Many outcomes important to startups are gated by pitching: the ability to quickly tell someone about your company and make them intrigued enough to want to learn more. You pitch your company every day to everyone—potential employees, investors, and prospects.

This is not a natural skill for many founders. Stripe Atlas has helped a few thousand companies get started and assisted dozens with refining their pitches. We distilled these lessons to help the community, and they focus on pitching investors, from the perspective of an early-stage company.

We’ve worked with Y Combinator (YC)—the startup accelerator—to update this guide, with input from YC CEO Michael Seibel and YC founders who started their companies on Stripe Atlas.

Know your audience

Put yourself in the shoes of the person receiving your pitch. They’re well educated, savvy about startups, and sharp about business. They want to like you—this is a job for optimists.

They won’t know this field as well as you do. They have not breathed the problem space like you have. The jargon and industry assumptions may be new to them.

Explain assumptions in your pitch like you would to a smart friend in a different field. Be explicit about connecting dots.

They’re also highly likely to be overwhelmed.

Yours will not be the first pitch they’ll see today, or the last. Reviewers at top accelerators and early-stage VC firms see hundreds of pitches every day, generally only for about two minutes each. Your objective is to pack as much positive signal into those two minutes as possible and ensure the reviewer retains one core idea to communicate to someone who asks about you.

The review process continues for days on end. It is exhausting. Far too many pitches retread the same tired ground.

Help your reviewer quickly understand (and remember) what parts about your company are exceptional.

Sell yourself and your team

Most businesses are not impressive at conception. That is what early stage means. Reviewers will use you as a proxy for whether you will plausibly create a massive business given the opportunity.

The hardest or most impressive thing you’ve ever done, even if not strictly professionally related, should be in your application. This is especially true if it is “off the beaten path.” The applicant pool is replete with founders who have been conventionally successful in conventional ways. They were good students, they studied hard, and they went to excellent schools.

Many of them are going to have a very difficult time adjusting to life in a startup, where there is no set curriculum, there are no points awarded for effort, and there is no teacher establishing and enforcing deadlines.

Investors know this; they’ve seen it before.

If you’ve ever succeeded in something that is obviously difficult or impressive, say so. Olympic-level swimming doesn’t sell enterprise software, but it suggests you are no stranger to hard work over long timescales. You don’t have to be a Nobel Prize winner either—complex projects with many moving parts that you brought to completion despite obstacles are also interesting, since that’s a startup in a microcosm.

If you’ve made money on the internet before, even in ways you’re vaguely embarrassed about (affiliate marketing, eBay arbitrage, etc.), that is a very useful signal.

Mention the most impressive thing you’ve ever built. Startups make products and convince people to use them, and the first skill is considered harder to learn from scratch; hopefully you’re not starting from scratch.

If you can’t point to either an impressive project or sterling engineering credentials, aggressively sell your technical ability by pointing to specific evidence of it. In general, this will sound less like “I have a degree in a technical field” and more like:

The sort of engineer with this anecdote probably wouldn’t get hired at Google, but the anecdote demonstrates enough skill to ship software, enough product sense to make something people want, and a certain scrappiness in the face of constraints. Those are more important than the fact that that app probably took a day to write and has two dozen users total.

Don’t sell yourself short. We all have doubts, insecurities, and places where our skill set isn’t where we want it to be. You don't have to belabor these in your application. Remember, you only have two minutes. Describe what you’re doing and what you’ve accomplished, not what you couldn’t do and couldn’t accomplish. Failure is only as relevant as the insight you drew from that experimental result.

You’re likely applying as part of a team. Explain how the team knows each other—a key risk factor for startup teams is that they will break up during one of the many, many stressful moments the next few years will bring, and close relationships mitigate against this risk. Explain how your skills complement each other. If your team is heavily skewed in a particular direction (all hard-core technologists but no demonstrable sales ability, for example), either explain how being lopsided helps obviate weaknesses in other areas, or explain your plan for bringing on someone to help you balance the team.

You also want to convey a sense of urgency and velocity. Successful startup founders work on projects that take years but also cover a heck of a lot of ground in any two-week period. If you’ve ever done an ambitious project on a ludicrously short time frame, that is a good story to tell. Your current project should show an “impressive amount of work done given the period of time you’ve had to do it,” to paraphrase Michael Seibel of YC.

Communicate concrete details

Include a concise description of what exactly you are building and who should use it.

The most common problem by far among the pitches Stripe Atlas reviews is a lack of clear detail about what is being built. Reviewers use ability to describe what one is doing as a heuristic to determine whether one can successfully do it.

This pitch says nothing, in 18 words or less:

This doesn’t evince enough understanding of either machine learning or ecommerce to make a reviewer think the person who wrote it could successfully build this product. It also doesn’t say what the product actually is. Be particularly explicit about what the end-user experience looks like. Your reviewer is likely an early adopter and product enthusiast; they want to be able to envision themselves using the product.

Here’s the same product, again in 18 words or less:

This is packed with concrete, compelling detail. This lets the investor immediately visualize the product by comparing it to a familiar one. It says a product or prototype exists (note the past tense “built”). It mentions the tech stack (Magento) and alludes to the market for the product (nontechnical SMBs who use Magento). It demonstrates that the founders understand ecommerce as a business model. It gives two compelling reasons to use the product. It also avoids marketing fluff (like “cutting-edge”), which pitch-weary reviewers are utterly immune to. This pitch makes the reviewer want to know more.

Clarity is particularly important when you’re tackling recently popular ideas, like blockchains or machine learning. Some founders interested in these concepts might be world-class experts; many more read about them in the newspaper and cannot discuss them in more depth than a newspaper article. You want to quickly demonstrate that you’re not just another founder chasing the fad.

When you’re talking about your product, you want to describe exactly what you have now and what is single-digit weeks away. You can include longer-term aspirations when talking about your market.

You can almost never go wrong with stating facts and making those facts more concrete. You are expected to be an expert in your space or on your way to becoming one. If you know the number, say the number. If you can avoid a flight to abstraction by including specific details, give the specific details. This helps establish that you are indeed informed enough about this field and market to spend the next few years of your life improving it.

Target an attractive market

Investors are optimists. They’ll project success for your business and try to envision what the success case looks like. A company whose success case is too small can be an excellent business but a poor investment.

Venture investors are looking for companies that can sustain revenues of hundreds of millions of dollars per year, minimum. “Niche” products where the ceiling is millions of dollars are only interesting to the extent that they unlock adjacent, bigger markets. If you’re building Altair Basic, give some thought to convincing the investor, “Today, it’s an interpreter for hobbyists. Tomorrow, it’s an operating system for all microcomputers. We’ll call it Windows.” You don't have to fill in the blanks between Windows 1.0 and present-day Microsoft; who could possibly have known that in advance?

Many investors want to invest in not just a winner in a large market but the winner. You want your success case to result in you dominating the market, not sharing it. This means both that you need to displace current players and that your product or go-to-market strategy needs to be able to build a moat against someone doing the same to you.

A fragmented market that could be centralized is always interesting. Taxis before Uber or hotels before Airbnb are both great examples; the core innovation in both cases was a transformatively better product that had strong network effects. This turned geographically disparate many-players markets into a single worldwide winner-take-most market.

Some markets are obviously billion-dollar markets without elaboration: cancer medication, search engines, online advertising, etc. If your market is not obviously a billion-dollar market, show a sketch of math on why it is (or a citation to a credible source that says it is) or an argument as to how winning your niche market means you’ll win a larger, more attractive market.

Fermi estimates are good ways to concisely demonstrate market size and quality of analysis. Start with a tight bound on the number of customers there are. Multiply by your share of the market (given success). Multiply by how many interactions they’ll have with the product per year. Multiply by the revenue (or margin, if margins are slim in your industry) associated with each interaction. Although you might think claiming a larger potential customer base is better, you’re better off claiming a tight, well-defined customer segment; this suggests that you’re likely to aggressively optimize your product and marketing for them and win them, rather than making a mediocre attempt at reaching everyone and delighting no one.

A thumbnail sketch is good; a rigorously tuned financial model is probably overkill. Leave the reviewer with the impression that you’re able to make good snap judgments about opportunities; you’ll be doing that a lot.

Share unique insight

Doing a startup means you are going to spend several years learning every nook and cranny of a problem space. You should already be thinking about this problem, in detail, and doing your research, both in reading voraciously and in talking to potential customers.

Make it obvious to your reviewer that you’ve already started this work.

Your reviewer is smart but likely not an expert in your space. Reading your pitch should teach them things, including things that are not obvious to a well-educated individual. Much like conversations with customers, a novel insight leaves someone with a reason to remember you and think of you fondly.

Here’s an example from someone selling learning management software to startups:

The statement “Accountants at startups have similar tenure to fast-food cashiers” is a striking fact. Most people, even those in startups, have never had to think about that before…and the implications come quickly after it. Of course continuing education for normal accountants is not responsive to their needs. Now that you’ve demonstrated that insight, you can talk about how it drives specific product, marketing, and sales decisions.

Your pitch should also explain how your take on this space is unique. Acknowledge your competitors—if you don’t have any, that suggests that either you’re targeting a very unattractive market or (more worrisomely) that you haven’t done even minimal research on the space. Explain how you’re different and what that accomplishes for you.

Focus on nascent greatness

There are many great businesses in the world that are not great startups. People who make equity investments in startups do so because they believe there is a small chance that the company grows to be massive.

Investors remember what the current giants of the industry—Google, Facebook, etc.—looked like in their earliest days. It was not typically like a well-rounded business run by a competent operator that was doing pretty much everything acceptably. No, the huge companies of today looked like they were shoestring operations doing almost everything wrong run by very unlikely people…but they had a little seed that was just irresistible.

Facebook was a hacked-together toy that could do basically nothing, and yet it spread to everyone who touched it like wildfire. Google was just another search engine in a world that had dozens, but it was so much better.

Many pitches look like “business plans”—an overview of every aspect of the company. Business plans, as an institution, can’t decide to optimize for breadth or depth and so optimize for shallow verbosity. You have very, very limited bandwidth. Focus it on the things you are great at.

Understand that, while the eventual goal of the enterprise is to make a lot of money, investors understand that they’re looking at acorns, not trees. You don’t have to show profitability now, and focusing too much on your timeline to getting to it suggests misunderstanding of what investors want. You are selling growth, growth, and growth. Less than half of companies accepted to YC in a recent batch had revenue, to say nothing of profits, as of their application.

Being able to talk about the economics of your business is great. Definitely do, particularly if they are attractive. But the most attractive economics are the ones that suggest that you're able to grow. If you are capable of calculating your unit economics in, for example, a food ordering business, knowing that you are profitable on a per-order basis is a very positive thing; show that math. If you can calculate customer acquisition costs given a channel and demonstrate that it could be scaled upward, that’s wonderful.

If your revenue for this month exceeded expenditures by $2,000, on the other hand, that’s not all that interesting. Similarly, plans on how you’re going to shave a few hundred dollars off your budget by refactoring your app or changing hosting are not interesting. You can’t cut costs into becoming Google.

Highlight evidence of success

There are three hurdles regarding your metrics: Do you know what you should be tracking? Are you actually tracking it? Are the numbers you have exceptionally good relative to the amount of time you have been working?

You should know the common metrics that startups in your industry or that use your business model are judged by. Andreessen Horowitz has written extensively about startup metrics. You can also ask more experienced friends or colleagues in similar businesses. Don’t reinvent the wheel; your industry already has prior art that everyone knows how to quickly evaluate.

If you have shipped a product, you should have in-product analytics installed so that you thoroughly understand the use of your product. In addition to increasing the rate at which you can improve your product, this will let you demonstrate confident mastery of user behavior. This makes you more credible to investors.

Consider this insight: “Sixty percent of our users immediately invite their teammates, because collaboration on a team is the key value proposition for this; that’s why users migrate to us from Excel.”

Cite metrics correctly, especially revenue. Do not cite gross merchandise volume (GMV) as revenue; if you facilitate a transaction between two parties and collect a fee then the total transaction is GMV but only your cut is revenue. Do not cite one-off payments for a long service as revenue in a particular month; an annual contract for software, for example, might be invoiced in March and collected in April, but the revenue is recognized evenly over the period of service, not all in March or April. Fudging revenue suggests malevolent intent or incompetence.

You might be so early in your company that you don’t have good instrumentation built yet to report metrics. This is common, but you should at least demonstrate understanding of the metrics to convey that you will make data-driven decisions in the future.

If you have “traction,” highlight your traction. Get used to this word; it is a vague term of art, but you'll be asked about it incessantly for the next few years. The best definition of it is “quantitative evidence of product/market fit.” Definitions are diverse because businesses are diverse—both Apple and Berkshire Hathaway are worth billions of dollars, but one look at their websites will tell you they value very different things.

Startups often ask us what numerical targets reviewers are looking for. This is difficult to generalize about, because they’re not static targets. Reviewers want evidence of great progress given the stage you're at and hints that you will accelerate in the future.

You don’t need every number to be stunning, but there should be some evidence that you have something working and special. At Snap it was that users opened the app an average of seven times per day (when most apps struggle to pass the “toothbrush test”—twice a day).

Your reviewer will read literally thousands of data points today—the average pitch includes more than 10. No one can remember that many arbitrary numbers, so reviewers compress them to “zero,” “non-zero,” and “impressive.” Impressive is in the eye of the beholder, but since people find it useful to have numbers, here are some rough estimates for where very early-stage startups level up from “great; they shipped a product” to “excellent; something is probably working about their customer acquisition.”

For B2B SaaS sold on the low-touch model, like Basecamp: $10,000+ monthly recurring revenue; 100+ active, paying accounts

For B2B SaaS sold on a high-touch model, like Salesforce: At least one pilot for a software product (not a consulting engagement) worth $50,000+ or more actively underway; some evidence of a sales pipeline (one to two letters of intent signed, etc.)

For mobile apps: Hundreds of thousands of free users or thousands of paying users, ideally with rate of acquisition increasing over a period of at least several weeks

For ad-supported websites: Millions of visitors per month, ideally with sustainable growth in traffic over a period of at least several weeks

The longer you’ve been working, the higher the bar is. “We have 20k active users” sounds much more impressive for a company that launched last Tuesday than it does for one that has raised an angel round and been active for three years. This suggests a way to (truthfully) help reviewers see your project in the best possible light: If you have been interested in a project for a while but have gone full-time or pivoted or “gotten serious,” etc. more recently, date your startup to that more recent milestone rather than the earliest date you had thoughts in the shower about it.

These targets may strike you as high. They are high. At least some companies in the applicant pool have them. If you do not, you need extra effort on the qualitative parts of your application to demonstrate that you have the skills, drive, and opportunity to achieve these numbers quickly and then grow explosively from them.

Ship a compelling prototype

Include a prototype if possible. Prototypes demonstrate the ability to ship working software. Many fewer early-stage tech entrepreneurs than you would expect can successfully ship working software.

You get two minutes. You might get as many as five or ten if your written pitch suggests that your company is likely quite interesting. This is not enough time to evaluate an entire software product.

You need to make the reviewer’s first few minutes with your software absolutely sing.

Nobody has ever written in their comments on a company “Wow, their sign-in screen blew me away. I want to invest in that sign-in screen.” Give people a link that drops them into a fully authenticated session on a pre-populated account that already has dummy data in it.

This is not a training session. You don’t have to go through the standard first-run experience or explain to them how to use every feature of the software. You should, instead, immediately show the most impressive interaction or output of your application. Often it will involve showing the goal state of your application and allowing the user to work backward. For example, if you were pitching email marketing software, you’d start with a finished or almost finished email in your application and ask the user to try editing or sending it.

Fake things liberally. If there is an interaction that is asynchronous or requires the involvement of another party in real use of the software, fake an immediate response. (You can be totally transparent that you’re doing this.)

Bring a level of design and polish appropriate to what you are doing. Many accelerators specifically prioritize founders who have “product sense,” which is a nonspecific blend of artistic taste, keen appreciation for good user experience, and an understanding of what the best products in a market already do. You should make sure to demonstrate product sense, particularly if you are in a market where it is fundamental to the success of your company, such as B2C mobile apps, for instance.

Some companies can’t reasonably demonstrate a prototype—Stripe Atlas has seen pitches for everything from medical devices to sewage systems. If you have a physical product, consider video or photos of it. If you are too early stage to have functioning software, mock-ups will at least let you show that you have good design and product sense.

Have fun, but calibrate well

Many pitches have a sense of humor, optimism, and joy for life in them. This is a good thing to have if it is natural for you; don’t feel the need to make your writing more stilted simply because that would sound more professional. If you’re not naturally blessed with the gift of gab, you can work on your storytelling later; just write the facts like you’d describe them to a smart friend.

Can you have too much fun in a pitch? Well, while many investors look for a certain cheeky irreverence for meaningless rules, remember that startups are a workplace and your investors operate under substantial public scrutiny. Don’t tell stories that suggest that you have catastrophically poor judgment, particularly not in a fashion where you expect the reader to applaud you for poor judgment.

Risk-taking is encouraged in startups; stupid risks are not. Walking into the office of a person in a position of authority without a meeting scheduled is a risk, but it suggests ambition and sales ability. Describing crimes you’ve committed generally suggests poor judgment.

Get an introduction if it would be great

Silicon Valley has a subtle culture around warm introductions (“intros”). As Marc Andreessen explains:

If you have a high-quality intro available to you, ask for it. A high-quality intro needs both of two things: It needs to come from someone who is credible to the decision maker, and it needs to include specific evidence of you being above the market average in some respect.

This is a high-quality intro if it comes from, for instance, a person that an investor has previously funded:

Any intro from someone who is not known to be credible from the perspective of the decision maker is a low-quality intro, virtually regardless of what it says. Any intro that does not include specific evidence of above-market ability or a strong, costly personal recommendation is a low-quality intro, virtually regardless of who sent it.

A low-quality intro is, at best, zero signal; at worst, it is negative signal. Don’t spend time getting low-quality intros just to have an intro.

People who are inculcated into the Silicon Valley intro culture are generally very careful with intros, because they carry a very real social cost. An intro isn’t just an email; intros that lead to successful outcomes work to the advantage of the person giving it by tightening their ties to both parties through the intro. If one makes a habit of making other-than-successful intros, one loses the capability to make intros that matter.

You don’t need an intro to apply to YC. That fact is one reason why they’re successful; they have had their pick from the vast pool of talented people who do not have a high-quality intro to a Silicon Valley VC available to them. About half of a recent YC class had no preexisting social connection to YC. That said, if you have a high-quality intro available to you, take it; it is not the case that it has zero value.

Keep going regardless

Many companies that go on to substantial success are passed on by investors, often multiple times. This isn’t necessarily a reflection on the company or founders—YC explains that constraints in the number of startups they can help per batch mean that for startups that are near the bubble, the reason for nonacceptance is likely to be something in another group’s application that made them just marginally better to include in the class.

If your pitch doesn’t succeed this time, that’s totally fine. Continue executing on the plan for your business. You will have other chances to pitch investors, just like there are always more prospective customers or employees available. You might even close funding from the same investor. Half of YC companies in a recent class had at least one founder who had applied unsuccessfully to YC before.

Don’t give up. This will not be the hardest moment in your company’s life; dust yourself off and try again. Improve yourself, your company, and your pitch for the next opportunity.

If you’re accepted to an accelerator or you raise investment: Congratulations! But you haven’t won yet. Take an evening to celebrate, then get back to the work. The prize for success is that it unlocks harder challenges with more at stake for next time.

Regardless of whether you’re just getting started or ready to move onto harder challenges, we’re here for you. Stripe Atlas is happy to help our companies with individualized pitching advice. If you are a Stripe Atlas member, we’ll be happy to walk through your YC application with you.