Long-form notes from inside live rounds.

Slide 1 — Title. Company, one-line wedge, round and amount. Not a logo on a black background. If a partner closes the deck after slide 1, they should be able to tell their colleague what you do and what you want. Most decks fail this test.

Slide 2 — The bet. Not the problem. The specific thing you believe that other smart people don't. Investors get paid for variance, and variance lives in the bet. If your bet reads like consensus, the deck is dead by slide 5.

Slide 3 — The wedge. Who you are for, doing what, instead of what. One sentence. The wedge is narrower than your TAM slide will later suggest, and that is correct. Win the niche; the expansion path is slide 9.

Slide 4 — Why now. A real change in the world: regulation, model-cost curves, distribution unlocks, behavior shifts, GTM channel openings. 'AI is hot' is not a why-now. 'OpenAI's Realtime API made our wedge possible in March 2025' is.

Slide 5 — The product. One screenshot. Not a Figma mockup, not a ten-second loom in a placeholder, not three feature tiles. A real screen, with the magic moment captured. Partners can tell mocks from real product faster than founders think.

Slide 6 — Traction. The curve, not the number. Slope is the asset; the y-axis label is just punctuation. New, expansion, and churn shown separately when you have it. Pilots labeled as pilots; LOIs labeled as LOIs. The cleanest read in the deck is also the cheapest credibility you'll buy.

Slide 7 — Market. Bottoms-up. Accounts × price × attach. Cite the source for each multiplier and you'll skip a 20-minute partner meeting argument about TAM math. 'Per Gartner' is not a citation; it's a tell.

Slide 8 — Business model. Price, who pays, how you'll know it's working. NRR target if you're past seed. Not your full pricing page; the unit and the trajectory.

Slide 9 — Expansion. Land here, expand here, then here. Three steps. Not nine. The expansion slide is where partners decide whether the wedge is a company or a feature.

Slide 10 — Team. Why this team for this problem, named explicitly. LinkedIn links live, recognizable past logos. Acknowledged gaps with the hire plan. 'Hiring head of sales with this round' is a sentence that closes meetings; it does not lose them.

Slide 11 — The ask. Round size, instrument, current commitments, close date or process timeline. 'Closing in six weeks' creates urgency. 'Open-ended' kills it. If you have a lead in conversation, say so without naming the firm.

Slide 12 — Vision. One slide, one sentence, one year-five image of the world if you win. Not a roadmap. Roadmaps live in the appendix.

Appendix — Cohort detail, hiring plan, competitive map, financial summary, prior round terms, customer references with permission. The appendix is not where decks die; it's where partners go to sell internally. Treat it like the asset it is.

1. 'We're disrupting the multi-trillion-dollar X market.' Trillion-dollar TAMs read as unfocused. Lead with the wedge; let the partner do their own TAM math after.

2. 'There are no real competitors.' Either you haven't looked or you're solving a problem nobody has. Both are bad. Name three competitors and how you'll lose to each one before you win against them.

3. 'Our hockey stick begins next quarter.' The hockey stick that lives in the next quarter is the hockey stick that is in the deck two quarters from now. Show the slope you have, not the slope you've drawn.

4. 'We just need to be 1% of the market to be a $X company.' This is the phrase partners use to teach analysts what bad TAM math looks like. Delete it.

5. 'We're like Stripe for X.' Sometimes true. Usually a tell that the wedge is borrowed, not earned. If you must analogize, do it once, late.

6. 'We're going to build a category.' Categories are observed in retrospect by analysts. Founders who build categories rarely use the word in seed decks.

7. 'We have several LOIs from Fortune 500 companies.' If they're real, name two. If you can't, the LOIs are weaker than the deck's tone implies.

8. 'Our team has 50+ years of combined experience.' Combined-experience math is a flag that the individual résumés don't survive on their own.

9. 'We're raising our seed round at a $40M post.' Anchoring high without traction reads as either misreading the market or expecting a partner to teach you about it. Neither closes.

10. 'We'll figure out monetization later.' The seed market in 2026 is not the seed market of 2021. Have a price point and a payer in the deck, even if the model isn't proven.

11. 'We have an unfair advantage.' Founder-as-asset is rarely as defensible as the deck claims. Name the advantage and how it compounds; otherwise omit.

12. 'We're cash-flow positive.' Often true on a thin definition. State the definition before a partner asks: contribution-margin positive, fully-loaded, including founder salaries — be specific.

13. 'Our churn is essentially zero.' Either you have nine months of data and 30 customers (true, also not informative) or your definition of churn is generous. Show GRR by cohort.

14. 'We were oversubscribed in the last round.' The bar for this phrase is higher than founders realize. If the round was $1.5M and closed in five weeks, it was 'on plan,' not oversubscribed.

15. 'We can scale to $100M ARR with this team.' Almost no team scales to $100M without 30+ hires post-Series A. The phrase reads as either naïve or evasive.

16. 'It's a winner-take-all market.' Most aren't. Markets segment. Phrasing it this way puts the partner on edit instead of underwrite.

17. 'Our biggest risk is execution.' This is true and also tells the partner you haven't thought through capital, GTM, hiring, or pricing risk. Replace with the second-most-important risk and how you'll know it's manifesting.

18. 'The market is highly fragmented, which is our opportunity.' Fragmented markets are usually fragmented for structural reasons. Name the structural reason; explain why it's changing now.

19. 'We're not raising right now, just having conversations.' If a partner believes this, they read slower. If they don't, you've spent a tell on nothing. Pick a posture and own it.

20. 'Happy to share the financial model — it's quite detailed.' Detail isn't the bar. Defensibility is. Send a one-page summary and an appendix-grade model in the same email.

Send monthly. The 1st or 2nd of the month, before you have time to soften the prior month's truth. Plain text email. No PDF, no Notion link, no Loom. Subject: 'Company update — Month YYYY.'

Section one — Highlights, three bullets. The wins that matter, not the wins that exist. A logo, a hire, a milestone, a product unlock. If you have to write a fourth, demote one.

Section two — Lowlights, three bullets. What broke, what slipped, what you got wrong. This section is the one that builds trust and the one founders skip. Skip it twice and your investors stop trusting the highlights.

Section three — Metrics. Five lines max. Whatever you committed to last month, in the same units, with the prior period delta. ARR, new customers, gross margin, cash on hand, runway in months. Add one line for the metric of the quarter — the thing the next round will be underwritten on.

Section four — Asks. Two bullets, specific. 'Intros to heads of platform at logistics SaaS' beats 'always looking for intros.' The narrower the ask, the more it gets done. Investors who can't help on a specific ask will tell you, and that data is also useful.

Section five — What's next. Three bullets, no more. The things that will be in next month's highlights or lowlights. This section is your public commitment device, which is exactly the point.

Cadence rules. Skip a month and you owe the audience an explanation in the next one. Acquired or dead are the only acceptable reasons to stop sending. Existing investors get the full update; prospective leads get a 60-day-old version on request — keep them warm without resetting their information advantage on the current round.

What this compounds into. By month 18, you've built a written record that is easier to fundraise from than any deck. Partners ask for the last three updates and the round is half-underwritten before the first call. This is the closest thing to a fundraise cheat code that exists, and most founders never get to it because the third month is the hard one.

The wedge slide answers three questions that every other slide depends on: who, doing what, instead of what. Without those three, the problem slide is abstract, the product slide reads as feature soup, the GTM slide has no ICP, and the market slide has no anchor multiplier. The wedge is the slab the deck stands on.

Reason one founders bury it: it feels small. A wedge that fits on one slide always feels small relative to the vision. That tension is the point. Partners who fund category-defining companies almost always fund a wedge that, on paper, looks like a feature. Linear was a wedge against Jira's frontend; Stripe was a wedge against PayPal's API; Figma was a wedge against Sketch's lack of multiplayer. Each was small enough to be defensible against a giant; each was big enough to compound.

Reason two: the wedge requires a real choice. To name an ICP is to deselect twelve others, and seed founders are still in the optionality phase mentally. The deck reads as hedged because the founder's strategy is hedged. The work to write a one-sentence wedge is mostly the work to make the choice the wedge expresses.

Reason three: the wedge is testable. Once it's written down, partners can tell whether you're winning it. 'We help mid-market e-commerce ops teams cut return-shipping time' is testable: you have either signed mid-market e-commerce ops teams or you haven't. 'We're building the future of commerce infrastructure' is not testable, which is also why it doesn't close.

How to construct it. Start with the customer call you most recently closed and write down the three things that customer said in their own words about why they bought. Strip every word that wouldn't make sense to a smart generalist. Cut the result to one sentence. That sentence is the working draft of your wedge slide. If you can't do this exercise — because the call doesn't exist yet, or the words from different customers don't rhyme — that finding is also the deck's finding, and the wedge slide is premature.

Where it goes. Slide three. Before product. Before traction. After the bet (slide two), because the wedge is the bet rendered as a customer. After the title (slide one), because the title earns the right to ask for thirty seconds of attention.

What it doesn't do. The wedge slide does not pre-empt the vision slide on slide twelve. The vision is allowed to be larger; the wedge is the part that compounds toward it. Partners read decks with a discount factor on vision and an amplifier on wedge. Order accordingly.

Bridge rounds in 2026 are common, structurally fine, and reputationally fragile. The fragility is almost entirely linguistic. Founders who frame the round well close it on the same SAFE in three to five weeks; founders who don't spend nine months and lose the option to compound through it.

Phrase to retire one — 'extension to our seed.' This phrase tells the partner that the prior round did not produce the milestones it was raised to produce. Sometimes this is true and the right move is to own it on slide eleven. Often it's not true: the prior round produced the wedge, the early traction, and the platform on which the next product is being built. In that case, the round is not an extension; it's a Seed-2, and the language should reflect that.

Phrase to retire two — 'a small top-up.' Small top-ups read as 'we are not in control of capital strategy.' If the round is $1M on $20M post into an existing SAFE pool, say so. Specifics are reassurance; vagueness is alarm.

Phrase to retire three — 'just to give us more runway.' Runway is a constraint, not a use of funds. Use of funds maps to milestones. If the milestone is to ship multi-tenant and close five enterprise design partners before the priced round, that is the use of funds, and the runway is a side effect.

Phrase to use one — 'A SAFE on top of our existing pool, capped at $X, into the next eighteen months of milestones.' This is the cleanest version of the truth most bridge rounds are. It is also the version that is easiest for the partner to write a check against because it is closest to how the cap table will look the day after the wire.

Phrase to use two — 'We have N of the $M closed from existing investors.' Existing-investor commitment is the single best signal you can put on the bridge slide. If you have 60% of the round soft-circled from existing, lead with that, by name, with permission. If you don't, work the existing investor list before the deck goes out and come back with the number.

Phrase to use three — 'The priced round will be raised on [specific milestone] in [Q-quarter]'. Bridges only make sense if the priced round is named. Named, milestone-anchored, and time-bound. 'When the metrics are right' is not a plan; it's a wish.

Structural moves that change how the round reads. First, anchor the cap. A capped SAFE with a discount that meaningfully respects the prior round's holders reads as a founder who will fundraise again with this audience and knows it. Second, bundle the use-of-funds and the milestone in the same sentence. Third, write the round close date down. A round with a close date in the deck is a round; a round without one is a wish.

What partners are actually solving for. Partners on the existing cap table are solving for whether you've absorbed the lesson of the prior round and whether the next round is underwritable. New partners on a bridge are solving for whether they're paying a premium for information they don't have. Both audiences read the same deck. Write for both — the existing investor reads it as a check-in, the new investor reads it as a first-meeting deck. That dual-audience constraint is exactly why bridges are linguistically hard, and exactly why getting the language right is most of the work.

Appendix slide one — Cohort detail. Monthly cohorts, by month of acquisition, retention curves at 1, 3, 6, 12 months. Logo retention and revenue retention shown separately. If you have less than six months of data, say so on the slide. Partners build their memo's 'durability' section from this slide, almost always.

Appendix slide two — Hiring plan. Twelve names, twelve roles, twelve start-quarter targets, twelve cost loads. Not a headcount bar chart. The plan reads as either 'this team has a real GTM org by Series A' or 'this team is going to be the same five people in eighteen months,' and partners can tell the difference.

Appendix slide three — Competitive landscape. A 2x2 if it actually clarifies, a feature matrix if it doesn't. Three competitors with one sentence each on how you lose to them, three with one sentence each on how you win. Partners cite this slide internally to neutralize the 'won't another player win this' objection in the partner meeting.

Appendix slide four — GTM motion detail. The funnel by stage, conversion rates between stages, time in stage, ACV by segment. If you're pre-revenue, the design partner pipeline with status, last touch, and next milestone. Partners build the 'GTM is real' line in the memo from this slide.

Appendix slide five — Financial summary. Three years out, top-line, COGS, S&M, R&D, G&A, EBITDA, cash. One column per year. Not the seven-tab model; the summary partners can sanity-check against the wedge.

Appendix slide six — Customer references. Two named customers with permission, three sentences each on use case and outcome. Phone or email contact for the partner to verify, after the second meeting. The partner who calls a customer reference closes faster and discounts less. The slide that makes that call easy is the slide that compresses the round.

What not to put in the appendix. Press logos. Awards. 'Thank you' slides. Team-bonding photos. Each of those is a tell that the founder doesn't know what the appendix is for. The appendix is for the partner, not for the founder's confidence.

Order matters. Put cohort detail first because that's the slide partners ask for first. Put hiring and competitive next because those are the two objections that kill rounds in the partner meeting. Put financial and customer references last because those are the slides that close the meeting. The order is the order partners need answers in, not the order founders feel like building.