Your deck is supposed to be a glimpse of who you are.
That’s the whole point of it. Before we’ve met you, before we’ve spoken, the deck is the only version of you we have. It’s meant to tell us something about how you think, what you’ve chosen to build, what you decided mattered enough to put on slide three.
So when a deck arrives that was generated end to end, here’s what it actually tells us. You typed a prompt saying “here’s my idea, design me a deck and fill in the slides.” That’s it. That’s the entire contribution. You haven’t shown us who you are, you haven’t shown us what the company is, you haven’t shown us anything that makes you different from anyone else.
And we’re now receiving hundreds of these. They all look the same. They sound the same. They’re structured the same. They blur into each other.
This is about standing out, and AI isn’t helping you do it any more
For a while, using AI on your deck genuinely gave you an edge. If you weren’t a designer, it made you look like one, and that was worth something.
That window has closed. When everyone has access to the same tools and the same default prompts, the output stops being a differentiator and becomes wallpaper. A deck that looks like every other deck we opened this week doesn’t get attention, it gets skipped, and the founder who made it never finds out why.
You’re competing for attention against the other 2,000 deals we see every year. You need to stand out.
So use AI, just don’t hand the whole thing over
I’m not telling you to build slides from scratch in 2026. Use the tools. But use them as a starting point, not as the finished article.
Put your flair in it. Customise it. If you’re going to prompt, write a prompt that actually describes you, your company, your market, your way of doing things, rather than the standard one that produces the standard deck. And then go in afterwards and change it, because the version that comes out of the machine is by definition the average of everything it’s seen.
The test is simple. If I could swap your logo for another company’s and the deck would still make sense, you haven’t made a deck about your company. You’ve made a template.
What we actually read
I used to skip the TAM slide. Now I skip it even harder, because every single one of them is identical: a big number, a nice chart, no source, and a claim that capturing 1% of it makes you a unicorn.
Same for the competitor grid where all the ticks are in your column. Same for the hockey stick with no assumptions behind it. These were always weak slides, but at least they used to take effort to make, so they told us something about your judgement. Now they cost nothing and they tell us nothing.
What we’re looking for is the part AI can’t produce for you:
- What you’ve achieved. Real numbers, real months.
- Your traction. Who’s paying, how many, are they still there after six months.
- Who you are. Why this team, why this problem, what you know about it that other people don’t.
That’s what matters now. If those three things are clear and near the front, the design of the rest matters far less than founders think.
The personalised paragraph isn’t personalised
The email problem is worse than the deck problem.
Every email I get now has the same shape, including the little paragraph near the top designed to show that the sender knows who I am. They don’t. The AI wrote it, it did the homework, and I can tell. I glaze over it. The moment I see one of those openings I know exactly what the rest of the email says, because I’ve read hundreds of identical ones, and I stop reading.
That paragraph used to be the strongest signal in a cold email. Now it’s the strongest signal that a founder outsourced the one part of the process that was supposed to demonstrate effort.
So don’t use a standard prompt. Don’t have AI draft the whole email. Write it yourself, keep it short, and let it sound like a person.
And honestly, don’t send the email at all
Here’s the part that costs founders the most, and it takes two minutes to avoid.
Our process is published. It’s on our website, it’s said in every place we say anything. We don’t take pitches through random emails to a generic inbox. We have an application form at plus.vc/apply-now, and that’s the route in. We review every deal that comes in through our process, and spend 0 time on any that just blindly come in via a cold, AI drafted and sent email.
If you spend one minute on our site before you write to us, you’ll work out that you shouldn’t be sending us an email at all, you should be filling in the form. That’s it. That’s the whole homework assignment, and AI can’t do it for you because the point isn’t the information, it’s that you bothered.
A cold email to a generic address has already failed the first filter. And that filter has become far more useful to us lately, because it now separates the founders who pay attention to how to get our attention from the founders who handed the entire task to a machine. In my experience the second group doesn’t differentiate itself at the next stage either.
The worst ones
The genuinely bad decks are the ones where the AI has guessed at what we do, got it wrong, and then written around the guess.
It invents a mandate for us, describes an opportunity that has nothing to do with the one we actually have, and confidently explains to the founder that Plus VC would be interested in this. We have a structure. We know what we invest in, at what stage, at what cheque size. All of it is public. A machine’s confident guess about our mandate is worth nothing, and the founder is the one who pays for it, because they’ve spent their one shot on a pitch aimed at a fund that doesn’t exist.
Don’t hand your fundraise over
Use AI for the parts that are genuinely work. Cleaning up a model, tightening copy that’s gone flabby, a second language version of the deck, attacking your own numbers before we do.
But don’t give the whole thing away, and least of all on fundraising, because fundraising is the one exercise where the entire job is to differentiate yourself. Show us who you are. Show us how you’re different. Show us why you’re better than one of the other 2,000 deals we’re looking at this year.
And then follow our process. Do those two things and you’re already ahead of almost everyone who wrote to us this week!