Beyond the AI Hype: What Practical Adoption Looks Like in the Lubricants Industry – Beyond The Blend LIVE!
Artificial intelligence is being discussed everywhere, but what is actually happening inside lubricant businesses?
That was the question at the heart of a live Beyond the Blend recording at Lubricant Expo Düsseldorf 2026. With thanks to the Lubricant Expo team for hosting us, Rob Taylor and Steve Knapp were joined on stage by Egor Parienko, Ian Lloyd and Rafe Britton for an honest conversation about AI adoption, customer behaviour, technical risk and the changing role of the salesperson.
Rather than attempting to predict every possible technological development, the panel concentrated on the situation facing the industry today. People are experimenting, businesses are interested and useful applications are emerging. However, there remains a sizeable gap between using an AI tool occasionally and integrating it responsibly into the way a company operates.
Adoption is happening from the bottom up
Research shared during the discussion found that 85% of respondents were using AI either occasionally or regularly. However, organisational preparedness averaged only 2.1 out of five.
Egor described the industry’s position as both “yes and no”. Lubricant businesses are beginning to adopt AI, but many individual employees are moving faster than their employers. Some are even prepared to pay personally for AI services because they can see the productivity benefits in their everyday work.
This creates an interesting crossroads. Employees are finding useful applications while leadership teams are still deciding what appropriate adoption should look like. That gap raises important questions about security, reliability and the information being shared. A productivity improvement can quickly become a business risk if commercially sensitive figures, formulations or internal documents are uploaded without suitable controls.
Progress will vary across different parts of the business
There may never be one moment when the lubricants industry collectively moves from experimentation to formal adoption. Progress is more likely to happen one business function at a time.
Egor pointed to translation as an area where AI is already producing tangible value. An international lubricant company may need to prepare marketing materials, product information and event collateral in several languages. The right AI tools can potentially shorten that process by days or weeks while reducing cost and improving speed to market.
Other applications will move more slowly, particularly when the consequences of an error are greater. Adoption will therefore depend on the task, the organisation and the confidence of the people involved. The commercial question is not simply whether a business uses AI. It is where AI can deliver measurable value without introducing unacceptable risk.
AI needs direction as well as participation
When the panel considered who should own AI within a business, Ian made an important distinction between setting the direction and executing it.
Using AI cannot remain the responsibility of one enthusiastic individual. If it is going to become a meaningful business capability, people throughout the organisation will need to understand how and when to use it. At the same time, somebody must establish the purpose, priorities and boundaries.
Many smaller companies do not yet have a clear AI strategy. People discover tools independently, find their own applications and gradually build confidence through experience. That can be a sensible starting point, but it also explains why adoption often remains disconnected. Without training and a shared direction, AI is likely to stay at the edges of the business instead of becoming part of a joined-up commercial process.
Easier access to information changes the salesperson’s role
Customers can already access product data sheets, OEM recommendations and large amounts of technical information without speaking to a salesperson. AI makes searching, summarising and comparing that information easier.
The question is what the buyer does with the answer.
Ian argued that good lubricant salespeople, particularly in technical areas such as metalworking fluids, must contribute more than product information. They need to understand the customer’s operation, explore the underlying cause of a problem and explain how a proposed solution could create value.
Steve suggested that the future of the lubricant salesperson remains secure when that person operates as a trusted consultant or adviser. Salespeople who primarily provide prices, product alternatives and information that customers can find elsewhere may be more exposed. Expertise, interpretation and commercial judgement become more valuable as basic information becomes easier to obtain.
Self-service creates a new challenge for technical selling
Rafe raised an important difficulty for the industry. A growing number of customers will want to describe their problem to an AI system and receive an immediate product recommendation. In that journey, the salesperson may never enter the conversation.
However, the customer may not know whether the answer is correct.
Rafe estimated that answers produced from general AI models might currently be accurate around 80% of the time in some lubricant-related searches. That may sound encouraging, but it is not sufficient for many industrial environments.
The opportunity for lubricant businesses is to find an appropriate place for human expertise within this self-service journey. Customers can use AI to investigate a problem and narrow their options, but specialists may still be needed to validate the recommendation, examine the application and consider consequences the original search did not capture.
Accuracy and responsibility cannot be separated
An incorrect lubricant recommendation can cause equipment damage, downtime, financial loss and reputational harm. The consequences depend heavily on the application.
Egor compared the situation with people using AI to investigate their health. The system might help somebody understand their symptoms, but diagnosis and responsibility remain more complicated. A recommendation for a small piece of equipment carries a different level of risk from advice concerning a large industrial machine.
That is where the expertise of technical and commercial teams remains essential. Lubricant professionals are not simply agents who deliver products. They help prescribe an appropriate response to a particular operating problem.
The important benchmark may not always be whether AI is perfect. Human beings also make mistakes. Businesses need to understand the comparative error rate, the consequences of a mistake and where human validation must remain part of the process.
Bounded applications offer a practical route forward
Rafe offered an example of a task where AI could provide both speed and accuracy: analysing a plant’s collection of equipment manuals to extract OEM lubricant recommendations.
This is different from asking a general language model to create an answer from a broad and uncertain pool of information. The system is working with a finite, controlled dataset and carrying out a clearly defined task.
That distinction could be useful for businesses deciding where to begin. Applications involving bounded information, repeatable processes and verifiable outputs may provide a safer route into meaningful adoption.
Ian added that the level of risk should influence the pace of change. Drafting basic content or assisting with a routine administrative activity is relatively easy to review. Contract negotiations, strategic pricing and complex technical diagnosis carry much greater consequences. Different applications require different levels of oversight.
Read more here: https://plangrowdo.com/ai-authority-audit/
Productivity alone does not guarantee a better buying experience
The most common uses of AI among sales professionals identified in the discussion were prospect research and email drafting. Egor also described employees saving hours when working with large spreadsheets, price lists and quotations containing thousands of article numbers.
These are useful efficiency gains, but Rob challenged the panel to consider whether businesses are making themselves easier to buy from or merely helping sales teams produce more of the same activity.
This is an important distinction. AI can reduce the time required to complete a task, but faster output is not automatically better output. The real commercial opportunity lies in improving the customer’s experience. That could mean responding more quickly, making technical information easier to understand, preparing more relevant recommendations or ensuring an enquiry reaches the right person.
More AI content can easily become more noise
The ability to produce something does not mean it should be produced.
Rob highlighted the familiar example of an unnecessarily long AI-generated email. The same problem can appear on LinkedIn when somebody who has rarely posted suddenly publishes lengthy, highly polished content that bears little resemblance to how they normally communicate.
This type of output may allow a business to say it is using AI, but it can also weaken trust. The tool starts replacing the person rather than helping that person communicate more effectively.
AI should help commercial teams express their knowledge, experience and personality more clearly. It should not remove the human qualities that allow customers to recognise and trust them. The goal is not to create the maximum possible volume. It is to improve the relevance and usefulness of each interaction.
Trust will develop through experience and evidence
The research discussed by the panel found that accuracy was the leading concern for 58% of respondents. However, Ian suggested that the wider trust gap also includes technology, data, leadership, training and people’s understanding of the potential value.
Businesses regularly ask lubricant suppliers for evidence that a product has helped a similar organisation. The same expectation is now being applied to AI. Decision-makers want practical case studies showing where a particular application has worked, what changed and what risks were involved.
Rafe compared adoption with installing a newly designed industrial pump. Even if the pump appears more efficient, a responsible operator would be unlikely to install it throughout an entire plant before understanding its cost and reliability. Testing it at the edges, learning from the results and gradually extending its use is a rational response.
Integration is where the larger opportunity sits
AI currently operates around the edges of many lubricant businesses. People use it to write, research, translate or complete isolated pieces of work, but it is not commonly connected to the broader commercial system.
Ian described the potential to analyse sales data alongside CRM activity. A business could investigate relationships between sales performance and sales behaviour, identify patterns and use the resulting insight to make better decisions about resources.
This is a more mature application than asking an AI assistant to draft an email. It requires reliable data, a clear business question and integration with existing processes.
The opportunity is significant, but the purpose must come first. Connecting AI to every available system without knowing what problem it should solve will only create complexity. Integration becomes valuable when it helps people make a better decision or deliver a better outcome.
The future cost of AI remains uncertain
Rafe also introduced a consideration that receives less attention: businesses may not yet be paying the true long-term cost of using AI.
Many current services offer subscription plans that allow customers to consume substantial computing resources for a relatively predictable fee. That commercial model may change. A company that builds AI into a large number of processes could face very different economics if providers introduce more extensive usage-based pricing.
This uncertainty helps explain why some business owners are cautious. They are not necessarily resistant to innovation. They may be considering how dependent the business could become on a particular provider and what that dependency might eventually cost.
Rafe suggested that open models used for specific, repeatable tasks could provide businesses with greater control, while more advanced frontier models might be reserved for less frequent research requirements.
The successful businesses will remain focused on the customer
Looking three years ahead, each panel member returned to a similar principle from a slightly different direction.
For Rafe, success may depend partly on controlling the cost of AI and selecting the right type of model for each task. Ian focused on using AI to create value for customers, whether through lower costs, better tools or easier access to information. Egor argued for pragmatism, persistence and deliberate choices rather than following the hype.
Steve added that businesses need to understand the problems they are solving and be prepared to stop initiatives that are not delivering value. AI makes it easy to start numerous experiments. Commercial discipline means recognising which of them deserve to continue.
The winners may not be the organisations that adopt the greatest number of tools. They are more likely to be those that understand their customers, control their risks and apply AI thoughtfully to worthwhile problems.
From experimentation to purposeful adoption
The panel did not present AI as something lubricant companies should fear, nor as a technology every business must immediately integrate everywhere.
Instead, the discussion revealed an industry in a sensible period of experimentation. Employees are discovering useful applications. Businesses are beginning to examine strategy, security and cost. Customers are gaining easier access to technical information, placing greater pressure on salespeople to provide expertise rather than simply supply facts.
AI is here to stay and it will influence how lubricant businesses operate. The immediate task is to move beyond activity for its own sake. Start with a real problem, understand the risk, work with appropriate data, keep human expertise involved and measure whether the application genuinely improves the outcome for the customer.
This live episode of Beyond the Blend was recorded at Lubricant Expo Düsseldorf 2026, with thanks to the organisers for hosting the conversation. Subscribe to Beyond the Blend on Spotify, Apple Podcasts and YouTube for more discussions celebrating the people, ideas and practical experiences shaping the lubricants industry.
Listen here: https://plangrowdo.com/beyond-the-blend-podcast/
