.

For retail and CPG leaders, the practical implication is sequencing. The organizations pulling ahead are not attempting to skip from spreadsheet-level visibility to prescriptive AI in a single move. They are building clean data foundations, deploying analytics in the categories with the most complexity and spend, demonstrating measurable return, and using those wins to fund the next capability. The work is incremental and largely invisible until the advantage becomes undeniable.

What the CPO Role Is Becoming

The procurement function’s organizational position is shifting in ways that make these technology decisions more consequential than they might appear from the outside. Deloitte found that procurement’s influence has grown across finance, IT, and manufacturing and operations functions since 2023. Boards and executive teams are increasingly asking procurement to shape platform strategies, define data governance postures, and connect sourcing decisions to product, market, and sustainability outcomes. The function is being asked to carry more of the enterprise’s strategic weight, and the CPOs building the right capabilities now are the ones who will be positioned to carry it.

For retail, this elevation is particularly well-timed. Margin pressure is structural. Tariff volatility has made supplier diversification a board-level concern. Consumer expectations for sustainable and ethical sourcing have moved from marketing language into supply chain accountability. Procurement sits at the intersection of all of it. The CPOs who can translate spend intelligence into strategic positioning, walking into an executive committee meeting with forward-looking supplier risk analysis, proactive cost scenarios, and defensible data, will have influence their predecessors could not have commanded.

The technology path from descriptive analytics to prescriptive AI is not short, and the organizations investing in it now, through clean data, rigorous vendor governance, and a deliberate move from isolated pilots to embedded workflows, are building an advantage that will be very difficult to close from behind. Assuming a mature, turnkey solution will eventually arrive and make the foundational work unnecessary is a reasonable-sounding reason to wait. It is also how organizations fall two to three years behind.

Conversations On Retail

Conversations On Retail is a gathering place and resource center for retail and CPG executives, built to make it easier to stay current, discover the technologies and solutions shaping the industry, and connect with the people driving it forward.

We publish news, views, and reviews from staff editors, contributing experts, and trusted partners. Some articles are developed internally, while others are submitted by industry contributors or adapted from interviews and recorded conversations with industry leaders.