Case Study · Wholesale and Distribution

They came in with a $40M target. We found $56M.

A Fortune 500 wholesale and distribution group asked us to validate a savings estimate and build a plan to deliver it. Our assessment found 40% more than they were looking for, and $31M was realized in year one.

What they came in with

$40M

Savings target based on an internal audit. This was the number leadership had anchored on going into the engagement.

→

We found

What we identified

$56M

Addressable opportunity identified across 18 months. $31M realized in year one through AI-assisted spend categorization.

Fortune 500 Wholesale and Distribution Group · 18 months · 2023–2024 · Spend Intelligence · AI-Assisted Categorization

The Situation

A business that knew it was leaving money on the table. It just did not know how much.

At the scale a Fortune 500 wholesale and distribution group operates, spend is complex. Thousands of suppliers. Multiple regions. Categories that blur into each other. The organization had run an internal audit and arrived at a $40M savings estimate, a reasonable figure based on what their teams could see at the time.

The problem with internally generated savings estimates is that they are limited by what people already know to look for. When you run an audit manually, you look in the places you think the problems are. You apply the categories you already use. You measure the things you already measure.

Key insight

Categorization is not a step before insight, it is the insight. The extra $16M above the target was not hiding somewhere new. It was in the same data, seen properly for the first time.

They came to us to validate their estimate and build a plan to deliver it. What they got was a more complete picture of their spend, and a considerably larger opportunity than the one they started with.

The Numbers

Three figures that tell the whole story.

Target going in $40M

Based on internal audit

Savings pipeline identified $56M

40% above the original target found through AI-assisted reclassification

Realized in year one $31M

Through supplier consolidation and renegotiations made possible by accurate category data

How We Found It

We looked at the same data through three lenses. Each one showed something the last one missed.

Our assessment does not start with assumptions. It starts with the raw data and works outward. For this engagement, we ran three analytical lenses in sequence, each designed to surface a different type of opportunity.

Lens 1 🗂️

Spend Categorization

AI-assisted reclassification of spend that had been sitting in broad, inaccurate categories. When you categorize properly, the opportunities hiding in miscategorized data become visible for the first time. This is where the gap between $40M and $56M lived.

Lens 2 🔄

Supplier Consolidation

Cross-referencing suppliers across regions and categories to find duplication. At this scale, the same supplier often appears under ten different names and billing entities. Consolidating that spend creates immediate leverage in negotiations.

Lens 3 📐

Benchmark Comparison

Comparing category-level spend against industry benchmarks to find where the organization was paying above market. This required properly categorized data from Lens 1, without accurate categories, benchmarking produces noise, not signal.

The three lenses build on each other. You cannot run Lens 3 without Lens 1. And without all three, you will always undercount the opportunity, which is exactly what the original $40M estimate had done.

From Identification to Realization

Finding $56M is interesting. Getting $31M in year one is what matters.

We see a lot of assessments that produce a big number and stop there. The number goes into a presentation, gets reviewed by a committee, and slowly disappears without anything changing. We built this engagement to avoid that.

Months 1–3

Data preparation and spend reclassification

Before any analysis could happen, the spend data needed to be properly categorized. This is unglamorous work. It is also the foundation on which everything else rests. AI-assisted categorization processed the full transaction history at a speed no manual team could match.

Months 3–6

Opportunity sizing and prioritization

With properly categorized data, we ran the three-lens analysis and arrived at $56M. We then prioritized by ease of realization, not just size. Year one focused on what was actionable within 90 days.

Months 6–12

Year one realization at $31M

The first wave of savings came from supplier consolidation and renegotiations made possible by the benchmark data. AI-assisted spend monitoring kept the savings from eroding, a common failure mode most savings programs do not account for.

Months 12–18

Expanding the pipeline

The remaining $25M required deeper supplier negotiations and category strategy work. The AI system that categorized and monitored spend in year one became the ongoing intelligence layer for all procurement decisions going forward.

What We Learned

Why organizations almost always underestimate their own opportunity.

Internal audits are limited by internal knowledge

The $40M estimate was not wrong. It was incomplete. The team used the categories they had and looked where they knew to look. AI-assisted reclassification looked at everything, with no preconceptions about where the problems should be.

Categorization is not a step before insight, it is the insight

The extra $16M was not in a new place. It was in the same data, seen properly for the first time. Proper categorization does not prepare you for analysis. It is the analysis.

A big number needs a realization plan

Identifying $56M and realizing $31M in year one required two completely different things. The assessment told us where the opportunity was. The execution plan determined what was achievable quickly and in what order. Without both, you get a number that never moves.

Monitoring is what keeps the savings alive

Most savings programs realize their numbers and then watch them erode as old habits return. The AI monitoring layer means the spend intelligence keeps running after the engagement ends. It keeps watching, and it flags when things drift.