If you build it, they will do absolutely nothing.

Entrepreneurs have to actively recruit their first few customers. This is hard, important work, because the first few customers will help you refine your product, begin understanding how to position it in the market, and provide important social proof for your company. This will generally be via active sales rather than the customers passively stumbling upon your offerings. Early stage companies do not yet have the marketing machinery that exposes them to steady streams of qualified leads acquired in a repeatable fashion. You will instead convince people to use your product one at a time.

The good news: You can do this, and do it well, even if you’ve never done sales before and even if you feel uncomfortable with it. You have advantages most sales reps lack: the ring of authenticity, commanding understanding of the problem domain, and the ability to make changes to the product to close early deals almost in real time.

It won’t be trivial, but it is easier to get started than most entrepreneurs think. You don’t need to have a perfectly polished product or superior sales skills to count to 10.

Why 10? You likely won’t be able to sustain your business with only 10 customers, but by the time you have 10 happy customers, it isn’t a fluke anymore. You’ll start to hear recurring themes about their needs. You’ll have nascent but real goodwill to turn into references and referrals. You’ll be able to say to potential employees and investors, “This isn’t a pipe dream anymore; every software company with 10 paying customers gets to a hundred, and every company with a hundred gets to a thousand. We just have to execute on this. Do you want to come along for the ride?”

We’re going to break down the basics of hardscrabble sales in a business-to-business context: finding companies that can use your software, quickly narrowing those down to the ones worth your time, and making the crucial first conversations happen.

Somewhat surprisingly, early sales conversations resemble pitching—you’re attempting to quickly intrigue someone with the seed of your idea just enough to earn the privilege of a longer conversation.

How do you find prospects for your company?

You’re going to do things scrappily in a spreadsheet to start out, likely with only a handful of columns: name of company, name of person, email address. That spreadsheet starts blank and terrifying. You need to carve out dedicated time to get names on it. You don’t need many to get started—a few dozen will keep you quite busy until you get good at the mechanics of pivoting from a company to a decision maker’s contact information and getting in contact with them.

Most entrepreneurs are not coming to a product area or industry entirely from a blank slate. You likely already have colleagues, beta users, or similar folks in your network. They should likely be the first names on your list—if not to sell to then certainly to ask for introductions to other folks in similar circumstances. You will close more sales from warm introductions than from an equal number of cold pitches, so start by picking this low-hanging fruit.

While you’re doing that, you’re also going to attempt to find new leads: companies and contacts who could potentially use your offering. At a base level, this means writing down a list of industries and job roles you expect to make use of your offering, then Googling until you can’t stand it anymore. For example, if you’re selling to office managers of dental practices, locking yourself in a room for an hour will give you the names and contact information of several dozen dental practices in Topeka, Kansas. Repeat for as many cities as needed to develop a few dozen names.

You can do this efficiently by finding places where customers congregate online, such as industry associations, conferences, communities and forums, and specialized directories. These are worth investigating, both from the perspective of developing leads and from the perspective of learning about organic behavior among your target customers. What issues routinely come up when they talk shop with their peers? What pain points do they have, and how do they conceptualize them? What else do they use? Knowing the answers to these questions helps your development, marketing, and sales efforts.

A note of caution: You can also purchase lists of leads from various providers, but for a variety of reasons, they’re harder to make use of effectively than leads you develop yourself. You’re also likely to feel worse about the experience. Defer using them until you’ve done things the hard way; the lessons you learn about the mechanics of developing and qualifying leads will inform your sales processes for years to come.

How do you determine if prospects are good or not?

As you get more experience running your business, you’ll get better at qualifying customers—determining which ones are likely to be easiest to convert, most profitable to service, and least hassle to support. In the early days, though, you don’t have much data to go on and you may not have an intuitive sense for your market and ideal customer profile yet.

Good customers for early startups share a few attributes, though, so you can prioritize these:

Early adopters: Whatever you’re selling, whether that is consulting services, software, or a physical product, you want to sell it to someone who has bought similar things before. They know the value proposition, and they know how to work their internal decision-making processes to successfully purchase the thing. You do not want to be someone’s first rodeo—they may give you very unrealistic signals with regards to how close their organization is to making a purchase, they may not be able to properly value your offering, and they very likely will be a difficult customer to onboard and service.

The term of art in the industry is “early adopter,” someone who is comfortable using products before their friends or colleagues are ready. They were likely on the internet first among their peers. They switch phones the day the new model comes out. They positively enjoy trying new software and are always on the lookout for new useful things.

How do you identify early adopters? One way is to look for early or reference customers of firms that aren’t competitors but that are lateral to you. For example, if you sell invoicing software for web development firms, look for firms that use project management tools created by similarly positioned startups. (You can often find customers mentioned on a company’s blog or homepage.) This lets you dodge a major objection to going with a new company, which is the risk factor: Clearly this customer has sufficient tolerance to adopt a solution from an “unknown” provider; they’ve done it before.

You can also look at hangouts for early adopters, like Product Hunt, and scope down to members who are likely to be in your industry. You do not need many names to count to 10.

On the internet: It is often easiest for internet entrepreneurs to sell to people who are aligned with the internet.

What does alignment look like? They spend time on the internet voluntarily, they’re active in online communities, and they mingle their online identities and professional worlds. This is convenient for you from an identification perspective (since their online activity gives publicly visible evidence of their existence and contact information) and from a sales perspective—they’re likely not to need reference customers, high-quality printed collateral, or other examples of Serious Sales Efforts™ because they’re comfortable transacting online.

When you’re organizing your spreadsheet, prioritize upwardly mobile firms and people who have developed web presences. Firms still exist that have almost no footprint on the internet. Many of them make excellent customers for many businesses, but they are unlikely to be excellent customers for your business today.

Light sales cycles: Even within a particular market, some firms have more process for particular purchases and some have less. You want to rigorously qualify prospects to find ones where a single decision maker can green-light adoption of your offering out of their discretionary budget. This will disqualify many potential users, and perhaps even the majority of particular markets (those for enterprise software, for example), but often there will still be enough users left to count to 10.

How do you identify users who have decision-making power and budget readily available? Prior to having conversations, you’ll explore and experiment. Small business owners typically have total discretionary authority and are still intimately involved in many significant purchases; this makes them a good point of entry for products costing $50 to $50,000. Even if they’re not going to be the end user at their company, they can make the decision to adopt you and forward you to the person who would use your thing directly.

After you’re in conversations with a prospect, ask point-blank: Can they independently decide to buy your thing? This is not a question many entrepreneurs are comfortable asking, but it is one you should get used to. You are not the first person to ask it of your counterparty, and you will not be the last. They are highly unlikely to be offended by it. If they demur, round that to “not interested right now” and move to the next prospect; you can always circle back after you’ve counted to 10.

Intrinsically reachable: Some businesses and some decision makers are intrinsically easier to reach than others.

Which businesses are intrinsically reachable? Typically, the ones where that is a prerequisite for operating their own businesses. Lawyers, accountants, consultants, and other service providers are always willing to pick up the phone and talk to someone on spec because that is the primary way they find new customers, and they are in constant search of new customers. This is very different behavior than that exhibited by folks who work in back offices or, for example, individual contributor engineers, who would generally prefer not to take a call because it distracts from “real work.” Engineers are intrinsically reachable (by dint of being tied to a computer all day, every day) relative to field service professionals.

If you sell software across a number of verticals, prioritize firms and decision makers who are intrinsically reachable; this maximizes the effectiveness of your early sales efforts.

How should you communicate with customers?

Use what is natural for you and for customers, but if you have any doubts, overwhelmingly the most common option is a cold email to get a phone call, with the main sales effort either happening on the call or after it.

Keep your cold emails concise and action-oriented, since you want to balance imposition on your prospects with the benefit of speaking with you. Generally, your first outreach will be a single paragraph, three to four sentences long, with a single ask.

Prove you put in some work

The first filter to getting good results with a cold email is sounding like a human, not a spambot. So sound like a human. Since you’re in the Do Things That Don’t Scale phase of your company, and likely not sending a huge number of emails per day, you can easily spend a few minutes to research your prospect and open with a bit of genuine connection.

The best cold email I ever received was selling recruiting services. There are millions of recruiters in the software industry; everyone with hiring authority has been pestered by them before. The recruiter with the best pitch identified a problem I clearly had based on evidence in a recent blog post of mine, proposed a sensible solution, and offered a call if I wanted to discuss matters in more detail. He also said very clearly that he was going to pitch me on the call.

You can generally do this in about a sentence after your salutation. (This should go without saying, but 50% of pitches in my inboxes need it said: get your salutation right. You should know your prospect’s name, since you did your homework, and you should use their preferred form of it.)

This sails over the prove-you’re-a-human hurdle, because it shows that you were intelligent enough to locate a presentation Karen has done and understand at least one point about it. (If you can comment intelligently on their work, so much the better, but even being able to paraphrase back to a customer what they consider to be unique about their company makes you better than the vast majority of pitches they get.)

Start providing value

After you’ve gotten someone’s attention, you want to convince them that a conversation with you is likely to be a valuable use of their time. This is often a combination of tell-and-show: tell them that you know what you’re doing, show a brief example. You have a cheat code available to you as a founder, which is that you literally started a company in your area of expertise, and regardless of how underprepared you feel, that reads to most customers as “likely ambitious and well-versed in that problem domain.”

After identifying yourself, if at all possible, give the prospect one or two useful takeaways for reading your email.

Ask for a specific, small thing

Close your email with a specific “ask”: You want a short phone conversation. You can explicitly label it as a sales conversation if you want to, although this is generally redundant when speaking to business owners or savvy professionals. Your prospect has been sold to before; they know you didn’t send them an email out of the blue just because you were feeling lonely. If they agree to take the call, it will be because they’re not opposed to hearing what you have to offer.

You should make it very easy for them to say yes to your ask. Take all the cognitive load out of it; rather than burdening them with scheduling, for example, suggest an appropriate amount of time and an appropriate day.

Inexperienced entrepreneurs often try to anticipate the objections (“If you’re busy Thursday then... ”) in the email, but this is not maximally respectful of your prospect’s reading time. If they’re busy Thursday, let them tell you that; they’ll likely propose a time if they’re very interested. If they come back with simple “I’m busy Thursday,” propose a different time, or offer to accommodate their schedule.

How should you go about following up?

One of the hardest and most natural parts about sales is to follow up. In most interactions in life, we’re trained to read social signals adroitly and interpret nonresponse as disinterest.

This is not an effective habit for selling. You should, instead, treat nonresponse as nonresponse. Your prospects are busy people with many things on their plate. If they were simply too busy to send you a response last week, dusting off their inbox with a one-sentence follow-up email might catch them at a better time or cause them to prioritize you more highly.

Some sales reps don’t take no for an answer. Particularly for your earliest customers, you can take anything less than “Wow, that’s very interesting” as an answer... but don’t take no response for an answer. Keep pinging, once or twice a week, until your customer engages, then play the conversation naturally from there.

There are tools that automate this process, but they create time efficiency for sales reps while creating time inefficiency for customers who use nonresponse as a way to discretely communicate disinterest. Since you can afford to be very liberal with your use of time as an early stage founder, just write your follow-up emails by hand and send them manually. You can do 20 in 10 minutes if you’re organized.

What should you ask for on the call?

Your calls will generally be very different depending on whether you’re selling a low-cost product or a high-cost product. Different industries and geographies have different understandings for what “expensive” actually means, but broadly speaking, from a startup perspective, software that costs less than $500 a month supports low-touch sales, and software that costs more than $5,000 or so annually requires high-touch sales.

Low-touch sales can eventually be completed passively without human interaction, but you’re using yourself as an accelerant for learning for the time being.

For low-touch deals, where a single decision maker can make a snap decision on whether they’re in or out, you should ask for the sale by the end of the call. Ideally, you’re ready to move immediately into taking their order, whether that is getting their payment and shipping information (for a physical good) or getting them an account provisioned for your software with the payment information to come later.

Many startups offering low-cost software offer a free trial. For your first few customers, free trials are likely to give people an easy way out of actually successfully adopting your software: They simply won’t get around to adopting it and cancel before the end of the trial. Instead of offering free trials, strongly consider offering ludicrously good help with onboarding or integrating the customer. If you have a data import step that has to happen, for example, offer to do all the button pushing yourself (including data cleanup) if they just email you the files.

You can offer a 30-day moneyback guarantee to decrease resistance to adopting you, since this has a very different dynamic than a free trial does. A customer who has committed to buying your software is committing to using it, and will only exercise the guarantee if the software fails due to defects. A customer who has not committed to buying your software can simply defer their purchasing decision for the time being, fail to develop sufficient conviction to purchase, and then default to not purchasing. Since you’re optimizing primarily for learning from committed customers for the time being, try to secure commitments.

Some products are intrinsically complicated and are not amenable to simple order taking; the customer will need to actually see evidence that the product will fit their purpose prior to getting the ball rolling on purchasing the product. In that case, you should spend your time on the call establishing rapport with the customer, learning about their problems and motivations, qualifying them, and if appropriate, asking for an opportunity to do a longer ~45-minute demo call or present a “formal proposal.” If they’re amenable to that, then you’re off to the races. We’ll cover the mechanics of more formal sales processes at a later date.

How can you get better at this over time?

The scrappy sales approach you start with is effectively a prototype for a future sales process that you’ll be capable of performing repeatedly, at scale. At the moment, you can change your process as quickly as you can change your mind, so experiment boldly with what customers you go after, how you pitch them, and what the structure of your offering is.

As time goes on, you’re going to start codifying this process so that you can execute on it repeatedly, as if from a script. Then you’re going to likely teach that script to other people. Then you’re going to build systems and processes that enable a team of people to execute that script more efficiently, at a higher quality level.

In addition to building the sales process itself, as you get more mature as a company, you’re going to produce things adjacent to the product that make the sales process better. For example, as you start to develop a marketing footprint (e.g., guides on how your customers could improve something in their businesses), you can use those as conversation-starters for cold outreach. You can build software that allows you to productize creating unique value for your prospects, such as a website speed analyzer if you’re selling web hosting, so that you can deliver outstanding value to people from your first conversation with them.

You can also feed what you learn from sales directly into the product itself: using the words customers do to describe things in the interface and documentation, for example.

This synergistic interplay of product development, marketing, and sales is one of the joys of running a business. We’re always interested in hearing what you’re doing—drop us a line at atlas@stripe.com.