Why Tender Teams Lose Time Before They Even Start Writing the Proposal

While winning tenders may feel like a sales problem, a large portion of the challenge occurs before the proposal even gets to the client.
A tender is usually received as an email or through a procurement portal. Someone needs to find the tender, analyze if the opportunity fits the organization's strategy, assess the requirements and estimate the commercial value, and determine if the organization has the time and resources to respond.
Depending on the decision, another challenge presents itself.
Estimators and commercial teams will have to assess a large volume of tender documents and determine the requirements, review the Bill of Quantities (BOQ), search for relevant past submissions, and draft a new proposal, which may have already been written a number of times previously.
The end result is a significant amount of valuable employee time spent locating and restructuring information, rather than using their expertise to engage with the opportunity.
Tendering creates an information retrieval problem.
Knowledge traps tend to develop around a company’s past proposals.
Past proposals may contain a plethora of information, including pricing, techniques, compliance, project descriptions, and even ways similar requirements were addressed. Unfortunately, this information is not accessible, valuable knowledge, and takes a significant amount of time to locate and adapt.
When a new RFP is released, an estimator is likely to know that the company responded to a similar RFP two years ago. However, finding the exact submission, pulling the relevant sections, and adapting it will take hours.
The problem becomes much more problematic when multiple tenders are released at the same time.
A business may have the technical ability to act on a particular business opportunity but may not have the needed administrative ability to prepare the response in the time available.
Research in procurement generally and AI within procurement specifically is more advanced. One such recent study in the Journal of Purchasing and Supply Management (2023) provides a systematic overview of AI procurement tools, outlines the challenges and benefits of use, and acknowledges the need to apply a critical lens to operationalize these tools while accounting for bias, explainability, and operational risk.
The AI procurement domain is gradually growing, and shifting attention to supplier offer analysis and large language models (LLMs) is a positive step in this regard. One of the first studies of this kind published in 2026 looked into formal client offer checking procurement tasks, where the author argues that LLMs have the potential to significantly reduce procurement work and time.
One of the main sources of wasted effort is related to bid or no-bid decisions.
Not every RFP warrants a response.
A company’s minimum contract value, preferred tender sectors, preferred geographic areas, and customer requirements may preclude them from responding to RFPs.
Many organizations employ an informal, judgment-based approach when deciding to respond to RFPs. The first person to see an RFP typically makes the initial assessment, and the assessment criteria may vary across employees.
This creates two problems.
Good RFPs may go unanswered, while significant time is spent preparing responses that will invariably be rejected.
An AI system to conduct the initial assessment within predetermined criteria will require less time compared to employees while providing the necessary commercial structure.
What AI Tendering Systems Should Actually Do
AI assistants used to handle more than document summarization.
They should have the ability to discover tenders that match their user’s preferences, evaluate them based on some rules, and connect tenders to previous relevant information.
That is already possible with ZentraBid.
ZentraBid checks RFP sources that match the user’s Rules of Engagement. It advises whether a tender should be answered, suggests the reason(s) behind the advice, and uses previous RFPs to draft a response. The platform explains that users should spend only a few hours reviewing and finalizing the responses.
That is significantly different from other platforms.
AI tools of this nature are designed to assist the estimator. They are not designed to replace the estimator. These tools are designed to help estimators spend their time and effort on things that require expertise.
AI makes archived tenders useful.
AI makes use of information previously captured. Winning tenders can serve as an excellent example of how new tenders should be answered. However, the information and documents captured in the archive are useless if employees do not have access to them. AI would allow employees to easily draft the response to a new tender by automatically filling in the relevant information from a winning tender that was archived.
That doesn't mean historical content should be copied verbatim. Project conditions change, prices change, and contractual requirements change. The idea is to provide estimators with a good starting point, which can then be checked and adjusted.
Research into AI and procurement decision support is increasingly analyzing AI in the role of assisting procurement professionals, as opposed to replacing them. A 2025 systematic literature review in Artificial Intelligence Review found considerable potential for AI and machine learning in procurement and purchasing decisions, as well as challenges associated with implementation.
The tendering process becomes easier as preparation is automated.
The commercial value of tendering automation goes beyond the faster production of tendering documents.
It is the capability to assess a greater number of opportunities while reducing the administrative burden at a proportionally lower rate.
A firm can apply its qualification criteria consistently, provide estimators with access to relevant old documents, and begin tendering with draft proposals, as opposed to blank documents.
That is the operational problem ZentraBid is addressing. It performs the first stage of tender analysis and proposal drafting while retaining the final commercial decision with the bid's responsible persons.
Further research
The role of artificial intelligence in the procurement process: State of the art and research agenda, Journal of Purchasing and Supply Management, 2023.
Artificial intelligence and machine learning in procurement and purchasing decision-support, Artificial Intelligence Review, 2025.
Enhancing Procurement Processes in Supply Chain Management with Large Language Models, Procedia Computer Science, 2026.