• Written by Azhar Aliyar, Director

A category manager at a chemicals business is deciding this week where to allocate the IT budget, and currently bases it on a long list of requested reports and system integrations, each one aimed at answering a specific question about production, inventory, or regulatory compliance. This approach is misguided, as it focuses on adding new systems and reports rather than getting more value from the existing ones.

Where IT money actually goes in a chemicals business

Most IT spending in chemicals is devoted to extracting insights from various systems, such as laboratory information management systems (LIMS), batch control systems, enterprise resource planning (ERP) systems, and environmental health and safety systems. According to the US Census Bureau's Business Trends and Outlook Survey, 22.4% of US businesses reported using AI in the two weeks to July 2026, but in the chemicals sector, this often means adding new reports and integrations to existing systems. A shift supervisor checks the maintenance log, while a quality control specialist compares batch data, but each new question seems to require a new report or integration.

This approach is costly, as each new report or integration requires significant resources to design, develop, and maintain. Furthermore, the US Census Bureau's Business Trends and Outlook Survey found that 12.4% of US manufacturers, which includes chemical producers, reported using AI in the previous two weeks, but much of this effort is spent on building and maintaining point-to-point links between systems.

The report and dashboard backlog

The traditional approach to answering new questions in a chemicals business is to build a new report or dashboard, which can be time-consuming and expensive. This leads to a backlog of requested reports and dashboards, each one competing for limited IT resources. The category manager compares two suppliers, but the data is scattered across multiple systems, and a new report is needed to bring it all together.

AI can replace this approach by allowing users to ask questions of the existing systems in natural language, without the need for new reports or dashboards. This can significantly reduce the report and dashboard backlog, and free up IT resources for more strategic initiatives. For example, a consultant can help implement an AI system that allows users to ask questions of the existing systems, and get answers in real-time.

Extending LIMS, batch control, ERP and the environmental health and safety system

Rather than replacing existing systems, such as LIMS, batch control, ERP, and environmental health and safety systems, AI can be used to extend their capabilities and get more value from them. This approach can defer a replacement programme, and reduce the costs associated with implementing new systems. A quality control specialist can ask better questions of the existing LIMS system, and get more insights from the data, without the need for a new system.

According to the US Census Bureau's Business Trends and Outlook Survey, 64.3% of US businesses that use AI made no change to their existing systems or processes, and only 6.8% changed how they collect or manage data. This suggests that AI can be used to get more value from existing systems, without the need for significant changes or replacements.

Integration spend

Traditional system integration approaches often involve building and maintaining point-to-point links between systems, which can be costly and time-consuming. AI can replace this approach by joining data at the point of the question, rather than building and maintaining a separate integration for each question. This can significantly reduce integration spend, and make it easier to answer new questions.

For example, a category manager can ask a question about production costs, and the AI system can bring together data from the ERP system, the batch control system, and other relevant systems, without the need for a separate integration. This approach can also make it easier to switch between different systems and data sources, and reduce the costs associated with maintaining multiple integrations.

Where AI adds cost rather than removing it

While AI can reduce IT costs in some areas, it can also add costs in others, such as data preparation and governance. According to the US Census Bureau's Business Trends and Outlook Survey, 11.4% of US businesses used AI in their information technology function in the previous six months, but this often requires significant investment in data preparation and governance. The data work and governance that is required to implement an AI system cannot be skipped, and must be budgeted for honestly.

A strategic leader must be aware of these costs, and plan accordingly. This may involve investing in data preparation and governance, and ensuring that the AI system is properly integrated with existing systems and processes. It may also involve changing how the organisation collects and manages data, and ensuring that the data is accurate and up-to-date.

How to tell whether IT spending actually fell

To determine whether IT spending has actually fallen, it is necessary to measure the reduction in requests and licences, rather than just comparing it to a projection. This can involve tracking the number of reports and dashboards that are no longer needed, and the number of integrations that are no longer required. It can also involve measuring the reduction in IT resources that are required to support the existing systems.

For example, a category manager can track the number of requests for new reports and dashboards, and the number of integrations that are required to support the existing systems. By measuring these metrics, it is possible to determine whether the AI system is actually reducing IT spending, and to make adjustments as needed. As discussed in our blog, AI can have a significant impact on IT spending in the chemicals sector.

What this does not do

This approach to reducing IT costs in chemicals with AI does not replace the need for strategic planning and leadership. It also does not eliminate the need for IT resources and expertise, but rather changes how they are used. Furthermore, it does not apply to situations where the existing systems are outdated or inadequate, and a replacement programme is necessary. In these cases, a more traditional approach to system integration and replacement may be required.