The 30-second answer: send a PDF under 10MB via a tracked link (unique per investor), in an email under 200 words, with real customer names and a specific 30-minute ask. Don't attach PowerPoint. Don't BCC. Most investors will not sign an NDA before reading an initial deck. Keep that first-send deck non-confidential, then follow counsel and the investor's process when sensitive diligence begins. Follow up after 7 days if you get no response.
That's the summary. Everything below is the detail: what investors actually told us, what worked for founders who raised, and what the 2026 data says.
What you'll get in this post
Six verbatim VC quotes on what to do and avoid. Four real cold emails that landed funding (including the same founder's failed v1 vs. the successful v2). Fresh 2026 benchmarks from DocSend, Storydoc, and Carta. A stage-by-stage map of what to send at pre-seed vs. Series A. And the deeper workflow: how a fundraising data room changes the conversation once investors ask for more.
What investors actually want
Before the mechanics (PDF vs. link, 10 slides vs. 20, warm intro vs. cold), a handful of named partners have said the exact same things for two decades. The founders who get funded read these before they write their first email.
On NDAs: start with a non-confidential deck
Brad Feld, Foundry Group, 2006:
"Even if I was inclined to sign an NDA, I'd have to go through the process of reading it and deciding if it had any problems… In some cases, I'd probably spend more time dealing with the NDA than with the entrepreneur and his idea. How stupid."
Fred Wilson, Union Square Ventures, same week:
"I have been in this business for 20 years and to my knowledge, I have never signed one."
Guy Kawasaki was sharper:
"If you even ask them to sign one, you might as well tattoo 'I'm clueless!' on your forehead."
Mark Suster added the mechanical reason: VCs review three or four similar deals per market at any given time. Signing an NDA puts them in legal jeopardy every time they hear two founders pitch an adjacent idea. "Asking for one to be signed shows naïveté."
Those comments describe the first pitch, not every diligence process. Do not put trade secrets or privileged material in the outreach deck. Share a non-confidential version first, then follow counsel and the investor's process when sensitive diligence begins.
On deck length: use the smallest deck that makes the case
Kawasaki's original 10/20/30 rule remains a familiar reference point, but current platform reports do not support one universal deck length. Their cohorts and formats differ: recommendations range from compact decks to roughly 20 pages.
Use the fewest slides that make the stage-specific case without relying on live narration. Our source-audited pitch deck benchmarks for 2026 keeps each length recommendation attached to its underlying cohort.
On email gating: investors hate it
Renata George, Zenmen Venture Fund:
"I've been sent a link to a pitch deck that I could only view if I type in my email in the pop-up window (this reminds me of lead magnets on all kinds of spam websites, so I am very reluctant to do that)."
Mark Suster, in 2018:
"VCs are acutely aware that with a link they're being tracked, so those benefits to you may actually limit consumption. If your deck is something I should open 3-4 times as I'm contemplating an investment, why add any consumption friction to the reader?"
That quote is seven years old and it's still the critique founders need to take seriously. If your tracking link requires an email to open, some investors will not continue. A separate Open link for each investor keeps activity grouped by intended recipient without adding a gate, but it does not verify who used the link.
On cold email: warm intros aren't sacred
Hunter Walk, Homebrew:
"Send me a great cold email. One which tells me why you're reaching out, directs my attention to something, and suggests what you'd like as a next step… There's nothing wrong with a solid cold email."
Walk adds that "at least one third, and maybe as much as 50%" of the warm intros he receives are from people who barely know the VC. A strong cold email beats a weak warm one. The 2026 data confirms this: cold decks average 2:31 of read time and 3-5% meeting conversion; warm decks average 4:18 and 40-50%. Warm intros help, but only when they're actually warm.
Pick your delivery: single deck or data room
The biggest mistake founders make isn't the deck itself. It's sending the same package at every stage. What a pre-seed investor wants to see is not what a Series A partner needs to open a diligence folder.
Use stage as context, not as a fixed document count.
Pre-seed: lead with the smallest deck that proves the case
Send a focused deck and one paragraph of context. Keep the current cap table, formation records, IP-ownership evidence, financial snapshot, roadmap, and pilot or beta evidence ready, but share supporting material when an investor asks for it.
Seed: use one tracked link per investor
Send the current deck through a separate personalized link for each investor. If another unique viewer appears on the link, treat it as an additional-viewer signal. It can reflect internal sharing, another device, or another stakeholder. It does not prove who used the link or who shared it.
Series A: open a focused room when diligence starts
Prepare the financial, ownership, traction, product, team, and company records early. Open a focused room when the investor requests evidence or begins diligence, then add reviewed material against the actual request list.
The workflow also changes what you track. Deck analytics show how a first-send document was used. Room analytics add document-level context. How to track an investor pitch deck explains what those signals can and cannot tell you.
Later rounds: choose the diligence system the process requires
Use a formal virtual data room when counsel or the transaction requires granular file permissions, formal audit exports, structured questions and file requests, NDA acceptance records, SSO, or contracted compliance evidence.
Prepare the records early, lead with a tracked deck, and expand the room as requests arrive.
How to send a single deck
At pre-seed and seed, the base case is still "send one deck." Here's what to get right.
Format: PDF, under 10MB
Mark Suster again:
"I know you love your high-quality images that cause your file to be 82MB, but the reality is that if your goal is to have your deck be read, this is ridiculous… I recommend sub-10MB."
PDF is non-negotiable. PowerPoint and Keynote files break fonts across machines, bloat in size, and can be edited by the recipient. Pitch decks live or die on layout, and layout breaks on someone else's software.
Export to PDF. Check the file size. If you're over 10MB, compress images (most export tools have a "reduced size" setting that keeps quality acceptable and cuts file size by 70%).
Delivery: tracked link, not an attachment
A tracked link gives you two things an attachment doesn't:
- You can update the deck without resending. Spot a typo on slide 7? Swap the file. Every existing link now serves the new version.
- You see supported activity: recorded opens, slide-level attention, return visits, and whether an additional viewer context appeared. A personalized link keeps the record grouped by intended recipient but does not prove identity or intent.
Tools like HummingDeck, DocSend, and Papermark all generate tracked links. Our DocSend alternatives breakdown covers the tradeoffs.
One thing to check: the link should work without requiring the investor to enter an email. If your tool asks investors to "sign up to view," investors will close the tab. That's where Suster's friction warning comes in.
Mobile-first design
Storydoc's 2026 data: 32% of decks are opened on mobile. For initial-outreach decks specifically, that number approaches 50%.
Most pitch decks are built for desktop projection. On a phone, the 10pt footnote becomes invisible, the team photo grid breaks, the four-column comparison table becomes a cramped mess. If your deck looks bad on iPhone, you lose a third of your audience on first impression.
Things to check on mobile:
- 24pt minimum for body text, 36pt+ for headers
- Hero slide readable at thumbnail size (the Slack and iMessage link-preview card)
- No four-column layouts; stack vertically
- No animations that don't render in a PDF viewer
- Charts that legible when the page fits in a palm
Open your own deck on your phone before you send it. Most founders never do this.
Watermarking: use it when the process requires it
Watermarking stamps recipient or tracking information across a document. Choose a formal VDR that supports it when counsel or the transaction requires that control. HummingDeck does not provide dynamic watermarking, and fundraising stage alone should not decide whether you need it.
When to move from a tracked deck to an investor data room
Keep the first send simple: one current deck on a personalized link for each investor. Prepare the underlying records early, but open a focused room when the investor asks for evidence or starts diligence.
Use the investor data room checklist to choose the documents and folder structure for pre-seed, seed, and Series A. Then review HummingDeck investor data room software to decide whether a lightweight Room fits the process. For what views, return visits, and additional-viewer signals can and cannot tell you, read how to track an investor pitch deck.
The cold email script that worked
The best way to understand a good cold email is to read one that worked. Even better: read one from the same founder that didn't.
Allie Janoch's v1 vs. v2
Allie Janoch is the founder of Mapistry, an environmental-compliance SaaS. In 2017 she emailed Jason Lemkin of SaaStr Fund to raise her seed. The first email got ignored. The second landed her a $2.5M seed round, with a reply from Lemkin within a few hours.
Same founder. Same recipient. Same company. Entirely different outcome.
Her first email (v1, no response):
- Didn't explain what Mapistry does (only alluded)
- Self-deprecating: "our traction is pretty lame"
- Weak ask: "Can I reach out to you in a few months…"
- No follow-up
Lemkin later said the first email "didn't tell me what the company did, didn't tell me why I should care, and didn't give me anything to react to."
Her second email (v2, $2.5M seed):
Subject: SaaS for environmental compliance - powering 3M, Tesla, etc. - Seed
Hi Jason,
What Mapistry does: We solve the compliance problem for industrial
companies. Environmental regulations are confusing and fragmented,
and current solutions require hiring expensive specialists.
Key highlights:
- Big-name customers: 3M, Tesla, Republic Services, and dozens of others
- ~15% MoM revenue growth
- [Current revenue figure] in ARR after [X months]
- Large, underserved industrial regulatory-tech TAM
We're raising $[X] to accelerate product development and expand our
customer base.
[Personalization paragraph: Referenced Lemkin's SaaStr conference
interview with Peter Gassner of Veeva; tied Mapistry's mixed
software-plus-services model to Gassner's vertical-SaaS philosophy
intentionally contrarian.]
Would you have 30 minutes for a phone call or coffee next week?
Best,
Allie
Lemkin's own analysis of why it worked:
"It does a good job of summarizing the opportunity, early customers and traction, growth profile, and market size. Perhaps just as importantly, it is truly personalized. Also for me at least, it's low drama. 'Do you have any time the week after next?' The confident and relatively data-rich but low drama emails work best."
Three changes turned a dead email into a $2.5M round:
- Real customer names replaced vague claims. "Big-name customers" is fluff. "3M, Tesla, Republic Services" is signal.
- Personalization tied to the investor's own work. Referencing Lemkin's published interview with Gassner proved Allie had done her homework. More importantly, she disagreed with the canonical vertical-SaaS line. Contrarian is memorable.
- Low-drama ask. "Would you have 30 minutes…" is specific but not urgent. No "limited time," no "raising this week," no desperation.
Janoch later shared that this template converted 33% of her cold emails into meetings across the full outreach sequence.
Three more that worked
Michael Bamberger, Tetra Insights → multiple B2B SaaS funds (2020, reply within 2 hours from first batch, led to $1.5M seed + $5M Series A).
Subject: RE: Pre-seed round? B2B, SaaS, qualitative data analysis
Hi ___,
I'm the co-founder of Tetra Insights, a B2B SaaS startup based in
Boulder. I'm reaching out to see if you're available for a 30-minute
call to learn about our company.
Our [transcription] software transforms audio and video into business
intelligence. We launched in January and generated [$X] in revenue
while in beta, [with MRR] from companies including LexisNexis, Yara
and Segment where our technology powers their customer insights
practices.
[Additional tailored paragraph on fit to this specific investor's
thesis/portfolio]
Would you be open to a 30-minute call next week?
Best,
Michael
Why it worked: the RE: subject line trick looks like a reply. The subject names stage, business model, and category in eight words. It's a thesis filter. The body names real customers (LexisNexis, Yara, Segment) for instant credibility.
Kemdi Ebi, Versus Africa (2021-22, seed round from Techstars-adjacent investors, ~$950K to date):
Hi [REDACTED],
Trust all is well. We're currently in our Seed round of investment
and see Versus Africa as a great fit with your investment thesis.
Key Highlights:
- Techstars-backed company
- Raised over $950k to date
- Full product built and owned 100% by us. Patent (pending) & Trademark secured
- Soft launched product Aug 2021 & revenue generating
- Top brands from vast industries are paying customers: Arsenal F.C, Vodafone, Safaricom, GIZ.
Please let us know if you need any further feedback on the
investment opportunity. We have also included a link to our deck.
Look forward to hearing from you.
Sincerely,
Kemdi Ebi, Co-founder/CEO, Versus Africa
Around 90 words. Five bullets, each earning its space: accelerator signal (Techstars), IP signal (patent, trademark), revenue signal (generating), trophy customers (Arsenal, Vodafone). No filler.
StoreCycle → Sapphire Ventures (2021-22, raised $15M seed):
[...] It's a data-powered multiplier that directly connects to the
store via API.
- We have over 1,000 Shopify merchants on our waitlist.
- We're ex-Thrasio brand management & supply chain ops.
It's a proven, profitable business model in the Amazon FBA space
($10bn valuation @ Thrasio), and we're joining a young and booming
industry on the Shopify side of things.
We're raising a $15m seed round to finalize our backend project
management solution and deploy capital for further acquisitions.
If you're interested in a quick chat, I'd love to explain how I
feel StoreCycle is a good fit for the Sapphire portfolio.
Two things worth stealing from this one: pattern-matching to a known outcome ("Thrasio, $10B valuation") gives the investor a mental anchor, and naming the specific fund in the ask ("fit for the Sapphire portfolio") proves the email isn't a blast.
What all four have in common
- Subject line names the stage, model, and category in under 10 words
- Body under 200 words
- Real customer names, not categories
- A specific 30-minute ask, not "let me know if you're interested"
- No sensitive details or NDA gate in the first contact, no BCC list, no "investment opportunity"
- Personalization tied to the investor's actual work, not their fund's About page
What investors actually do with your deck
Once your email lands, recorded review time varies by stage, platform, format, and session definition.
The 2-to-4 minute window
Published figures cluster around a few minutes, but they are not interchangeable. DocSend reports stage-specific figures, Papermark reports complete reviews on its own platform, and Storydoc studies interactive presentations. The practical lesson is to make the core case understandable quickly, then compare your own likely-human sessions instead of dividing a headline average evenly across every slide. See the full benchmark methodology and source links.
Where the time actually goes
The first page gets more than twice the time of any subsequent page. Papermark's 2024-25 analysis of 2,239 decks and 8M data points: the cover slide averages ~20 seconds; every later page averages ~15 seconds. Investors scan.
Among sections, team and financials get the most attention. Product gets the least. This is useful for structure: put your team and financials earlier than most templates suggest, because attention falls off a cliff after slide 10.
Bounce rates
Storydoc 2026: 31% of readers bounce within the first 10 seconds. Another 15% drop off within the first minute. But of the readers who reach slide 4, 82% finish the deck.
The first 10 seconds of your cover slide are doing most of the work. If your cover is vague (a logo, a generic tagline, a "confidential" watermark), you lose a third of investors before the hook. Your cover slide should answer three questions in five words: what it is, who it's for, why now.
Mobile
32% of decks are opened on mobile across all outreach. For initial cold outreach specifically, the number is closer to 50%. Desktop average: 4:51. Mobile average: 3:27. Tablet: 3:39.
The mobile viewer is not a lesser version. It's a meaningful share of first impressions. If your cover slide doesn't survive a 375px-wide preview, you're losing half your first-read audience.
The "95% unfundable" framing
SaaStr's AI Pitch Deck Analyzer, launched September 2025, has now ranked over 5,000 seed decks. Jason Lemkin's summary:
"Most founders are in the B range… A B- is not enough to get funded in today's market."
The specific numbers: 1.3% of decks score A+, 3.8% score A, and 95% are classified as unfundable as submitted. This doesn't mean 95% of companies fail to raise. It means most decks are rewritten before they close a round. If your first-pass deck doesn't get a response, that's data, not a verdict.
For a deeper dive into the benchmarks and what they mean, see our pitch deck benchmarks 2026 breakdown.
Follow-up timeline
Most founders either follow up too fast (within 24 hours, which reads desperate) or never follow up at all (and miss the rebound). The consensus range across fundraising writers (Allied VC, Sifted, Digify) is tight.
- Day 2: No follow-up yet. Give them time to actually read it.
- Day 7: First substantive follow-up. Reference one specific thing: a new customer, a metric update, a product ship. Don't just resend the deck.
- Day 14: Second follow-up. Same structure, different update. After this, you're on their radar.
- Day 21+: Move on. Add them to a quarterly update list. Investors who passed often come back at Series A.
Engagement data changes what you can prepare, not what you know. If a likely-human visit concentrates on one topic, make the next message relevant to it. If another viewer appears on a personalized link, offer context for other reviewers without claiming the link was forwarded or that the deal advanced.
If the deck has never been opened by day 5, your follow-up shouldn't be "wanted to check in." It should be "resending in case the first email got buried," with a one-line hook for why they should open it now. Never assume the first email was read.
When another viewer appears on your deck link
Tracking platforms observe uses of a link. They do not observe a forwarding action. A new viewer on a personalized link is consistent with internal sharing, but it can also come from another device or another stakeholder using the URL.
Use the signal as context for a low-pressure question, such as: “Happy to send supporting material or include anyone else who is reviewing.” Do not infer a partner meeting, term sheet, or pass from viewer count alone.
For the mechanics and limits of additional-viewer signals, see how to track an investor pitch deck.
12 common mistakes to avoid
A quick-reference checklist. Screenshot it before you send your next email.
- Putting sensitive material in the first-send deck. Most investors will not sign an NDA before an initial review. Start with a non-confidential deck, then follow counsel and the investor's process for sensitive diligence.
- Sending
.pptxor.keyinstead of PDF. Fonts break, sizes bloat, formatting shifts. Always export to PDF. - Oversized attachment (>10MB). Deliverability issues above 10MB. Compress images or host via tracked link.
- "Sent from iPhone" sloppiness. Typos, casual formatting, or mobile-auto signatures signal low care. Write from a desktop.
- Generic mass-blast with no personalization. Feld: "I can't tell you the number of people who pitched something and have no idea whom they are pitching it to." Do the homework.
- Vague subject lines. "Investment opportunity," "10 minutes of your time," "investment idea." All dead on arrival. The subject line is a thesis filter.
- Wrong stage match. Pitching a Series A fund at pre-seed wastes everyone's time. Know the check size before you write.
- Vague ask amount. Say exactly how much you're raising and how you'll spend it. "I'm open to discussing" is not an ask.
- Weak warm intro. Hunter Walk: one-third of the warm intros he gets are barely better than cold. A strong cold email beats a weak warm one.
- Following up wrong. Too fast (before day 7) = pushy. Never = no signal. Use the 2/7/14/21 cadence, modified by engagement data.
- Not mobile-optimized. 32% of decks open on mobile. If your cover slide is illegible at 375px, you're losing a third of first impressions.
- BCC'ing multiple investors. Transparent and disqualifying. Send one email per investor. Period.
Most fundraising posts have a mistakes section. Most are filler. Every mistake above has a named VC or founder saying exactly this, on the record. Don't take our word for it. Click through the citations.
What's different in 2026
Five things shifted in the last 18 months that change how pitch decks land.
AI-generated deck flood. Gamma, Tome, Beautiful.ai, Pitch, and Figma Slides all shipped AI deck-generation features between 2024 and 2026. The result: VCs see dramatically more decks with the same generic visual language. A 15-minute AI draft is not a pitch. The SaaStr AI Pitch Deck Analyzer has now rated 5,000+ decks; 95% are unfundable as submitted. The signal advantage now goes to decks that are visibly human-authored: founder voice, company-specific metrics, custom visuals that aren't stock-library polish.
Email security bot inflation. Microsoft Defender SafeLinks and Proofpoint URL Defense can inspect links in corporate emails before the recipient opens them. Those scans can register as views. An immediate, single-page event can be consistent with automation, but timing alone does not classify it. This is why suspected-automation filtering matters: review the classifier and the full event pattern before treating activity as likely human.
DocSend post-Dropbox. DocSend was acquired by Dropbox in March 2021 for $165M. Five years later, the product is still maintained (DocSend shipped a "Diligence Tracker" in February 2026), but the pricing has moved upmarket, and founders who used the old free-tier Send & Track feature have been cut off. The market repriced what DocSend charges a premium for; our DocSend alternatives breakdown covers the tradeoffs.
Carta's AI concentration. Carta's State of Private Markets 2025 report (published February 2026): AI startups captured 44% of all US startup capital in 2025. For non-AI founders, this doesn't mean you can't raise. It means your deck now has to work harder on defensibility, unit economics, and team depth. Investors are seeing AI-native decks hitting $50-100M ARR with under 100 employees. That's the efficiency benchmark you're being compared against, even if you're not in AI.
Prepare the room before diligence. Do not treat a branded room as a stage badge. Open a focused room when an investor asks for evidence or the process enters diligence. Choose a formal VDR when counsel requires per-file permissions, structured Q&A, NDA records, audit exports, SSO, or documented compliance controls.
FAQ
What format should my pitch deck be in?
PDF under 10MB. PowerPoint and Keynote files break fonts across machines and can be edited by the recipient. Export to PDF, check file size, compress images if you're over 10MB.
How long should my investor pitch deck be?
Use the fewest slides needed to make the stage-specific case clearly. Current platform reports use different cohorts and point to different ranges, so there is no universal slide count. See our source-audited pitch deck benchmarks for 2026 before applying a benchmark to your deck.
Should I send the deck before or after the meeting?
Before, if it's a cold intro: you want the deck to qualify the meeting. After, if the investor specifically asked for a meeting without seeing the deck first (rare, usually only with warm intros from people they trust).
How long do investors actually spend on a pitch deck?
Published benchmarks vary by stage, platform, format, and session definition. Several current reports cluster around a few minutes, but none is a universal average. Use the source-audited pitch deck benchmarks to choose the closest disclosed cohort, then compare it with your own likely-human activity.
Can I track who opened my pitch deck?
Tracked links record supported activity such as opens, time by page, completion, return visits, and additional-viewer signals. A personalized link separates activity by intended recipient, but an open does not prove who was behind the device or what they thought.
How do I know if a VC forwarded my deck?
An additional viewer on a personalized link is consistent with internal sharing, but it is not a forwarding record. It can also reflect another device or another stakeholder using the URL. Treat it as a prompt to ask whether anyone else needs context, not proof of who shared the deck.
Should I require an NDA before sharing my deck?
For the initial pitch, share a non-confidential deck without requiring an NDA. Sensitive diligence is different: follow counsel and the investor's process, and use appropriate access controls for the requested material.
How long should I wait before following up?
7 days for the first follow-up and 14 days for the second are common starting points. Use the investor's own response and your agreed process first. A likely-human visit can help you prepare relevant context, while no recorded open can justify one concise resend without proving the email was missed.
The short version
If you only remember five things:
- PDF under 10MB, via a tracked link that doesn't require an email gate.
- Email under 200 words. Real customer names, specific 30-minute ask.
- Lead with one tracked deck. Open a focused room when the investor asks for evidence or begins diligence.
- Follow up at day 7 and day 14. Use engagement data to adjust timing.
- Treat return visits and additional viewers as context, not proof of intent or forwarding.
The bar is higher in 2026. More AI-generated decks, more bot-inflated view counts, more capital flowing to a narrower set of companies. Most of what separates founders who get funded from founders who don't is boring: the format, the email, the data room, the follow-up. None of it is glamorous. All of it is learnable.
Related:
- Investor data room checklist and folder structure: prepare the evidence investors request and open the room at the right stage.
- How to track an investor pitch deck: interpret recorded engagement without treating activity as identity or intent.
- Pitch deck benchmarks for 2026: compare deck length and reading patterns against source-audited cohorts.
- Why your deck analytics are wrong: understand how email security scanners can inflate view counts.
- DocSend alternatives for fundraising: compare fundraising workflows after defining what your raise actually needs.