The short answer
To spend the YC $500K well, plan it as a 12-month budget that buys evidence for your next round: roughly a third on product and engineering, a quarter on founder salaries, 10 to 30% on go-to-market, around 5% on tools and the rest held in reserve. Put the go-to-market share into one system owned by one person, with the founder as the voice, rather than into channels bought one at a time.
Key takeaways
- The YC standard deal is $500,000: $125,000 on a post-money SAFE for 7%, plus $375,000 on an uncapped SAFE with an MFN provision.
- Carta data puts the median time from seed to Series A at 2.2 years in 2025, so the $500K is the first leg of a longer runway, not the whole of it.
- Choose a Product-heavy, Balanced or GTM-first split based on where your evidence gap is, and always keep a reserve.
- Go-to-market money at seed buys a foundation, an account list, proof assets and controlled tests, not volume.
- Spending it through one team with one accountable lead stops the budget leaking across channels that nobody connects.
The money arrives fast. YC commits to its standard deal the day a company is accepted, and from then on you are spending $500,000 that comes with no instructions attached. If you have never run a budget this size, nothing in the deal tells you how.
This article gives you one. It sets out what the deal is, how long the money needs to last, three ways to split it and a worked example of monthly burn. It then covers the line founders get wrong most often, which is go-to-market.
What do you actually get in the YC $500K deal?
You get $500,000 in two parts. According to YC's deal page, $125,000 comes on a post-money SAFE in return for 7% of your company, and $375,000 comes on an uncapped SAFE with a most favoured nation (MFN) provision.
The MFN SAFE converts in your priced round on the terms of the lowest cap SAFE, or other most favourable terms, issued from a set start date. In practice that means the $375,000 takes the best terms you give later investors. YC also takes a pro rata right to invest in later rounds, and it charges no fees.
What the deal does not include is any rule on spending. That freedom is useful, and it is also why so many founders spend the first three months reacting to whatever feels most urgent.
How long does seed funding need to last?
Plan for well over a year. Carta's data puts the median time from seed to Series A in 2025 at 2.2 years, and 1.9 years for AI companies, so the $500K funds the start of a long road and has to buy the proof that pays for the rest of it.
The quick raises in the headlines are the exception. In Carta's 2025 data on more than 3,000 US startups, only 15% raised a Series A within a year of their seed round, and 39% took three or more years to get there. Carta's own advice is to plan for seed money to last about 1,000 days, or 2.7 years.
Paul Graham's test for this is whether you are default alive: if expenses stay constant and revenue keeps growing at its recent rate, do you reach profitability on the money you have left? You do not need to be profitable at seed. You do need to know which side of that line your plan puts you on.
“Hiring too fast is by far the biggest killer of startups that raise money.”
That is the warning behind every budget below. The biggest decisions in a seed budget are headcount decisions, because salaries repeat every month and are the hardest line to cut.
How should you allocate seed funding?
Allocate it against your biggest evidence gap. If you cannot yet show that the product works, product and engineering takes the largest share. If customers already use it and you cannot show repeatable demand, go-to-market moves up.
The three splits below are a recommended framework, not market data. Each divides the same $500,000 across five lines. Founder salaries stay at a quarter in every version, because two founders who cannot pay rent make worse decisions, and the reserve never drops below 10%.
Three ways to split the YC $500K
For teams still proving the product works, with heavier compute and infrastructure costs.
For teams with design partners using the product and a need to show demand beyond the founder's network.
For teams with a working product and paying customers whose next round depends on pipeline.
Codax recommended framework, shown as percentages of the $500,000 total. Not market data.
What each line covers
- Product and engineering: any engineers beyond the founders, contractors, design and the compute the product itself consumes.
- Founder salaries: enough for each founder to live without a side income, and no more.
- GTM people and programmes: the team or provider that builds pipeline, plus events, content, data and paid distribution.
- Tools and infrastructure: hosting, CRM, sending infrastructure, analytics and the software the company runs on.
- Reserve: cash held back for a hire that cannot wait, a slow month or a raise that takes longer than planned.
Keep legal, accounting and incorporation costs inside tools and infrastructure, or take them from the reserve. They are small next to salaries, and they should not quietly eat into the GTM line.
Most B2B companies with a handful of design partners belong in Balanced. Move to GTM-first when the product question is answered and the question investors keep asking is where the next 50 customers come from. Stay in Product-heavy only while the product itself is the risk.
What does a $500K budget look like month by month?
On the Balanced split, a 12-month plan burns about $35,400 a month and leaves $75,000 in reserve. The arithmetic below is a worked example, so you can repeat it with your own numbers.
Fourteen months is far short of the 2.2 year median to Series A. That is the point of the exercise. The $500K is not meant to carry you to the A on its own. It is meant to produce the traction that lets you raise the rest of your seed on good terms.
If you want the same money to last longer, stretch the plan and the burn falls. The table shows the same $425,000 of planned spend over three lengths.
Worked example: the same $425,000 plan at three lengths
| Plan length | Monthly burn | Pay per founder per year | Months on $500K |
|---|---|---|---|
| 12 months | $35,417 | $62,500 | 14.1 |
| 18 months | $23,611 | $41,667 | 21.2 |
| 24 months | $17,708 | $31,250 | 28.2 |
Founder pay is where the trade-off bites. Pilot's 2026 Founder Salary Report, from 1,623 founders, puts the median salary at $103K for companies that have raised $100K to $999K. The 12-month plan gets close to that only by splitting $125,000 between two people. Decide this line first, because it is the one you will least want to revisit.
How much should you spend on marketing at seed?
At seed, plan go-to-market as a share of the raise, not of revenue. In the framework above that is 10 to 30% of the $500,000, or $50,000 to $150,000 across the year, covering both the people and the programmes.
Revenue benchmarks do not translate to this stage. SaaS Capital's 2026 benchmarks, from more than 1,000 private B2B SaaS companies, put median spend at 15% of ARR on sales and 8% on marketing. Those percentages assume millions in ARR. A seed company with a few design partners has almost no ARR to take a percentage of, so a revenue-based budget rounds to nothing.
The same report found that equity-backed companies spend 100% more on marketing than bootstrapped ones. Investors expect you to spend on growth. They expect that spend to produce evidence.
What does go-to-market spend buy at this stage?
At seed, go-to-market spend buys the ability to sell without the founder in every first conversation. It pays for a foundation that converts, an agreed account list, written proof and a small number of controlled tests. It does not buy volume, and it should not try to.
Where the GTM line goes, in order
Repair the foundation
Fix the site, the CRM and lead definitions, sending domains and tracking, so every later pound and dollar is measured.
Agree the account list
Name the accounts you will pursue this year, and the roles inside each buying group, with whoever sells.
Turn proof into assets
Write up design partner results as case studies and capture the founder's thinking as executive content.
Run controlled tests
Start each channel as a four to six week experiment against the account list, with one variable changed at a time.
Move money to what converted
Put the next quarter's budget into the channels and messages that produced qualified pipeline.
This order matters because tests read nothing on a broken foundation. Sending outbound before the domains and tracking are fixed burns the list and teaches you nothing.
It also explains why the GTM line looks small next to product. At seed you are not paying for reach across a whole market. You are paying to learn, quickly and cheaply, which accounts, roles and messages turn into qualified pipeline, so that the next round can fund more of what works.
The proof you already have is usually the biggest unused asset. Design partners who would speak on a webinar, a partner programme you joined and never mentioned, a founder who explains the problem better than any landing page. Turning those into assets costs far less than buying new attention. Our 90 days after Demo Day plan sets out the same sequence as a calendar.
Why do channels bought one at a time waste the budget?
Channels bought one at a time waste the budget because nobody owns the result across them. Each provider optimises its own number, and the founder becomes the only person trying to connect them.
The usual pattern is a performance agency for ads, an outbound or SDR agency for sequences and a content studio for posts. Each is a separate contract with a separate report. None of them reads what sales hears, and none of them can move money to another channel when theirs underperforms.
The bigger cost is that buyers do not arrive through one channel. In Codax's cybersecurity services engagement, 8 in 10 opportunities were touched by three or more channels, and it took an average of 5 touches before the first call. Three vendors each see one touch and claim none of the outcome.
- Three providers report three different numbers, and none of them is qualified pipeline.
- Messages drift, because nobody writes from one account list.
- Budget stays where the contract put it, not where the results are.
- The founder spends hours a week coordinating vendors instead of selling.
Our comparison of your first GTM hire options covers the hiring version of the same choice.
Where does a growth department fit in a seed budget?
A growth department fits in the GTM people and programmes line, as one team instead of several hires and vendors. A growth department is one senior team that owns qualified pipeline end to end, from strategy to execution, under a single accountable lead.
At Codax that lead owns the qualified pipeline number and can move budget between channels within the quarter. The account list is agreed with whoever sells, positive replies reach the founder the same day and pipeline is reviewed account by account every month. The full model is on how we work.
The founder stays the voice. Codax asks a founder for 1 to 2 hours a week, captured from recorded interviews, and turns that into content, invitations and outreach sent in the founder's name. The founder stays the sender and the closer while the team builds everything around them.
Compare that with the alternative on the same budget. A $100,000 GTM line buys part of one experienced hire, or a junior hire plus a vendor or two. Either way, the founder ends up managing the pieces. One team under one lead puts strategy, content, outbound, events and measurement behind a single number, and the founder's job narrows to the conversations only they can have.
The case for this is in the numbers. In the agentic AI healthcare engagement, the CEO as the sender produced 3x the replies, and the company closed $1.02M ARR in seven months from LinkedIn and email with 0 hires added. In the cybersecurity services engagement, founder-signed invitations produced 3x the acceptances and founder ads beat company ads on click-through by 2.4x.
How do you know the budget is working?
You know it is working when qualified pipeline grows month on month from accounts outside your own network. Review it every month, account by account, against the plan and the burn.
- Recalculate monthly burn and months of runway against the plan.
- Rerun Paul Graham's default alive test with the latest revenue growth.
- Count qualified pipeline by source, and how much came from outside the founder's network.
- Move GTM money from tests that missed to tests that converted.
- Release reserve only for a decision you would defend to your next investor.
Investors at the next round ask for these numbers. Our guide to the pipeline metrics Series A investors want covers which ones matter most. The full series for accelerator founders is on the accelerator founders hub.
Questions and answers
How much money does YC give startups?
YC's standard deal is $500,000 in total. $125,000 comes on a post-money SAFE for 7% of the company, and $375,000 comes on an uncapped SAFE with a most favoured nation provision. YC charges no fees.
How long should a seed round last?
Plan for well over a year. Carta's data puts the median time from seed to Series A at 2.2 years in 2025, and 1.9 years for AI companies. Carta recommends planning for seed money to last about 1,000 days.
How much should a seed startup spend on marketing?
Plan go-to-market as a share of the raise rather than of revenue, because seed companies have little revenue to measure against. A useful range is 10 to 30% of the round across people and programmes, weighted by whether your biggest gap is product proof or demand proof.
Should founders pay themselves from YC money?
Yes. A budget that leaves founders unable to cover living costs pushes them into worse decisions. In the worked example above, founder salaries take a quarter of the $500,000, which is $62,500 a year each for two founders over 12 months.
What is the biggest mistake founders make with seed money?
Hiring too fast. Paul Graham calls it by far the biggest killer of startups that raise money, because salaries repeat every month and are hard to cut. The second is spreading go-to-market money across channels that nobody owns together.
Sources
- YC's Standard Deal, Y Combinator
- Planning to Raise VC Every 18 Months Is Planning to Fail, Carta
- Ignore Headlines About Startups Raising A Rounds in Six Months, Carta
- Default Alive or Default Dead?, Paul Graham
- Founder Salary Report 2026, Pilot
- 2026 Spending Benchmarks for Private B2B SaaS Companies, SaaS Capital
