Simon Frost

In procurement, agents can analyse spend, research markets, discover suppliers, prepare sourcing events, compare responses, create and monitor contracts, track performance, identify risks, support negotiations and create reports.

• The practical questions are what agents existing systems already provide, what to buy versus build, what context and permissions an agent needs, and where autonomy stops and human oversight starts.
• Frost warns that putting an agent on top of a broken process does not fix it.

Frost's worked example is a request to find five potential glass bottle suppliers in Europe. The agent plans what data it needs, searches databases and websites, assesses suppliers against the request, rejects poor fits, repeats the search and delivers a shortlist with supporting details.

Read original

They apply it to strategy, meetings, product reviews, contracts, factory visits and crises.

• In contract management, AI can analyse, search, summarise, draft, monitor, automate and upload; people remain responsible for judgement, risk, relationships, negotiation and decisions.
• Frost argues that AI can help manage contracts but cannot replace an understanding of contract fundamentals.
• Without that understanding, people can sign away rights, misunderstand obligations, leave value on the table, expose the business to risk or lose a face-to-face negotiation.

Frost says AI can compare a contract against preferred positions, flag obligations and expiry dates, analyse legacy contracts and identify risky clauses. Their counterexample is someone who does not understand contract fundamentals and therefore signs away important rights or misses obligations.

The distinction starts as a general AI-versus-human contrast, then narrows to contract management, where Frost names the tasks AI should handle and the decisions people should keep.

Read original · Original 2 · Original 3

The audit method is to collate and catalogue contracts, then check end dates and commercial terms.

• Common failures include buying materials from weak suppliers without a signed contract, unfavourable direct and consequential loss clauses, contracts without SLAs, large companies losing contracts in emails, negotiations run by people without contract training, and unclaimed rebates.
• Frost also discusses when an SME should build a simple contract lifecycle tool versus buy proper CLM software, with a rough threshold of fewer than 20 contracts for the simple build.
• A separate co-created post with lawyer Miles Barnes, a partner at Howes Percival, explains termination clauses; it presents the view as co-created, not Frost's alone.

Frost describes a client recovery of more than £1M from a contract audit. The hard cash came mainly from collating and cataloguing contracts, checking end dates and commercial terms, recovering unclaimed quantity rebates, recalibrating price cards to the current market, renegotiating supply that continued past contract end dates, correcting wrong batch quantities and activating service penalties.

Read original · Original 2 · Original 3 · Original 4

Fact-based negotiation reduces emotional hype; an index price is not the same as a supplier's actual buy price; specifications can often be relaxed when prices surge; larger batches can drive supplier efficiency if cash and storage allow; and spot buying probably should not exceed about 20% to cap upside risk.

On index price challenge, Frost asks suppliers what percentage of their cover is fixed by month, so increases can be aligned with actual exposure rather than headlines. On specification challenge, they suggest listing attributes such as variety, size, granularity and colour, setting them to the cheapest option and dialling up only as needed.

Read original

A poll of their audience returned 81% for category management skills versus 19% for category knowledge.

• They list hard skills that AI increasingly assists with, including spend and data analytics, value-stream mapping, supplier assessment, cost modelling, strategic options and recommendations.
• Human skills become more valuable: stakeholder engagement, commercial judgement, communication, influencing, curiosity, storytelling, leadership, critical thinking and relationship building.
• Frost stresses the fundamentals: understanding principles, giving AI the right context, asking the right questions, challenging answers, spotting errors and applying judgement.
• A career post extends the argument from Mars in 2003 through analyst, junior buyer, senior buyer, manager and director roles, predicting that the old triangle becomes a diamond as AI reduces analyst and junior-buyer headcount and pushes managers toward broader commercial oversight.
• The 81/19 result is a poll of Frost's own audience, not a representative survey.

Frost traces the traditional procurement ladder from analyst to director and argues that AI changes each rung: analyst numbers decline as research is automated, junior buyers decline as tail spend is automated, senior buyers are guided by agentic source-to-pay tools, managers oversee agents and spend more time on relationships, and directors must keep reassessing a fast-changing battlefield.

The poll post makes the skills-versus-knowledge argument directly. A later career post turns it into a forecast about role shapes and headcount.

Read original · Original 2

The offer is ten modules over ten weeks, including contract management.

• They also mention a half-day contract-management masterclass.
• Access is a recurring priority: Frost offers regional discount codes for people outside Western Europe and North America where typical salary bands are lower, and describes a vision of world-class procurement training for everyone globally.
• The promotions follow substantive posts rather than replacing the teaching.
• The Academy dates differ across the posts; ask for the current schedule.

Frost says lawyer Miles Barnes co-teaches the contract management module. They also offer a discount code to people in regions with lower salary bands, making the training available beyond Europe and North America.

The posts give different Academy start dates, indicating multiple intakes or a schedule that changed across them. Ask for the current timetable.

Read original · Original 2 · Original 3 · Original 4

Frost answers the question with a concrete sourcing example.

Frost defines an AI agent as something that pursues a goal—planning, acting, checking, adapting and repeating—rather than simply answering a question.

• They give a worked example of finding five glass bottle suppliers in Europe and list ten procurement tasks agents can handle, from spend analysis to contract monitoring.
• They also raise questions about build versus buy and human oversight.

For brands planning a campaign

A ready-made structure for a co-branded webinar on AI agents: what they are, procurement use cases, build-versus-buy decisions, data and guardrails, and limits on autonomy.

Frost lists common contract problems: tens of thousands of tons of material bought with no signed contract; high-risk goods with unfavourable clauses; contracts referencing performance but missing SLAs; contracts lost under desks or in emails; negotiation by untrained staff; and unclaimed rebates.

• They describe collating and cataloguing contracts, then checking end dates and commercial terms to recover over £1M.
• Examples include unclaimed quantity rebates, price cards calibrated to last year's market, supply continuing past contract end, wrong batch quantities, and unactivated service penalties.

For brands planning a campaign

A compelling case study for a co-branded webinar on contract management: it shows the financial value of basic contract hygiene and gives Frost a credible, numbers-led teaching moment.

Frost argues that AI handles the mechanics, but human understanding prevents costly mistakes.

Frost lists what AI can do with contracts: compare them against preferred positions, flag obligations and expiry dates, analyse legacy contracts, prepare for disputes, draft from scratch, and identify risky clauses.

• But without understanding the fundamentals, people can sign away rights, misunderstand obligations, leave value on the table, expose the business to risk, or get eaten alive in face-to-face negotiation.
• They mention co-teaching Module 4 on contract management with lawyer Miles Barnes, a partner at Howes Percival, and offer a half-day masterclass.

For brands planning a campaign

Frost already co-delivers contract management training with a lawyer, offering a working model for co-branded webinars and expert collaboration.

Frost reports a poll showing skills beat category knowledge, then explains how AI shifts the balance.

Frost shares a poll in which 81% chose category management skills over category knowledge.

• They argue that AI can close knowledge gaps quickly, making human skills—stakeholder engagement, commercial judgement, communication, influencing, curiosity, storytelling, leadership, critical thinking and relationship building—more valuable.
• They also stress the fundamentals and the need to challenge AI outputs.

For brands planning a campaign

A thought-leadership angle on the skills that matter in an AI-enabled procurement team. It could frame a webinar around what AI cannot replace and promote the Procurement Adventurer Academy.

It offers a quick reference for setting expectations.

Frost contrasts AI verbs (automates, detects, calculates, predicts) with human verbs (judge, trust, sense, gauge, smell, taste).

• AI-suited tasks include spend analytics, risk monitoring, RFX automation, supplier discovery, contract management and workflow automation.
• Human-touch areas include supplier relationship management, stakeholder engagement, cultural understanding, change management, crisis management, ethical judgement, innovation, negotiation and leadership.

For brands planning a campaign

A strong visual framework for co-branded content, setting up a balanced discussion of human-AI collaboration rather than AI replacement.

Frost traces a procurement career that began at Mars in 2003 and progressed through analyst, junior buyer, senior buyer, manager and director roles.

• They predict the triangular org design will become a diamond: fewer analysts and junior buyers as AI handles research and tail spend; senior buyers guided by AI; broader, more commercial managers; and directors constantly reassessing.
• They say knowledge will become easier to access, human judgement more valuable, tech and AI literacy essential, and learning, unlearning and relearning critical.

For brands planning a campaign

Positions Frost as a guide to career adaptation, with material for audience development and a webinar helping procurement professionals navigate AI-driven role changes.

Frost outlines nine levers with brief explanations: should-cost modelling, index price challenge, specification challenge, network optimisation, phased price increases, give-and-get trades, spot versus contract, packaging and order optimisation, and demand challenge. Each includes a practical tip, such as using fact-based negotiation to reduce emotional hype or checking inventory levels.

For brands planning a campaign

Frost produces practical, list-driven training content that could be adapted into a downloadable resource or a workshop segment for a co-branded webinar.

They define an AI agent as goal-seeking rather than question-answering, using a supplier-search example: the agent plans what data it needs, acts across databases and websites, checks candidates against the supplier-search requirements, rejects poor fits, repeats and delivers a shortlist with evidence.

• They then map agent tasks across procurement: spend analysis, market research, supplier discovery, sourcing events, response comparison, contract creation and monitoring, performance tracking, risk identification, negotiation support and reporting.
• The operating questions are which agents existing systems already provide, what to buy versus build, what context and permissions an agent needs, and where the line sits between autonomous action and human oversight.
• They warn that putting an agent on top of a broken process does not fix it.

Definition, worked example, numbered questions and task inventory in a single post.

The post ends by asking the audience what question six should be, leaving the model open-ended rather than settled.

For brands planning a campaign

This is the strongest ready-made co-branded webinar spine for an AI-procurement brief: the definition, task map, buy-versus-build questions and autonomy debate give a partner two teaching blocks and a discussion prompt. A platform vendor can supply product evidence while Frost supplies the framing and audience questions.

Examples: Source 1

They apply it to strategy (gauging appetite for change), meetings (reading body language), product reviews (tasting samples), contracts (trusting the other side), factory visits (seeing how clean a site is) and crises (being in command).

• In contract management, they list what AI should do — analyse, search, summarise, draft, monitor, automate, upload — and what people remain responsible for — judgement, risk, relationships, negotiation and decisions.
• They also argue that AI can help manage contracts but cannot replace understanding their fundamentals, warning that weak contract knowledge can lead to signing away rights, missing obligations or losing value.

Two-column contrasts repeated across contexts.

The split is a strongly stated position, not a tested framework. Ask for evidence from controlled trials before presenting it as one.

For brands planning a campaign

A partner can build a webinar around the handoff between AI contract tooling and human judgement, using Frost's two-column framing. The strongest angle is not 'AI will replace buyers' but 'AI frees time for the higher-risk decisions', which suits a training or advisory brand.

Examples: Source 1 · Source 2 · Source 3

One post describes a contract audit that found over £1M for a client through steps such as collating contracts, cataloguing them, checking end dates and commercial terms, then recovering unclaimed quantity rebates, recalibrating price cards, renegotiating supply running past contract end dates and activating service penalties.

• They list common failures: material bought from weak suppliers with no signed contract, unfavourable loss clauses, contracts with no SLAs attached, large companies losing contracts in emails, negotiations run by people without contract training, and unclaimed rebates.
• A co-created post covers termination clauses, and another covers when an SME should build a simple contract lifecycle tool versus buying proper CLM software.

Case story, failure checklist, co-created specialist post and build-versus-buy advice.

The £1M recovery is a client anecdote from Frost's own post. Ask for substantiation and permission to use it as a client example.

For brands planning a campaign

Contract management is the natural second half of a co-branded webinar with an AI-agents first half. The audit story gives a concrete before-and-after, and the co-created clause content shows Frost already works with legal specialists, so a brand can plug in a legal or CLM partner without changing the format.

Examples: Source 1 · Source 2 · Source 3 · Source 4

Each lever has a short how-to and an insight note — for example, that index is not the same as a supplier's actual buy price, that specifications can often be relaxed when prices surge, that larger batches can drive supplier efficiency if cash and storage allow, and that spot buying probably should not exceed about 20% to cap upside risk.

Numbered levers, each with action steps and a distinct insight.

The 20% spot-buying guidance is a rough rule of thumb, not a modelled threshold.

For brands planning a campaign

A training partner can use this as a ready module or handout for a cost-inflation webinar. The format is practical and source-linked, so co-branding mainly means adding partner examples and a live worked case.

Examples: Source 1

A poll of their audience returned 81% for skills versus 19% for knowledge.

• They list hard skills that AI increasingly assists with — spend and data analytics, value-stream mapping, supplier assessment, cost modelling, strategic options and recommendations — and human skills that become more valuable — stakeholder engagement, commercial judgement, communication, influencing, curiosity, storytelling, leadership, critical thinking and relationship building.
• A career post traces the traditional path from analyst to director, starting with Frost's own 2003 start at Mars, and argues that the old triangle will become a diamond as AI reduces analyst and junior-buyer headcount and pushes managers towards broader commercial oversight.

Poll result, two-column skills list and a career-ladder retrospective.

The 81/19 poll reflects Frost's self-selected followers, not a representative survey. Ask for representative data before using it as market research.

For brands planning a campaign

This gives a thought-leadership partner a future-of-procurement narrative for a webinar or report, especially if the brand sells training or capability tools. The skills-versus-knowledge split is a debatable position that can carry a live debate.

Examples: Source 1 · Source 2

Frost also mentions a half-day contract-management masterclass and regional discount codes for people outside Western Europe and North America where salary bands are lower. The promotions sit at the end of substantive posts rather than replacing the teaching.

Substantive posts with a clear call to action, plus a region-sensitive discount offer.

The Academy dates differ across Frost's posts, so they describe more than one intake. Ask for the current schedule before promoting it.

For brands planning a campaign

A brand assessing a thought-leadership partnership should note that Frost already sells training and has an established funnel. A co-branded webinar can work as a top-of-funnel asset if the partner agrees how it relates to the paid Academy and whether they appear in module content.

Examples: Source 1 · Source 2 · Source 3 · Source 4

One post is a weekend recap about the tech on show, conversations, speakers and new connections, with the message that the tech revolution is in full flight.

• Another promises an unofficial translation of 30 procurement terms encountered at the event.
• A third references DPW while arguing that agents were already everywhere, in conversations with software providers and in presentations.

Conference recap and planned glossary.

Frost's posts include only the introduction to the glossary post. Ask for the full term list before planning to use it.

For brands planning a campaign

Event-based co-branding around DPW is a viable route, and a jargon-busting piece suits a training partner. However, the promised 30-term glossary is not present in Frost's posts, so that asset would need new material rather than editing an existing draft.

Examples: Source 1 · Source 2 · Source 3

Frost uses direct questions to readers, sometimes polls, and asks people to repost or follow.

• They end most posts with 'Frost Procurement Adventurer'.
• One post explicitly asks readers not to reply using AI, reinforcing the human-judgement theme.
• The structure makes each post skimmable and teachable, and the same format appears across AI, contracts, cost and career topics.

Numbered lists, arrow bullets, insight callouts, direct questions, polls and calls to action.

The format is consistent, but the posts are not edited as a series; some cross-references depend on other posts, including the missing 30-term list.

For brands planning a campaign

A partner can expect co-branded content to be structured and practical, with clear takeaways and a question for the audience.

• A co-branded poll or joint list post fits naturally.
• The no-AI-comments request signals how Frost wants engagement to work.

Examples: Source 1 · Source 2 · Source 3 · Source 4 · Source 5

Ask whether they own the program or hold a formal role.

View example

View example

They co-created a contract management post with Barnes and run a Procurement Adventurer Academy module with him.

View example

They described the event’s energy and translated its terminology.

View example

Taylor also focuses on the gap between the contract and the invoice, checking whether what is being paid still matches what was agreed—the same value-recovery logic as their audit story.

Taylor writes from the perspective of a contract technology and audit service, describing a named product and a customer case with dollar figures.

• Frost writes as a trainer and practitioner, offering numbered steps and a checklist readers can use themselves, and pointing to training at the end.
• Their examples come from retail and manufacturing; Taylor's come from finance and healthcare invoice review.

Read original · Original 2

Hill's distinction between AI for procurement and procurement for the AI era sits close to Frost's split between what AI automates and what people keep. Both also sell training built on that argument.

Hill's frame is construction: tenders, subcontractors and technology vendors competing for work packages that used to sit elsewhere.

• Frost's examples come from category management, contracts and cost inflation across sectors.
• Hill's posts ask what the buying decision itself is becoming, including data and IP ownership and supplier dependency; Frost's ask what procurement teams should do differently, and teach contract and category skills.

Read original · Original 2

Both write for people working in procurement rather than a general audience.

Jenks curates other people's observations and credits the speakers, making the posts event notes rather than a framework of his own.

• Frost develops recurring arguments across posts, including AI agents, the AI-and-human split and contract audits, and links them to training.
• Jenks's posts are shorter and less prescriptive.

Read original · Original 2

The posts make a specific, testable argument: compliance fails when contracts live in scattered inboxes and shared drives, so centralisation, obligation tracking at signature, standardised clause libraries and a named owner per contract are the structural fixes.

• Taylor describes an operating method a partner could co-present: an independent third party sits across procurement, AP and ERP to connect contract to invoice to payment.
• Rather than auditing everything at once, it tests rebates, payment terms, rate caps or pricing tiers across every contract and invoice.
• The posts offer an unusually concrete example for a name-search comparison.
• A national retailer ran more than 900 supplier agreements through contract review, found that over half its suppliers were paid earlier than the contract required and that terms ran 30 or more days shorter than peer benchmarks, and released $22M in working capital with $2.8M or more in interest savings.
• For a finance-focused webinar needing a contract-management segment with a quantified case, this is a stronger platform than teaching-first content: the argument comes with numbers and a repeatable audit design.

The content is written from the sell side of a single commercial offer, and the posts name it.

• Advice on obligation tracking, AI-assisted review and audit-ready repositories points to a services-plus-technology partner model, with the retailer case credited to that vendor’s contract product.
• A co-branded session would therefore be product-led rather than neutral teaching—a different proposition from one a training or advisory partner is offering.
• There is also no agent material here: no working definition of an AI agent, goal-pursuit example, or buy-versus-build or autonomy debate, so a session covering AI agents would have no support from this creator.
• Nothing in the two posts shows a training curriculum, cohort or practitioner community.
• The posts also omit a follower count, per-post performance and commercial terms; ask for those before assessing reach and cost.
• For an audience of working procurement generalists rather than contract-compliance and finance specialists, the range is narrower than a campaign aimed at that audience requires.

Examples: Source 1 · Source 2

The posts draw that line cleanly.

• AI for procurement means using tools to execute existing tasks better, such as research, tender analysis, contract review, supplier assessment and spend analysis.
• Procurement for the AI era is different because the buying decision itself changes: not supplier A against supplier B, but buying technology, outsourcing a capability, partnering, building internally, procuring a managed service or buying the outcome, with questions about data and IP ownership, where risk sits and how dependent the business becomes on the provider.
• That frame would give a co-branded webinar its debate segment, and a partner could reuse the distinction in its own course materials.
• The second post adds a construction-specific market-intelligence argument: estimating, procurement and commercial teams sit on the boundary with the supply chain and are first to see new suppliers, delivery methods and technology vendors displacing traditional subcontractors, but that signal is never recorded systematically; AI can now read and connect that qualitative pattern across dozens of tenders.
• Hill also states a training offer focused on the second, strategic piece, which maps directly onto a training partner’s needs.

The posts offer no support for the contract-management segment.

• There is no contract clause material, termination or audit content, or agent definition or worked agent example, so a session split between AI agents and contract management would need to source both halves elsewhere and use Hill for strategic framing only.
• The framing is construction-anchored: the buyer decisions discussed sit in contractor and supply-chain work packages, sharpening the message for a construction audience while narrowing its fit for a cross-industry procurement training partner.
• Two posts provide a thin basis for judging a partnership, and they are position pieces rather than demonstrated teaching: there is no exercise, prompt, checklist or worked example a reader can run.
• The training offer appears only as a closing line, so ask what the collaboration would actually sell.
• Request a follower count, engagement data, rate card and sponsorship policy before assessing reach and terms.

Examples: Source 1 · Source 2

The posts are field reports rather than essays: takeaways from a procurement conference and a partner note from a spend-analytics event after the day’s talks.

• That format suits a partner seeking event-side co-branding, a conference recap or a panel moderator, and it gives the partner usable lines to quote.
• The reported points are specific and attributable: make the compliant way the easy way, because people will not follow a process that is harder than the shortcut; calculate the cost of inaction alongside ROI; embed supplier relationship management into the procurement process rather than treating it as a sideshow; dirty data makes AI investment worthless; supplier knowledge is the basis of resilience; procurement should own its spend, supplier, category and contract data; bring insights to users or adoption lags; and recognise that change management and user experience make or break a technology rollout.
• A webinar opening with what buyers in large-enterprise procurement are actually saying would gain real texture from these points, and the creator’s event presence is itself a distribution asset.

Much of the thinking is relayed from and attributed to named speakers and panelists at other companies.

• That works for a recap but offers little basis for a co-presented teaching session with a named author on the agenda.
• There is no contract-management depth, agent definition, framework or training product, so the partner or other guests would need to supply both halves of a webinar on AI agents and contract management.
• The posts also show an existing relationship with a software vendor, which affects the neutrality calculation and may limit which partners this creator can appear with.
• Two posts from a single month do not establish sustained output or audience reach; request follower, registration and per-post performance data, along with commercial terms, before booking.

Examples: Source 1 · Source 2