Generative AI & Bid Writing: Levelling the Playing Field or New Challenges?

Generative AI & Bid Writing: Levelling the Playing Field or New Challenges?

Picture of Plan A

Plan A

Generative AI tools, like ChatGPT, are the talk of the town in various sectors, including bid writing. As someone entrenched in the tendering process, I’ve witnessed firsthand how these tools can significantly impact the way we approach and execute bid submissions. While generative AI offers numerous advantages, it also brings several challenges that must be addressed. In this article, I’ll explore both sides of the coin. We’ll unpack the opportunities, as well as key aspects where companies need to be careful.

Levelling the Playing Field

Evaluators tell us that tender submissions are judged on the quality of the service or product offered – not the eloquence of the bid writer. That said, for tender evaluators who are assessing 10, 20, or more bids, better quality writing will make life easier. One of the most significant strengths of generative AI is its ability to democratise the bid writing process. Traditionally, companies with more resources would hire professional writers to craft easy-to-read tender responses, giving them a distinct advantage over smaller businesses. With AI tools like ChatGPT, even companies with limited resources have the potential to produce well-written bids.

As a bid writer, my job is not to fabricate amazing stories about a company’s skill or experience. A great bid writer’s skill is in taking what the company can do and crafting it to maximise the evaluators’ scores. Tenders win through the strength of the company’s product or service, not in their writing prowess. My hope is that, when used properly, Generative AI will level the playing field: it will prompt companies to focus on showcasing their strengths and innovations, rather than polishing the perfect narrative.

What’s more, if you can quickly generate well-structured responses, you can bid for more contracts. That’s a huge bonus for smaller businesses seeking to expand their reach and win more tenders. When properly established, AI tools can streamline the bid writing process, saving time and resources and enabling companies to focus on refining their products and services.

Risks of Misuse and Lack of Understanding

A common pitfall of AI-generated content is its default to generic content when quality information isn’t supplied. It’s actually quite easy to identify AI generated content. There are four ‘red flags’ of a Gen AI response. Here’s what they are, and how to avoid them.

  1. Generic content: If Generative AI is not provided specific content, it will often revert to providing a high-level summary of what you should consider in your answer. For example, instead of actually creating a Road Traffic Management Plan, it will outline in bullet points what a generic Road Traffic Management Plan could contain. I am genuinely intrigued to hear from evaluators the number of ‘how to’ guides they get for various policies, procedures, or plans, instead of a specific company documents.
  2. High school essay formatting: ChatGPT seems to have been coded to automatically produce outputs in persuasive essay format, unless told otherwise. When I see a response that has ‘Intro’, ‘three-key points’, then ‘to conclude’ or ‘in summary’ (which no adult ever has done unless writing a university essay), I automatically know it is Gen AI written. In general, I have nothing against this style of writing, however, I can imagine how frustrating it will get for an evaluator having ten or more tender responses where every question reads like a 15-year-old trying to persuade their teacher on ‘why a second lunch break would be beneficial to learning outcomes’.
  3. Americanised spelling: I do believe that we will all start using American spelling thanks to ChatGPT. In the interim, it is an absolute giveaway that you have used chat, and not edited your work, when I see ‘z’ instead of ‘s’, and ‘or’ instead of ‘our’ in words.
  4. Overly descriptive: Being click happy with adjectives, does not convince an evaluator of your skill in delivering on the contract. Take the following two examples:

Example One: “Born from the harsh realities of New Zealand’s rugged coastlines, our civil works company has woven a tapestry of engineering marvels through our extensive experience in constructing rock embankments for shoreline stabilisation. With each project, we harness the raw, untamed power of nature, meticulously selecting each rock with the precision of an artist crafting a masterpiece. Our dedicated team, a symphony of skilled engineers and passionate workers, works day and night, driven by an unwavering commitment to preserve the beauty and integrity of our cherished shorelines. The harmonious blend of nature’s elements and our innovative techniques results in structures that are as resilient as they are visually stunning, standing as proud sentinels against the relentless assault of the ocean’s waves.”

Example two: “Our civil works company has 15 years’ experience in the construction of rock embankments for shoreline stabilisation. Over this time, we have successfully executed 52 projects, demonstrating our proficiency in selecting appropriate materials and employing effective engineering techniques. Our team comprises highly skilled engineers and technicians, dedicated to delivering robust and durable structures. By leveraging our knowledge and experience, we ensure that each project meets the highest standards of quality and safety, providing reliable protection against coastal erosion, while preserving the natural aesthetics of the shoreline.”

Without doubt the second company would ace the scoring. It’s simple, straightforward and to the point. Evaluators can easily identify key skills and experience.

So, would I use Gen AI to respond to a bid?

No – or, at least, not in its entirety. Generative AI will only work well when you a) have  100% clean and robust data to provide it with and, b) you give it clear guidance through specialised training and prompts. Very few companies can do this!

But all’s not lost for Chat GPT. Here are some valuable uses that will help your bid writing:

  1. A collaborative ‘ear’ to explore each tender question. Help to understand what the Funder wants in the ideal response. Try asking Gen AI to ‘act like a council infrastructure manager’ and ‘outline key areas you would like a service provider to consider when delivering on xyz’. You may get new, and useful insights.
  2. High level structural outline for more complex, or multi-layered, questions. It’s often hard to wrap your head around the best way to respond. You have all the components, but you want to make sure your writing is clear and succinct. Generative AI can help suggest layouts or format options to maximise impact.
  3. Checking your responses for information ‘gaps’. Generative AI can review your content and identify potential missing information or spot a weak response. Word of warning here, don’t put proprietary or commercially sensitive information into any open-source Generative AI platform. It could surface in your competitor’s bids – nightmare!

Use Generative AI sparingly for content generation, and only if you are confident of your data is clean and strong; and you have given specific prompts to guide it.

Watch carefully that it doesn’t misunderstand the question – and then limit constructive thought on other possible responses, or – worst case – invent new, inaccurate information about your offer. Make sure someone on your team thoroughly reviews any AI-generated content, to make sure it’s true for your company and it meets the tender requirements.

Final thoughts

The skeleton in Generative AI’s closet may yet be data privacy. AI tools absorb sensitive company information and proprietary details into their systems. Consider how this data is stored, used, and protected, and how you can prevent unauthorised access or misuse of your information. Make sure your AI policies are robust and compliant with relevant data protection regulations. If a tender specifically mentions confidentiality, you must respect this request. Do not copy and paste that tender information documents into ChatGPT.

While AI can be a powerful ally, it should be used thoughtfully and responsibly to deliver its true benefits. Misunderstandings and myths regarding Generative AI still prevail in the bid writing space.

Despite the advantages, some organisations have imposed blanket bans on the use of AI in bid writing. While these bans are well-intentioned and aimed at preserving the integrity and originality of submissions, they can also stifle competition. Restricting access to AI tools may stop smaller companies that rely on AI from bidding. A tender should be about the best product or service, not about who can write the best response unaided by technology.

AI could revolutionise bid writing. It could level the playing field by prompting companies to showcase their true strengths. However, risks like data privacy, misuse of AI, and generic content are real.

Our team at Plan A can strike a balance between leveraging AI’s capabilities and maintaining the integrity of your tender responses. Let’s watch this space: used cleverly, you too can harness the full potential of these tools to succeed in the competitive world of bid writing.

Hannah is a freelance Bid Writer for Plan A with a background in Health, Infrastructure, and Humanitarian Development. Alongside her partner, an AI tech developer, Hannah has co-founded an AI consulting company, Strategic Focus AI, where she undertakes the role of Lead Strategist