← All articles
- Market Intelligence
- Category Management
- Supply Chain Risk
- Inflation Monitoring
Signal and noise: you are the integration layer
The synthesis happens in your head, between meetings.
By EXOS Research · · 9 min read
The synthesis happens in your head, between meetings
There is no system in your organisation that holds a current view of your categories. There is you.
The commodity tracker holds prices. The risk platform holds supplier events, if you have one. The trade press holds context. Your suppliers hold their own version of events, delivered with a purpose. Finance holds the budget assumption that was set nine months ago. None of these talks to any other. The place where they are reconciled into something a decision can be made from is a single professional, usually in the twenty minutes before the meeting where the decision is taken.
This is not a failure of diligence. It is an accurate description of where the integration layer currently sits in most procurement functions, and it is worth naming plainly because it is usually discussed as a personal time-management problem rather than a structural one.
The volume a category manager is asked to hold has genuinely increased
Three separate movements, each measurable, and they compound.
| Movement | What changed | Source |
|---|
| The supplier base is wider | Average supplier count rose 18% in a single year, from 75 to 92. Large organisations sit at around 120 | RS/CIPS Indirect Procurement Report 2025 (n=426, UK and Ireland) |
| Disruptions arrive together rather than in sequence | 82% report new tariffs affecting their supply chains, with 20–40% of supply chain activity touched | McKinsey — Supply chain risk pulse (December 2025, n=100) |
| The risk ranking changes faster than the review cycle | Inflation concern fell from 59% to 41% between Q1 and Q2 2025 as availability and political risk displaced it. By 2026, US protectionism had left the top three for the first time in over a year | CIPS Pulse Q2 2025 · CIPS Global State of Procurement & Supply 2026 |
Two of these deserve a note.
The supplier-count figure is the most direct evidence for the bandwidth argument, and it moved in the opposite direction to intent: half the same respondents said they intended to consolidate. Intention and outcome diverged inside one year.
The reordering point is the one most often missed. Anything established as a risk picture at the start of a year has been reordered by the middle of it. That is an argument for monitoring rather than for periodic review, and it is a stronger argument than any single price figure, because it is about the rate of change rather than the level.
The proportion of what a procurement professional reads that is worth reading has fallen
This part we cannot put a number on, and will not pretend to.
The observable position is that the volume of published procurement and supply chain content has grown considerably faster than the volume of original reporting or primary research behind it. A large share of what surfaces in a search or a feed is content marketing: the same handful of survey findings recirculated between vendor blogs, aggregator pages and statistics round-ups, frequently uncredited, occasionally altered in transit. Generative tools have lowered the cost of producing this material to approximately zero.
The practical consequence is that filtering has become a larger share of the work than reading. A figure now has to be traced before it can be used: named survey with a stated sample, or vendor page citing another vendor page. That check takes longer than absorbing the finding itself. Time spent on provenance is time not spent on the category.
The cost of a missed signal is a weaker position, not a missed insight
A signal that gets past a category manager does not present itself as a gap in knowledge. It presents itself as a contract concluded at the wrong price, on terms set at a moment when the supplier held better information than the buyer did.
That asymmetry is the actual exposure. Fewer than one in five companies in McKinsey's survey expect to pass more than 80% of tariff costs on to their own customers (McKinsey — Supply chain risk pulse 2025). That locates the balance in the margin, and by extension in the terms that were agreed before anyone modelled it.
Scale is available for the strategic end. McKinsey Global Institute's modelling across 23 value chains put disruptions lasting a month or longer at roughly one every 3.7 years, costing the average company in the region of 45% of one year's EBITDA over a decade (MGI — Risk, resilience, and rebalancing in global value chains). Sector variation is wide — consumer goods sits nearer 30%. The order of magnitude is not in dispute.
Headline inflation tells you nothing about your categories
Aggregate inflation can fall while a specific input rises, and in 2025 it frequently did. CIPS Pulse data tracked a growing number of spend categories expecting input price increases above 10%, up from three categories in the first quarter, with shipping and logistics leading for four consecutive quarters.
The demand signal behind this is the clearest in the profession. In the RS and CIPS Indirect Procurement Report 2026, inflation and higher costs were named by 68% of respondents as a top challenge for the coming twelve months. The same survey series recorded 37% in its 2024 edition. Concern has not risen gradually; it has roughly doubled in two years within a like-for-like series.
Getting to a usable category view means holding several streams at once: producer prices for supplier-side pressure, commodity and energy indices matched to the actual cost drivers, wage data where labour is a real input, freight indices, regional differentials. Different publication schedules, different institutions, each requiring interpretation before it touches a decision. Across several categories and geographies, that is not a task a person completes. It is a task a person keeps up with, partially.
The reason this matters commercially is arithmetic rather than rhetorical. A supplier citing general inflation is making a claim about their own cost structure, and that claim decomposes. Where materials account for roughly 22% of total cost, a 15% rise in material prices justifies about 3% at product level, not 15%. Without the decomposition there is nothing to test the claim against. (The decomposition, set out in full)
Most of the tooling stops one step short
The monitoring products available to procurement generally produce a stream that ends at a screen. An alert fires, someone reads it, and whether it reaches the negotiation it bears on depends on that person remembering that the contract exists and carrying it across by hand.
That last step is never automated, and it is the one that fails. Not through carelessness — through the same bandwidth constraint that produced the problem in the first place. A risk alert that never becomes a negotiating position has cost money and changed nothing.
There is a related observation in McKinsey's 2025 work worth noting, because it says something about what moves organisations. Most companies still understand their supply chain risk only as far as tier one, and awareness at tier two and beyond fell in 2023 and 2024. The exception was tariff compliance, which drove a 22 percentage-point increase in tier-two visibility. The constraint has been priority rather than feasibility.
What would actually reduce the load
Not more feeds. The constraint is bandwidth, and additional inputs consume it.
What reduces it is conversion: continuous inputs turned into a small number of category-specific findings, each already attached to the decision it affects. Disruption signals mapped to the categories genuinely exposed to them rather than presented as market news. Leading cost indicators expressed as an effect on a specific contract rather than as a chart that has to be interpreted.
The difference is not the sophistication of the model doing the reading. It is whether the output arrives attached to something.
Where EXOS sits
EXOS is a scenario-based procurement analytics system: it models total cost of ownership, negotiation scenarios and supplier risk before a contract is awarded. Monitoring runs within it on a scheduled basis across a defined supplier and category list, in the same system that holds the scenarios, for the reason set out above — a signal that has to be transported by hand usually is not.
Specialist supply chain risk monitoring platforms map multiple tiers considerably more deeply than we do, and where that depth is the requirement they are the correct instrument. Our position is narrower: fewer feeds, connected to the deals they concern.
How EXOS relates to the other classes of procurement AI → · Six kinds of procurement AI, as we see them →
Sources:
Supplier base and inflation concern: RS and CIPS — Indirect Procurement Report, 2025 edition (n=426) and 2026 edition (n=448)
Risk ranking and tariffs: McKinsey — Supply chain risk pulse 2025 (December 2025, n=100) · McKinsey — Decoding disruption to reshape manufacturing footprints (January 2026)
Disruption economics: McKinsey Global Institute — Risk, resilience, and rebalancing in global value chains (August 2020)
Category price pressure: CIPS Pulse Survey Q2 2025 · CIPS — Global State of Procurement & Supply 2026