AI doesn’t automate markets – it automates transactions where human involvement adds little value. That shifts where suppliers in chemicals and energy will differentiate in the future: away from standard business toward risk, security of supply, and advisory. Learn from our Simon-Kucher experts how sales must realign in a world of agentic AI.
The debate about “bots buying from bots” is often framed the wrong way. It sounds like a future in which autonomous machines replace people in purchasing and sales. But that is exactly what will not happen in the short term. The real shift is more fundamental: with agentic AI, machines are beginning to reorder industrial markets. Such AI agents plan, evaluate, and execute commercial tasks on their own within defined guardrails.
What gets automated first is not companies or customer relationships, but the undifferentiated middle of the market: routine transactions where human involvement adds little value. In these cases, the purchasing process now costs more than the human decision within it is worth. Wherever products are standardized, prices transparent, and decisions data-rich, traditional procurement and sales processes lose their value. Everything that can be compared, structured, and assessed by rules becomes machine-readable.
This is precisely why chemicals and energy are such interesting leading indicators. Both industries combine standardized commodity logic with high strategic criticality. Both show where AI accelerates markets – and where human decisions gain value.
AI therefore does not simply automate sales. It redistributes their value creation and helps determine where profit will be generated in the future.
The fallacy: bots don’t replace people
The real change is not that bots will suddenly conclude contracts autonomously. Far more likely is a gradual shift in the logic of decision-making.
AI agents read tenders, structure requirements, compare offers, simulate risks, and build shortlists. In standard cases, they already trigger transactions themselves. In more complex cases, they initially shift the balance of power in the market. In the future, negotiating advantage will arise less from information asymmetry than from the ability to simulate, structure, and compare better.
The bot doesn’t buy first. In many markets, it will first decide which supplier gets to speak with a human at all.
This changes the sales function from the ground up. In the past, suppliers could also differentiate through relationships, speed of response, or bespoke offer logic. A world of agents tolerates that far less.
It asks, matter-of-factly:
- Is the specification met?
- Is supply availability verifiable?
- Is the risk transparent?
- Are pricing formulas comparable?
- Is CO₂ information machine-readable?
Anyone who cannot answer these questions in a structured way may lose not just the negotiation, but access to the shortlist in the first place.
Three market logics that AI pulls apart
The question is not which industry will be more heavily automated. What matters is which part of a market follows which logic.
1. The machine-readable standard market
The first market AI changes is the machine-readable standard market. It is dominated by standardized products, recurring needs, and decisions of low strategic importance. At the same time, the process cost of human decisions is often higher than their actual added value.
In chemicals, this applies above all to long-tail products, standard grades, and recurring procurement with clearly defined specifications and a high documentation load. In the energy sector, the same logic appears in standard tariffs, standardized power and gas products, simple SME tenders, or algorithmic trading.
In exactly these areas, systems can already obtain quotes, compare prices, evaluate suppliers, and trigger orders automatically. Not because AI fundamentally decides “better” than people, but because speed, consistency, and process cost count for more than individual negotiation. Human decisions grow rarer wherever they no longer create additional economic value. For suppliers, this means: anyone who continues to serve this business personally is paying for a sales effort the customer no longer needs.
2. The hybrid risk market
Some markets are highly automatable at the transaction level and yet not free of risk. Algorithmic markets already work for equities, oil, power, or metals such as copper and aluminum. Wherever products are clearly comparable, prices transparent, and markets sufficiently liquid, systems take over price discovery, tendering, and trading.
In part, this now also applies to core petrochemical building blocks such as benzene, toluene, acetone, or butadiene. The transaction itself can be largely automated there – from tender to order.
But this is also where full automation reaches its limits: companies do not buy just a molecule or a kilowatt-hour. They buy availability, security of supply, logistics, and the ability to stay operational in the event of a disruption. Especially in volatile markets or with concentrated supplier structures, it is therefore not the product alone that counts – the stability of the entire supply chain proves decisive.
3. The strategic transformation market
At the other end, markets are emerging in which the purchase itself becomes a strategic decision. Rather than price or availability, what matters more here is a company’s future competitiveness.
In chemicals, this involves alternative feedstocks, the circular economy, Scope 3 reduction, or new raw-material pathways. In the energy sector, it concerns PPAs, electrification, hydrogen, storage, flexibility, or long-term site supply. Such decisions influence far more than cost: production models, CO₂ profiles, investments, and security of supply over many years.
Here, suppliers no longer sell just a product or a tariff. They sell risk management, transformation capability, and economic resilience. Sales thus shift from transactional to strategic.
Paradoxically, high-quality commercial advice gains value in an increasingly automated world. The more complex and transformation-driven markets grow, the more it pays to be the one who makes uncertainty calculable. Those who can do this negotiate less about discounts and more about contract models, terms, and how risk is shared.
Commodity becomes code, trust stays human
Commodity becomes code wherever specifications, prices, ability to supply, and contract terms can be evaluated. Competition therefore shifts from relationships and information advantage toward how well a supplier prepares its data and how quickly it responds.
At the same time, the value of what cannot be fully standardized rises: responsibility, security of supply, and the ability to translate complex risks into robust decisions.
Sales therefore do not disappear. They move away from routine transactions and toward decisions where complexity, risk, and transformation matter. Trust stays human – but the evidence that supports trust must increasingly be machine-readable.
What companies must do now
The most important task is not “more AI.” It is about becoming machine-readable and distinctive at the same time in an increasingly automated market environment.
For executives and sales leaders in chemicals and energy, this means four things:
1. Create machine-readability
Products, prices, certificates, supply availability, and contract logic must be comparable and available digitally. Without structured data, a supplier remains invisible to agents.
2. Segment by automation potential
Not every customer, every transaction, and every sales step requires human interaction. Standard business belongs in scalable digital processes – human selling time belongs to risk, complexity, and transformation.
3. Quantify and monetize differentiation
Supply-chain resilience, flexibility, quality, CO₂ impact, and security of supply must become measurable and economically demonstrable. In a world of agents, assertion is no longer enough. Only when a supplier translates these capabilities into tiered service levels, contract options, risk-sharing models, and prices does additional willingness to pay arise.
4. Redefine sales
The sales function of the future does not primarily sell products. It helps decide which risks a customer bears themselves and which the supplier assumes. That is exactly why good sales become more valuable – not despite AI, but because of it.
Conclusion
Chemicals and energy are a textbook example of how industrial markets are changing. Agentic AI will accelerate standardization, increase comparability, automate transactions, and break down opacity. Especially with interchangeable products, recurring processes, and rules-based decisions, transaction execution shifts from people to systems.
At the same time, a new scarcity emerges: genuine commercial value. The future therefore belongs neither entirely to people nor entirely to AI agents. It belongs to companies that understand what can be automated, what stays hybrid, and where trust, risk, and transformation matter more than the product itself.
Because in the end, this holds true:
Commodity becomes code.
Trust stays human.
Would you like to find out which transactions can be automated and where commercial advice matters more? Then talk to our Simon-Kucher experts.
