Ali Oztan

Cost and lead-time optimisation typically come later, once constraints appear.

• By then, RFQs are out, purchase orders are issued and tooling may already be in progress, making changes expensive.
• Oztan proposes modelling cost during design, so it informs decisions rather than serving as after-the-fact analysis.
• Engineers can then see what a choice saves or costs before parts go to suppliers.
• The approach works only if modelling is fast and effortless.
• Oztan describes that visibility as what they wished they had had as a design engineer.

The post lays out the sequence that makes later savings hard: a compliant first version ships, RFQs follow, then POs, then tooling. Cost modelling has to happen early, not after that window has closed.

Read original

In the design post, a slow or manual process simply does not happen.

• The margin post says procurement teams rarely have time to market test supplier pricing or build cost models, so parts sourced for years quietly get more expensive.
• The negotiation post recounts a personal failure: Oztan tried the method while leading procurement, but each model took two weeks or more and the supplier did not take them seriously.

Oztan sets out a four-step method: build cost models, give suppliers targets backed by those models, market test if suppliers will not move, then negotiate again or switch suppliers. The method failed in practice because each model took two weeks to build and a supplier brushed them aside.

Read original · Original 2 · Original 3

Suppliers know exactly which cost drivers make up their price and which have changed, while buyers often lack time to reconstruct that for every part they buy or judge whether a change justifies the ask.

• The capability Oztan describes identifies a part's cost drivers and compares the requested increase with relevant producer price indices over time, alongside a should-cost model, giving the buyer objective data.
• An earlier launch post makes a larger claim for the same technology: it is patent-pending, can predict true cost drivers in seconds, and is more accurate than every large language model and procurement tool available.
• That accuracy comparison is Oztan's own product claim.
• Ask for independent testing to assess it.

The 15% raw material request is the worked case: the buyer must establish which cost drivers moved, how they moved over time and whether that supports the increase before agreeing to anything.

Read original · Original 2

With manufacturing profit margins around 5%, that drift directly erodes profitability.

• Oztan calls it one of manufacturing's most common and least visible sources of margin leakage.
• This is the clearest statement in the set of why the other problems matter commercially.

The 5% margin figure, set against long-established sourced parts, is the example: no single decision looks wrong, and the damage appears as an unexplained squeeze rather than a loss event.

Read original

They call it the most useful product feedback loop the team built that year and note that agents hit friction before users do.

• The argument is that most procurement software was designed for humans clicking through screens, and the next generation needs to work for both humans and agents.
• This is the set's only position on software design rather than buying outcomes.

Agents run the same sourcing workflows as a buyer, hit friction earlier than human users, and help guide the design of the next version.

Read original

The cost-intelligence launch announces the capability and says customers are already validating quotes.

• The negotiation post closes by asking readers to message for a demonstration.
• The event post reports demoing Procurable AI at a First Resonance and Anthropic event and argues that manufacturing is being redefined, with a large opportunity ahead for the supply chain.
• These posts are shorter and more product-forward than the analytical ones.

The negotiation post closes with an invitation to see the playbook run by supplier, bill of materials or commodity on the reader's own parts. The argument ends with a demonstration, not a conclusion.

Read original · Original 2 · Original 3

Start here: it lays out the incentive argument underlying the rest of Oztan’s posts. Engineers are rewarded for speed, so cost becomes a problem only after it is expensive to fix.

Oztan argues that most design engineers are incentivized to move quickly: get a first version out that meets form, fit and function, and keep the program moving.

• Cost and lead-time optimization typically come later, once constraints appear.
• By then, they say, RFQs are out, purchase orders are issued and tooling may already be in progress, making changes costly.
• Their proposed fix is to model cost during design, so it becomes an input rather than an after-the-fact analysis and engineers can see what a decision saves or costs.
• They add a practical condition: cost modeling must be fast and effortless, because design teams under time pressure will skip a slow or manual process.
• The post positions Procurable cost intelligence as a way to align engineering, procurement and the business while cost is still an opportunity rather than a problem.

For brands planning a campaign

Useful opening framing for a collaboration on design-stage cost intelligence. Any content built on this argument has to address Oztan’s constraint: the tool must be fast enough to use, rather than treating cost visibility alone as the answer.

The clearest product-launch statement in the set.

Oztan opens with the problem: procurement loses billions every year because teams make decisions using inaccurate or non-existent cost data.

• They announce Procurable Cost Intelligence as patent-pending technology that predicts the true cost drivers of any good or service in seconds.
• The post claims accuracy that outperforms every large language model and procurement tool available, and says customers are already validating quotes instantly and finding savings across everyday products and components used in global manufacturing.
• It closes with the positioning line that this is how procurement becomes strategic.

For brands planning a campaign

The natural anchor for a launch-style piece. The post’s strongest claim is an unverified comparison of accuracy, so a partner publishing alongside Oztan should substantiate it or avoid presenting it as established fact.

The clearest walkthrough of a single negotiation scenario.

The post sets up a familiar exchange: a supplier says raw material costs have risen and asks for a 15% price increase.

• Oztan frames the challenge as information asymmetry: suppliers know which cost drivers make up their price and which have changed, while buyers often lack the time to identify the drivers behind every part they buy, track how those drivers have moved, or assess whether they justify the requested increase.
• The described capability identifies a part’s cost drivers, compares a supplier’s increase request against relevant producer price indices over time alongside a should-cost model, and gives buyers objective data before they negotiate.
• The closing line sums up the argument: “more information, better negotiations.”

For brands planning a campaign

Strong material for a negotiation-focused collaboration: it names a specific, testable mechanism — producer price index comparison plus a should-cost model — rather than offering generic advice. It also provides a clear before-and-after frame for demonstrating any similar tool.

The most candid piece in the set: it states the playbook, then describes their personal failure to make it work.

Oztan writes that every manufacturer wants to save money and few have a game plan, then offers one: build cost models, give suppliers targets backed by those models, market test if suppliers will not move, then negotiate again or switch.

• They say they tried this while leading procurement, but it failed: each model took two weeks or more, and the supplier did not take them seriously.
• The post then announces Negotiate in Procurable, which they say runs the whole game plan by supplier, by bill of materials or by commodity.
• It ends with an invitation to message for a demonstration on the reader’s own parts.

For brands planning a campaign

Provides both the method and the credibility story: an admitted failure that explains why the software exists.

• That is more persuasive for a practitioner audience than a feature list.
• The direct-response call to action gives a collaboration a clear handoff to keep or replace.

The shortest statement in the set of the financial stakes behind Oztan’s other posts.

Oztan observes that procurement teams rarely have time to market test supplier pricing or build cost models, so parts and materials sourced for years quietly get more expensive.

• With manufacturing profit margins around 5%, that price drift directly erodes profitability.
• They call it one of the most common and least visible sources of margin leakage in manufacturing.

For brands planning a campaign

A compact executive-facing hook: it turns an operational habit into a profit-and-loss consequence, making it well suited to finance or operations leaders rather than procurement specialists.

It shifts from procurement economics to product design philosophy and offers the clearest statement of a forward-looking position.

Oztan describes stress-testing Procurable’s interface by having AI agents attempt real sourcing workflows end to end.

• They call it the team’s most useful product feedback loop that year, based on the observation that agents hit friction before human users do.
• The larger point: most procurement software was designed for humans clicking through screens, and the next generation needs to work for both humans and agents.

For brands planning a campaign

Best suited to a collaboration about agent-ready or automated procurement workflows. It is a weaker fit for conventional negotiation or cost-modeling content because it focuses on software design rather than buying outcomes.

One post frames the problem as billions lost each year to decisions made on inaccurate or nonexistent cost information, and introduces Procurable Cost Intelligence as patent-pending technology that predicts the true cost drivers of a good or service in seconds.

• A related post names the underlying problem as information asymmetry: suppliers know which cost drivers changed and by how much, while buyers rarely have time to reconstruct that picture for every part.
• They describe a capability that identifies a part's cost drivers and compares a supplier's requested increase against relevant producer price indices and a should-cost model, giving the buyer objective numbers for the negotiation.

For brands planning a campaign

Strongest angle for a manufacturer or procurement software brand: joint content that walks through how a should-cost model is built for one real part and what changes in the negotiation once the buyer has the cost drivers. Oztan's posts stay at the level of principle; a collaboration could provide the worked example.

Examples: Source 1 · Source 2

The proposed fix is to model cost during design, making it an input rather than an after-the-fact analysis and letting teams quantify the financial impact of design choices before parts go to suppliers.

• They add a practical condition: this only works if cost modeling is fast and effortless, because design teams under time pressure will skip anything slow or manual.
• The post closes with a personal note: this is the visibility they wished for while working as a design engineer.

For brands planning a campaign

Useful for reaching engineering and product audiences, not just procurement. A collaboration built around design-stage cost tradeoffs, ideally with a before-and-after example, fits how Oztan already frames the problem and broadens the conversation beyond the sourcing team.

Examples: Source 1

Oztan lays out a four-step method: build cost models, give suppliers targets backed by those models, market test if they will not move, then negotiate again or switch suppliers.

• The same post records that they tried this while leading procurement, but it failed: each model took more than two weeks, and the supplier did not take the models seriously.
• The experience explains why speed and credibility matter in the tooling they now build.

For brands planning a campaign

The failure anecdote is the most usable content hook here. A brand collaboration could focus on what made the manual approach collapse and what a faster model changes, making it a better fit for a demo or case study than a general thought-leadership piece.

Examples: Source 1 · Source 2

With manufacturing profit margins around 5 percent, this drift erodes profitability directly. They call it one of the most common and least visible sources of margin leakage in manufacturing—a clear statement of why the other topics matter commercially.

For brands planning a campaign

A quantitative angle: help a brand build content that estimates the cost of price drift on a typical bill of materials. The 5 percent margin framing gives a manufacturer-facing audience a reason to read past the first paragraph.

Examples: Source 1

They call it the most useful product feedback loop of the year and note that agents hit friction before users do. The larger argument is that most procurement software was designed for humans clicking through screens, and the next generation needs to work for both humans and agents.

For brands planning a campaign

Best fit for a technically oriented manufacturing or AI audience. A collaboration could explore what agent-run workflows reveal about interface design in procurement tools, a subject few creators cover from inside a working product.

Examples: Source 1

One post from the First Resonance and Anthropic event argues that manufacturing is no longer old school and is being redefined, with a big opportunity ahead for the supply chain. These posts are shorter and more promotional than the analytical ones, and tend to end with a direct invitation to message for a walkthrough.

For brands planning a campaign

Oztan already produces announcement-style content tied to product moments and events, making a launch-window collaboration a natural fit. Expect a product-forward tone rather than a neutral industry essay.

Examples: Source 1 · Source 2 · Source 3

They also refer to “what we're building at Procurable AI.”

• The posts show a working role at the company but don't state a job title, ownership stake or contractual arrangement.
• Ask them to clarify their role and relationship with Procurable.

View example

The post presents this as an industry appearance, not a commercial partnership. Ask whether any paid or contracted relationship was involved.

View example

The post doesn't indicate a sponsorship, client engagement or other commercial tie. Ask whether any such relationship exists.

View example

Both argue that procurement should own the price question and that visibility into a vendor's costs or pricing can change the outcome of a negotiation.

• Etkin recounts a Carta deal in which the vendor said there was no more room, yet 21 percent more savings was still found.
• Oztan makes a similar case for bringing objective cost data to supplier conversations.
• Both use a single worked negotiation to carry the argument rather than making a general appeal for better sourcing.

Etkin covers software and AI spending, including token costs that land in COGS as a line item that moves daily; Oztan focuses on manufactured parts, raw materials and should-cost models. Etkin cites a named customer story with a quoted procurement lead, while Oztan describes a product capability in their own words and is more willing to admit a personal failure with the method.

Read original · Original 2

Toomey's point that buying AI has become the slowest, most expensive enterprise purchase because security reviews and extra stakeholders are treated as a bottleneck sits alongside Oztan's claim that procurement loses billions to decisions made with inaccurate or missing cost data.

Toomey focuses on process and platform: routing requests, pulling supplier and contract data, checking policy and moving work toward a decision autonomously.

• Oztan focuses on the economics of a single part or commodity.
• Toomey's posts are shaped by announcements and launches; Oztan pairs similar launch posts with longer analytical pieces that lay out a method.

Read original · Original 2

Vega's China+1 post argues that moving final assembly to Mexico, India or Vietnam does not move the supply chain behind it.

• Oztan makes a similar case when arguing that a supplier's 15 percent increase should be checked against cost drivers and producer price indices, rather than accepted or refused on feel.
• Both end with a structured method: Vega's questions about where components, tooling and raw materials actually come from, and Oztan's four-step plan of cost models, supplier targets, market testing and renegotiation.

Vega analyzes geography and concentration risk across countries; Oztan examines cost drivers within a single part or bill of materials. Vega's posts close with a promotion for a Mexico sourcing service, giving them an advisory tone, while Oztan's analytical posts lead to a product demonstration using the reader's own parts.

Read original · Original 2

The China+1 argument—that moving final assembly does not move the underlying supply chain, so a factory in Mexico may still depend on Chinese electronics and a supplier in Vietnam may still source tooling from China—gives a manufacturing audience a way to decide which China dependencies are genuinely critical and where alternatives should be developed.

• The Mexico post extends the same logic to nearshoring: when the peso strengthens, a strategy built on labor arbitrage looks different, while a business case built on proximity, lead times, inventory, USMCA access, and responsiveness survives currency and freight swings.
• That suits a campaign on supplier risk, sourcing footprint, or bill-of-materials visibility.

Ali Oztan's posts do not walk through a supplier price-increase negotiation, should-cost construction, or matching requested increases against producer price indices—the core of their argument.

• Oscar Vega explains where supply sits and what makes it resilient, not how to counter a specific 15% request.
• Both posts also point readers to a sourcing business Ali Oztan runs, so a software vendor would be borrowing a channel that promotes an overlapping service business.
• Follower counts, engagement, and partnership terms are unknown; ask for those before committing.

Examples: Source 1 · Source 2

They speak to how AI-native source-to-pay changes the operating model—routing requests, pulling supplier and contract data, checking policy, and moving work toward decisions autonomously.

• The second post adds a counterpoint that lands with a manufacturing procurement audience: buying AI is often the slowest, most expensive enterprise purchase because of security reviews and extra stakeholders, and the teams pulling ahead build visibility rather than skipping steps.
• That fits a collaboration on procurement process, autonomous sourcing workflows, or the friction of enterprise AI purchasing.

Neither of Ali Oztan's posts works through cost drivers or a should-cost example.

• Both are platform and category announcements rather than method teaching, so a collaboration built on the supplier price-increase scenario would need Oztan's kind of worked analysis instead.
• Both posts are tied to a current employer's product launch.
• Ask for audience and partnership data before comparing reach or format.

Examples: Source 1 · Source 2

The Carta story—an incremental 21% saving found after the vendor said there was no more room—turns on distinguishing a supplier that has genuinely hit its limit from one that is merely digging in, and why those situations call for completely different responses: pushing a maxed-out vendor burns goodwill, while not pushing one with room leaves money on the table.

• The AI-buying argument makes the same move at portfolio level: token spend moves daily, pricing changes every few months, and a signed model can be deprecated before contract end, so procurement should jointly own what is consumed, what will be consumed next quarter, and what should be paid.
• That suits a spend-management, negotiation-data, or AI cost-governance collaboration and reaches finance and engineering as well as procurement.

The examples are software and SaaS purchases, not manufactured parts, so they do not cover raw materials, tooling, or producer price indices.

• Both posts read as vendor and customer-story marketing for a spend platform, making comparison with Ali Oztan's should-cost teaching uneven: the theme of negotiation leverage is shared, but the object of spend and depth of method are not.
• Only two posts were available, dated late September and early October 2026; ask for more examples to assess recurring themes.

Examples: Source 1 · Source 2