Paste the narrative copy from each slide. Get a 0-100 Authenticity Score in thirty seconds that predicts whether a VC associate reads your deck as a real founder voice or another ChatGPT draft. Sentence-level highlights flag the exact lines that look generated so founders, sales teams, and accelerator applicants can rewrite before the deck hits an inbox. Investor seed, Series A, sales decks, partnership decks, and B2B SaaS demo decks all run through the same scorer.
No other document type concentrates pattern volume in a single inbox the way a pitch deck does. After two years of ChatGPT-drafted decks, the reader on the other side has a personal classifier sharper than anything polish can fool.
An associate at an active seed fund reads 40 to 80 decks in a normal week, and a partner reads the forward pile on top of that. Across a year, that is two to three thousand decks per reader. Signature ChatGPT phrases like revolutionizing the X industry or unprecedented market opportunity jump off the slide before the reader has parsed what the company does. Decision time per deck at most seed-stage funds runs under ninety seconds.
VCs are not grading prose quality. They are deciding whether the founder thought through the story. An AI-flavoured deck answers that question for them. If the founder outsourced the narrative, the working assumption is that the founder probably outsourced the strategy. The deck never makes it to the partnership meeting because nobody has anything memorable to repeat about it on Monday.
Several accelerators added AI-screening to their 2025 cohort review, and a growing number of seed and Series A funds quietly run incoming decks through a detector before the associate sees them. The reasoning is not that AI-drafted decks are disqualifying on principle. It is that a founder who could not be bothered to put their actual voice on the page is signalling something about how they communicate to customers and to a board.
Even at funds without detection tools, an associate who likes a deck has to forward it to a partner. Partners read in batches on weekends. A deck written in template prose does not survive the partner first pass because there is nothing specific to repeat in the Monday meeting. Specific founder voice is what gets a deck advocated for, not just received. Without advocacy, the deck dies in the forward pile.
These patterns appear across investor seed decks, Series A decks, sales decks, partnership decks, and B2B SaaS demo decks. Experienced readers spot them in seconds, often before they can articulate why a slide feels generic.
Disrupting the $X market. Massive opportunity in a multi-trillion-dollar industry. Game-changing technology in a rapidly evolving landscape. Every associate has seen this opening line hundreds of times. The fix is opening TAM with a bottom-up number tied to a real wedge: 1,200 mid-market dental practices, $9,000 ACV, $10.8M serviceable. That reads like a founder who has done the math.
ChatGPT defaults to three balanced bullets on Solution. Bullet one is a noun phrase about workflow, bullet two is a noun phrase about AI or automation, bullet three is a noun phrase about outcomes. Each bullet roughly the same length. The fix is letting the slide be uneven. One bullet is a sentence, one is a fragment, one is a question your customer actually asks. The seams are the conviction.
Passionate founder with deep expertise in [industry] and a proven track record of building category-defining products. This is the single most recognised AI tell in fundraising. Associates stop reading the Team slide three words in. Replace the framing with one specific: Built and sold the email-deliverability piece of [Company], now applying the same approach to [new wedge]. Specifics are the one thing ChatGPT cannot fake because it does not know your resume.
As digital transformation accelerates and AI adoption reaches an inflection point, the timing has never been better for [category]. Real Why Now answers cite a specific regulatory shift, a specific incumbent stumble, or a specific buyer behaviour that started inside the last eighteen months. Replace the macro paragraph with one date and one consequence: GDPR-2 enforcement starts March 2026 and every B2B SaaS company processing EU data now needs the answer we sell.
Building the future of [category]. Reimagining how [audience] works. Every reader has filed these under no specific claim. Replace the vision slide with a forward-looking specific: by Q4 2027 the default integration in this category, by 2028 the default in two adjacent categories. Concrete dates and bets read as conviction, not as filler.
Robust, scalable, AI-native, customer-centric platform. Four adjectives doing the work of zero specifics. The scorer flags these clusters because they show up almost exclusively in generated copy. Cut all four. Replace with one verb that describes what the product actually does in a single sentence, and the Problem slide moves from the high-AI band into the reply-worthy band in one edit.
Five bands that map Authenticity Score to expected associate behaviour. Calibrated for short narrative slides where signal density is high and decision time is under ninety seconds per deck.
Specific customer names, real numbers, lived founder context, and the kind of phrasing that only comes from someone who has been in the problem for years. This band advocates itself inside the fund. Slides in this range are what the partner remembers from the deck on Monday morning and what the associate quotes when forwarding to the partnership. This is the target band for any priority fund.
Reads as a real founder with a couple of slides that got cleaned up too much. The associate books the first meeting. Worth one editing pass on the stiffer slides before sending to the top three target funds, but acceptable for broad outbound. Most decks land here on the first authenticity pass.
The associate finishes the deck but has nothing specific to say in the Monday partner meeting. You usually get a polite no or a not now, keep us posted. Half the slides have founder voice, half read generic. Fix the narrative slides (Problem, Solution, Why Now, Why Us) before sending to the funds you actually care about.
The associate forms a pass decision in the first two slides. Problem and Solution read as template ChatGPT. Even strong traction numbers later in the deck do not recover the impression. Do not send at this score to any fund on your priority list. Restructure the narrative slides before another touch.
Reads as raw ChatGPT output. At funds with AI detection in intake, the deck is filtered before an associate sees it. At funds without, the associate pattern recognition rejects it inside thirty seconds. The fix is a full rewrite of the narrative slides, not surface edits. Use the pitch deck AI rewriter workflow instead of editing word-by-word.
Free covers three to four full deck narrative passes per day on the 10K daily detect budget. Founders iterating a seed deck weekly usually run on Starter. Sales teams pitching across portfolios start on Pro. Accelerators and RevOps teams running batch deck audits start on Business for the REST API. Full details on the pricing page.
Billed $89.88/year — Save $30
Billed $179.88/year — Save $60
Billed $359.88/year — Save $120
Yearly billing saves 25%. View full pricing →
A four-step loop that fits between draft and send. Most founders run it on every priority fund touch and on every revision after partner feedback. Total time under fifteen minutes for a seed deck after the first run.
Open the slide deck in Google Slides, Pitch, Notion, or wherever the founder draft lives. Copy the prose body of one slide at a time into TextSight. Start with Problem, then Solution, then Why Now. Skip bullet-only and table-only slides because the signal is unreliable on fragmented text and the score does not reflect investor reaction in those cases.
The scorer returns a 0-100 Authenticity Score for each slide plus a sentence-level colour map: green for human-aligned, yellow for mixed, red for likely AI. The slide score tells you whether the slide needs work. The highlights tell you exactly where. On a low-scoring Problem slide, two or three red sentences usually carry the entire score, and editing those three is faster than rewriting the slide.
Sort the slides by score and start with the lowest. Problem and Solution carry the most weight on associate first impression, so prioritise those even if Vision or Why Now scores lower. Read each red sentence out loud. If it sounds like something a McKinsey deck or a TED preview blurb would say, it is generic. Replace with the specific version: the named customer, the actual workflow, the exact failure mode you are solving.
Rewrite each red sentence with a concrete reference or a real number. If a sentence resists manual editing, run it through the AI rewriter in Light or Balanced mode. Skip Maximum for pitch decks because aggressive rephrasing can soften a value claim. Paste the revised slide back into TextSight to verify. Target 80 or higher on every narrative slide before the deck leaves your laptop.
A pitch deck is not one document for scoring purposes. Each slide has its own narrative density, and the AI signal concentrates differently on each one. These eight slides carry prose. Skip bullet-only or table-only slides.
The slide most likely to read AI. Target above 80. If you cannot get the problem statement past 70, the issue is generic framing, not word choice. Name a customer, name a day, name a tool that broke. The Problem slide decides whether the associate keeps reading.
Second-highest AI risk. ChatGPT defaults to philosophy here with words like platform, AI-native, and seamless. Target 75 plus. Describe the product mechanically. Cut every adjective that does not modify a noun the customer would recognise.
High AI risk because it tempts trend-bait language. Target 75. Anchor on one specific recent shift with a date. The rise of AI is not a Why Now. A specific cost change, regulatory event, or buyer-behaviour shift inside the last eighteen months is.
Easy slide to score well on if you write it yourself. Names, prior roles, specific shipped projects. Target 80 plus on the prose, ignore the bullet list of credentials. The passionate-team phrase is the single most recognised AI tell on any deck.
Mostly numbers, so the prose carries less signal. Target 60 plus is fine. Keep the chart, fix the caption. Strip phrases like demonstrating strong product-market fit around a perfectly good growth chart and let the number speak.
High AI risk. ChatGPT writes generic Gartner-style market prose. Target 70 plus. Cite source and date for each market number. A bottom-up wedge calculation beats a top-down trillion-dollar opener every time.
Medium AI risk. Avoid best-in-class, first-mover, no direct competitors. Target 75. Name specific competitors and one concrete differentiator each. The differentiator should be one the customer would describe the same way.
Vision is high AI risk, Ask is low. Vision target 75. Ask target 60 is fine because it is structured by dollar amount and milestones. Vision needs concrete dates and bets; Ask just needs the round size, lead status, runway months, and three milestones the money funds.
Each role has a different deck volume, a different decision-maker on the other side, and a different way the scorer fits into the week. Same scorer, different rhythms.
The highest-stakes use case. One deck goes to ten to thirty target funds across four to eight weeks. Score every revision after partner feedback. Track the narrative-slide average as the deck evolves; the goal is a Problem slide above 80 and a Solution slide above 75 before the deck reaches the top three target funds. Iterate on the slides scoring lowest before each new round of outreach.
VPs of sales and procurement leads read twenty to forty vendor decks a quarter. The same conviction-led test applies to partnership and B2B SaaS demo decks. AEs score the bespoke decks before sending to a target account, and sales engineers score the demo decks before customer playback. Two minutes per deck, much higher reply rate downstream.
YC, Techstars, and 500 Global review thousands of decks per batch and several added AI-screening to their 2025 cohort review. Score the Problem and Solution slides above 80 before submitting; that single threshold raises pass-through rates measurably on applications.
Mid-market and enterprise corp-dev teams read partnership decks with the same calibrated eye as VCs. Score the pitch slides before sending to a target partner. Even where the partner is open to the conversation, an AI-flavoured deck delays the first call by two to three weeks.
Batch-score every deck in a cohort using the REST API on Business. Surface the lowest-scoring decks for founder coaching before demo day. The audit happens inside the existing batch management workflow rather than asking founders to run scans individually.
The full slide-by-slide rewrite workflow once the score flags red sentences across your deck.
Open AI rewriter →Sister scorer for outbound founder messages, SDR sequences, and warm follow-ups.
See the workflow →The other short-form professional document where AI signal concentrates and a single phrase ends the conversation.
Read the guide →How the 0-100 score is calculated, what each signal weighs, and where the calibration comes from.
See the score →Free to try, no card. Slide-by-slide scoring, sentence-level highlights, 0-100 Authenticity Score, ten thousand detect characters daily.