EIC Accelerator Proposal Writing: Start With The Narrative in 2026

This week I received a very interesting response to an EIC Accelerator proposal. It stood out not because it was a very good score, but because it was a perfect one: four out of four evaluators gave it a GO, and it received the maximum grade in every single criterion.

On its own, that is not that rare. A good number of companies reach a perfect score in the EIC Accelerator. What made this one different was the comments. There were zero negative comments, and there were not even suggestions.

Even Excellent Proposals Usually Get Suggestions

With very good proposals, what you normally see is that the evaluators, especially in Step 1, still add a recommendation for Step 2. They write that this is a great proposal and a great project that should be funded, and then: the company should elaborate on its commercial strategy, or strengthen its IP position, or the evaluators would like to see a little more detail on the AI development and on the data available for it. There is usually at least one suggestion.

In this evaluation there was none. There was not a single negative comment, word or sentence that you could extract from the entire document. One evaluator even went out of their way to thank us for the great proposal. I have seen kind words before, but I am not sure I have ever seen an evaluator say thank you, simply grateful to have read such a proposal.

Even I Did Not Predict It

When I handed in the proposal, I did not predict that. I wrote the proposal and crafted the narrative, but I had no idea how it would go. Of course I can make an educated guess; I know what I am doing. But you never know how the response will be. You have four random evaluators, and you do not know exactly who they are or what they will look for. So even I was surprised, and it made me think about why this proposal received such an extremely good response.

I think there is a lesson in it for other applicants.

The Basics Still Have to Be Right

Of course, the writing has to be great, everything has to work, everything has to stick to the rules, and there should be no red flags. In this proposal there were some very critical things I had to adjust. I had long conversations with the company's CFO, because we had to adjust the financial projections to match what the evaluators are going to respond to. We did a lot of work to make sure there were no red flags, and that is obviously the most important part.

We also had to make sure the entire narrative was coherent within a very short proposal. Evaluators very often complain that something is missing which is not even supposed to be in Step 1. With only twelve pages, how much space do you have for the commercial strategy, the IP strategy and everything else? It is very difficult to squeeze everything in, and yet it had to be there.

The Narrative Was the Reason

But the one thing I believe drove this response was the narrative itself. It was something I had to think about for a long time, because I was not entirely sure how to position the project.

The project sits in a strategically very important industry, but the company does not produce the thing that industry needs. Its technology is not central to the industry; it is more auxiliary, a helper technology on the periphery. That made positioning very challenging. You would like to say that you are revolutionising the industry, but you are not, because the technology is not at the core of it.

So I had to find a narrative in the absence of real government policies showing that this was important. I dug very deep, found a few smaller pieces of evidence, and built a narrative around them that made the project very big. I framed it so that Europe has missed out on a very important technological shift, but there is another part of the value chain that can still turn Europe into a winner in this particular area. That hit a nerve.

Uncertainty Gives the Grant Writer Options

The narrative was not an obvious choice. The company came to me and said: this is the idea we have; we have a different business, and we want to transition into this other industry; these are the possible application areas; we do not yet know what the project will look like or how far up or down the value chain we want to go.

That is a lot of uncertainty, which is actually great for a grant writer. The more uncertainty you have, the more optionality you have in crafting the narrative. So I told them: this is what the product should be, this is how you should see it, these are the companies we will treat as competitors, and these are the companies we will treat as distributors or customers. That is how we built the narrative, and it took a while to come up with.

It is a very powerful set-up for the next evaluation stages, because the narrative is very simple. When you hear the idea, it makes a lot of sense, and when you dig deeper, you find enough evidence to validate it. The project remains risky and uncertain, but the narrative is linear and simple.

Reason One: In 2026, Everything Is Political

There are two reasons why I think this focus on the narrative matters more in 2026 than before. I have noticed that the narrative has become more important in the EIC Accelerator. In past years it was relatively easy to submit a low-effort narrative and a very low-effort problem description, and you would still do reasonably well if the technology was interesting.

Now, in 2026, it seems that everything is political. If a technology does not have a really strong impact on politically relevant areas, such as semiconductors, energy or artificial intelligence, or medtech, it starts at a disadvantage. If you cannot craft a really strong narrative in 2026, you will have a lot of trouble in the EIC Accelerator.

There are simply too many hurdles, and at every one of them different evaluators have to understand a relatively simple idea. If the idea is too complex, you will lose some of them.

• Step 1: four evaluators assess the short proposal.
• Step 2: four more people: one technology expert interviews you, and a panel of three prepares the evaluation, as described in One Evaluator Decides.
• Step 3: five or more jury members in the final interview.

That is at least thirteen people you have to convince. If the idea is too complicated and not impactful enough, you will have a very hard time.

Reason Two: AI Cannot Write the Narrative

The second reason is that artificial intelligence is really bad at coming up with high-level narratives. It is really good at the details: getting the grammar right and making sure everything is consistent. It hallucinates, of course, but its real weakness is crafting a truly impactful narrative.

I started out as a grant writer and I really enjoy writing, and this year I made an effort to use AI beyond research: to have proposals actually written by AI, but in a way that keeps me in control of the narrative and the content and keeps the quality I am used to when I write myself. The proposal that received this response was written by AI, but directed entirely by me. It was not a single draft. It was a very slow process in which I crafted everything, and the only thing the AI did was connect the words and put everything into correct grammar.

This is how AI writing has to be done. You cannot let the AI do the thinking. You have to know exactly what each section should look like and what information has to be in it, and you cannot rely on the AI to tell you. If you rely on the AI to tell you what to fill in, it will be a bad proposal.

Of course you use AI for research and for details: what are the exact benchmarks, what are the exact names of the technology's components. The AI can fill in all of those. But you have to know exactly what the proposal looks like, and above all what the narrative is. Left alone, the AI will hallucinate its way through a horrible narrative that is completely boring and will never convince anyone. So far, crafting the narrative is something only a human can do. AI will get it wrong.

I am sure that in one or two years it will get all of these details right. But right now, crafting a narrative means understanding how a human thinks, and AI is just not good at that. For an AI, a very generic and complicated narrative is good enough. For a human, it has to be simpler.

Simple Enough for the First Slide

That is not because people are stupid. People are very, very busy. If an idea is not simple enough to be understood immediately from the first slide of the pitch deck, they may struggle to understand it even after you have talked to them for an hour. That is the reality of the EIC.