How to get your first 10 customers for your sales-led B2B startup (in 2026)

Every few weeks a founder emails me to say they read my article on getting the first 1000 users, did everything in it, stood up a glossy landing page, hustled their way to a few thousand email signups, and then sat there wondering why none of it had turned into a single penny of revenue. Nine times out of ten the same thing is going on. They're building something sold to other businesses, with a salesperson in the loop - usually them - and a price tag big enough that nobody hands over a company credit card on a whim.

If that's you, the 1000-users piece is the wrong map. The tactics in it still mostly work. The trouble is that the whole shape of the game changes the moment one person has to persuade another person to spend their employer's money.

So this is the companion playbook, for founders selling business-to-business the hard way, through conversations and demos and a great deal of following up.

The first thing to get straight is the goal. In sales-led B2B you are not chasing a user count, and a vanity number of signups will actively mislead you. What you want is logos. The milestone that actually means something is roughly ten paying customers who resemble each other closely enough that you can start to see a repeatable pattern in how you won them. Ten is enough to prove the thing is real and to give you references to lean on. It's not so many that you can hide from the fact that you don't yet know what you're doing.

I've done founder-led B2B selling a few times now. In my previous startup, Firedrop, we pivoted from high-volume self-serve SaaS to large-scale enterprise integration projects with blue-chip companies, where sales cycles were up to a year. More recently I've been building a B2B software platform that I sell the old-fashioned way, by getting on calls, doing demos and chasing things down like the rent depends on it. So what follows is the playbook I actually use.

Usual caveat before we start. This is a set of activities that work for fairly boring, structural reasons, not a formula. Anyone selling you a guaranteed B2B growth formula is selling snake-oil, as ever. YMMV.

When to start selling

The instinct is to wait until the product is built and polished before you dare show it to a buyer. Resist it. In B2B you can, and absolutely should, start selling before you've written a line of code, because the people you sell to will tell you exactly what to build if you give them the chance to talk.

Validation here doesn't look like an email capture field. It looks like a conversation, and then a commitment. Go and speak to twenty or thirty people who match the buyer you have in mind. Don't pitch at them. Listen, and find out whether the problem you think you're solving is one they'd actually pay to make go away.

The real signal is when somebody tries to give you money or put something in writing. A letter of intent (LOI), a design partner agreement, a deposit, even a slightly impatient "just send me the contract". At Firedrop we signed our first enterprise purchase order based on a two-day prototype and a waterfall plan of how we would build out the full implementation. It was for a relatively small amount - £10k or so - but it started the snowball rolling. From there we built up until our POs were in the hundreds of thousands.

Start the conversations now. The product can catch up.

How long it takes, and what it actually costs you

The consumer founder lies awake worrying about traffic. The B2B founder lies awake worrying about the sales cycle, which is the one part of this that nobody has yet worked out how to shortcut. A deal into a large enterprise can take anywhere from six to eighteen months to crawl through evaluation, security review and procurement. Selling to small businesses is quicker, but it's still measured in weeks rather than the minutes a consumer takes to sign up for an app.

The currency you'll spend is your own time, not an advertising budget. Founder-led selling is slow, unglamorous, and full of meetings that go nowhere. Set aside nine to eighteen months before you have anything resembling a repeatable motion. As a rough rule, the lower your average contract value (ACV - the typical annual size of a deal), the faster and more volume-driven the game, and the higher your ACV, the slower and more relationship-heavy it becomes.

To be clear, I'm not talking about product-led growth (PLG) here - the bottoms-up, self-serve, swipe-your-card-and-you're-in kind of SaaS. That world lives much closer to the 1000-users playbook. This is for the deals that genuinely need a human conversation before anyone says yes.

1. The foundations

In the 1000-users article I told people not to bother registering a company yet. For sales-led B2B, do the opposite. Get your house in order properly, because business buyers will look under the bonnet before they sign.

• Register the company. A buyer's finance team will not raise a purchase order to your personal bank account, and no serious procurement department will deal with a sole trader. I'm not giving you tax or legal advice here, so talk to an accountant, but understand that a limited company, a proper invoice, and before long a basic data processing agreement (DPA) and some half-decent answers to a security questionnaire all stop being optional the instant you sell to anyone with an IT department. It's tedious, and deals die in procurement far more often than founders like to admit. In the modern age of AI and concerns around data, you will also want to build a proper compliance pack where you prove that you are not sharing data unnecessarily with the likes of OpenAI, if you are integrated with them.
• Get a website, but keep it simple. Your B2B site is a trust artefact, there to reassure a buyer who has already met you that you exist and look like a real company. It is not a conversion funnel and you shouldn't agonise over it as though it were one. It needs to say who it's for and what it does inside about five seconds, and then get out of the way. The key point to get right is the contact form, and I would also recommend a live chat if you have the ability to man it - this is how I got the first enquiry from our biggest client when I was running Firedrop. Just keep it open while you're working and jump on anything that looks promising.
• Build a founder presence on LinkedIn. This is the highest-leverage channel you have in B2B and it costs nothing but effort. Set up a company page, yes, but put the real work into your personal profile, because business buyers buy from people they've come to trust. Forget the consumer advice about grabbing forty social handles defensively. Pour that energy into the one platform where your buyers actually spend their working day.
• Stand up a light sales stack. You do not need Salesforce and a revenue operations team. You need somewhere to track conversations so they stop falling through the cracks - a free customer relationship management (CRM) tier such as HubSpot, or something leaner like Attio or Pipedrive - and a booking link like Cal.com or Calendly (or even NeetoCal if you're cash-strapped) so people can grab time with you easily. That genuinely is enough to begin with.

2. Know exactly who you're selling to

The consumer founder sketches a single customer avatar and gets on with it. You have a harder job, because in any business deal of meaningful size there is rarely one decision-maker sitting there waiting to be convinced. There's a committee, and they don't all want the same thing.

• Draw a painfully narrow ideal customer profile (ICP). Fight the urge to say "any company that could conceivably use this". Pick the smallest, most specific slice you can credibly serve - a particular company size, a sector, and crucially the specific situation that makes them feel the pain right now. A narrow ICP is faster to sell into and far easier to turn into a reference that opens the next door. You can always widen it once you're winning.
• Map the buying committee. There's usually a champion who loves what you do and sells you internally, an economic buyer who actually signs the cheque, the end users who'll live inside the product day to day, and then the blockers - procurement, security, legal, and the annoying middle management people who feel threatened by your solution. Each of them is afraid of something different, and your job is to work out what, then use that appropriately in your messaging approach. Then, find your champion who - and this is crucial - is actually able to get meetings with budget-holders, and work them.
• Put a number on the pain. Consumer pain is a feeling you can gesture at. Business pain has a figure attached - money lost, revenue left on the table, hours burned, risk being carried on someone's shoulders. If you can't roughly quantify the problem you solve, you'll keep losing to "we'll just carry on as we are", which is your real competitor far more often than any rival product. One thing I have personally noticed is that people are generally very poor at estimating their own time, so selling on time saving can often be a grind. But every budget holder loves a cost-saving opportunity, especially if it can go somewhat under the radar so as not to impact their budget for next year (yes - beware the cost-saving tool that ends up getting noticed by the higher ups and reducing your champion's budgets as a result!).

A useful trick, and one I lean on constantly, is to get an LLM to roleplay your buying committee and tear your pitch apart from each seat at the table. Gemini and Claude are both good at it. If you want the version with no bedside manner, Grok will happily tell you your value proposition is rubbish, but might also show you an unsolicited picture of an excessively endowed anime girl as well, if the rumours are true.

3. Your first design partners

This is the business-to-business equivalent of the original article's "first 100 users", except you're after a handful of the right companies rather than a crowd of curious strangers. These are companies that will take on your solution in the earliest days and help you both to refine it and sell it. They're called design partners (for the refinement help aspect), and getting three or four good ones is the most important thing you'll do in your first few months.

• Start with warm introductions. Your network is worth vastly more in B2B than it ever is in consumer. Go to the founders, ex-colleagues, investors and old clients who'll actually pick up the phone, tell them precisely what you're building and who you're after, and ask them to introduce you. One warm intro to the right person is worth a hundred cold emails fired into the dark.
• Offer them something genuinely worth their while. A design partner gives you their time, their honesty, and eventually their logo, which is a lot to ask of a busy company. In return, give them a heavy discount or the product for free, real influence over what you build next, and an almost embarrassing amount of attention. Be upfront that it's early and rough - the right partners find that exciting rather than off-putting.
• Keep the number small. Three to five is plenty. Any more and you simply can't give each of them the hand-holding that early software demands, and you'll end up doing all of them a disservice. Their job, beyond paying you, is to validate the thing, help you shape it, and become the references that win you customer number six.

4. Founder-led outbound

Let me start with the good news, because outbound has a deservedly grim reputation and I don't want to throw the whole thing out. Done with care, founder-led outbound is one of the most reliable ways ever invented to get in front of the exact companies you want to sell to. You go and knock on the right doors instead of sitting around waiting to be found.

Now the bad news, and this is the bit I rather underplayed in the consumer article, so let me make up for it here. Mass cold email is dead, and good riddance to it. AI-powered spam filters have made generic outbound blasts almost entirely pointless, and your lovingly mail-merged sequence to five thousand strangers will sail straight into a void nobody ever checks, and risk blacklisting you in the process. Not fun.

What works now is the opposite of volume.

• Go signal-based, and go tiny. Build a small list of perfect-fit accounts - think fifty, not five thousand - and watch for the signals that someone is in pain right now. A company hiring for a role your product would make redundant, a fresh funding round, a new VP cleaning house, somebody posting publicly about the precise problem you solve. Tools like Clay let you assemble and enrich a list like this without much fuss, and then you write to each company like a human who's actually spent five minutes doing their homework. Automate the research, not the human contact. That's a mantra for the modern era if ever there was one.
• Work email and LinkedIn together. A thoughtful LinkedIn message paired with a short, specific email, each referencing something real about that particular company, will outperform any blast you could ever send. LinkedIn Sales Navigator earns its fee here for the targeting alone, though it does have its limitations. The aim, though, is not a high number of sends, it's a small number of real conversations with the right people.

When your customer is B2B, this is about as effective as cold outreach gets, though it takes real effort to do well. That real effort and consistency is where most people fail - mainly because it's a grind and can be scary - but that also means getting this right is a genuine superpower.

5. Content and founder authority

Content marketing was the single most effective channel in the consumer playbook, and it remains so here, but the flavour changes when your reader is a business buyer.

• Write as the founder, on LinkedIn, regularly. Be useful, be specific, and have an actual opinion. Your buyer has a long cycle and won't purchase today, but if you're the person whose posts they've been reading for six months, you become the first call the day the pain finally gets bad enough to act on. Thought leadership, when you strip the grandeur off the phrase, just means writing the things your industry will later pretend they knew all along.
• Solve the buyer's problem out in the open. Not your product's features, which nobody cares about yet, but their actual headaches. Publish the framework, the teardown, the messy spreadsheet you built to make sense of something, your own real numbers. Don't be afraid to give the good stuff away. Just don't sell all the family silver, obviously.
• Feed it back into the sale. A strong post becomes a follow-up asset the moment a demo ends. "Good to chat earlier, this is the piece I mentioned" with a link keeps you useful and keeps you present. Your content engine and your sales process should be seamlessly handing work to each other.

If long-form is your thing, a founder-led newsletter on Substack gives you a direct line to your buyers' inboxes that no algorithm can throttle. It doubles neatly as the engine for everything above. Medium is also a good platform for this, especially as it's loved by the LLMs so will help you with your GEO (see next section).

6. Optimise for AI discovery (GEO)

Here's a scene that plays out constantly now. A buyer with a problem opens ChatGPT or Perplexity, types "best tool for [their exact problem]", and builds a mental shortlist out of whatever the model says back, long before they ever fill in a contact form. If your name isn't in that answer, you're not on the list, and you'll never even know you were skipped.

This is why generative engine optimisation (GEO) now sits alongside traditional search engine optimisation (SEO) for any B2B founder paying attention. The mechanics differ from old-school SEO in a pretty fundamental way.

• Be the primary source. AI models reward first-hand data, strong opinions and original research, the things that can't be found repeated in a hundred other places. So publish your own benchmarks, your own findings, the view only you hold. Generic listicles get ignored by the models exactly as they're starting to be ignored by people.
• Appear on the sites LLMs love to train on. It's well-known that sites like Reddit, Medium and Substack are goldmines for the AI training crawlers, and there's lots of circumstantial evidence to show that featuring on those sites, with quality content, makes an outsized difference to your LLM rankings. I do stress quality content though; don't be one of these people who creates bots to spam Reddit endlessly - you'll be downvoted into a reputation sinkhole that will be hard to crawl out of.
• Make your site machine-readable. Clear headings, short self-contained paragraphs, and an llms.txt file at the root of your domain to point the crawlers at your best material. It all helps.

B2B buyers research harder and longer than consumers ever do, and an ever-growing share of that research now begins inside a language model rather than on a results page.

7. Free tools and proof

The consumer playbook showed SEMrush giving away a free site audit and WPEngine offering to speed-test your WordPress site, both of which were subtly (or perhaps not so subtly) designed to tell you that your current setup wasn't good enough and they were the cure. The same logic works beautifully in B2B, because we all respond more sharply to being told something's wrong than to being sold something that's right.

• Build a free assessment or calculator. A return-on-investment calculator, a "benchmark yourself against your peers" tool, a free audit of whatever you help with. It delivers genuine value for nothing, captures a lead in the process, and plants a seed of the precise discomfort your product resolves. The honest ones perform best, because business buyers can smell a rigged result a mile off. There are loads of no-code sources on the web these days, or just vibe code one on Replit or Base44.
• Publish a benchmark report. Business buyers are oddly fond of a good benchmark - a "State of [your industry] 2026" with real numbers in it. If you're sitting on proprietary data, a report like this builds your authority, earns you the backlinks SEO still wants (despite what everyone says), and gives the AI models exactly the kind of primary source they love to cite. One decent report can keep working for you for years. Even better if it can be a quarterly or annual thing.

8. Measure the pipeline, not the traffic

The consumer founder lives in their analytics dashboard, watching where visitors come from. You need to live in your pipeline instead, watching where deals get stuck and why.

• Treat your funnel as a sales funnel from day one. Track the journey from first conversation to qualified opportunity to proposal to closed deal, and pay close attention to the stage where things consistently stall. So many founders lax in this area, and they lose out so much from it. You should understand your buyer journey, and adapt your funnel around it, so that you can qualify leads faster and faster and develop qualified opportunities with clarity. A qualified opportunity, by the way, simply means a real prospect with a real relevant need, a budget, and the authority to act - everything earlier than that is just a general lead. Be disciplined about this: the rigour will serve you well as time goes on.
• Do your lost-deal analysis. When someone says no, go back and ask them why, properly. It's an uncomfortable conversation which is why a lot of people shy away from it, but it's also the most useful product and sales feedback you'll ever get for free. Develop a thick skin for it.
• Talk to your customers, relentlessly. I said this in the 1000-users piece and it's even truer here. The handful of companies who trusted you early are your richest source of what to build, what to fix, and what to say to the next ten prospects. Email them, call them, buy them lunch. They'll tell you almost everything you need to know if you simply keep asking. I assume it goes without saying that you shouldn't drive them completely mad, of course, but feel free to test where that line actually is.

A few other tactics worth knowing

The activities above will, applied with patience and consistency (and consistency is the real keyword here), get a sales-led B2B startup to its first ten customers and well beyond. There are other routes that need more money, more specialist skill, or more time before they pay off, and most founders are better off coming to them once the basics are humming. Briefly, here are the ones worth having on your radar.

• Channel and partnerships. This means borrowing someone else's customer base, through integrations, referral partners, or resellers who already sell to your buyers. It's slow and fiddly to set up and occasionally maddening, but a good channel partner becomes one of the most durable sources of growth you'll ever build. Just beware that, even after the initial deal, partnerships need constant nurturing, so treat them like a customer.
• Account-based paid advertising (ABM). Rather than spraying budget across the internet, you aim LinkedIn ads and retargeting squarely at the small list of named accounts in your ICP. It works best layered on top of your outbound and content, warming up the exact companies your founders are already chasing.
• Events and speaking. In the consumer world I filed these under "later". In B2B they're frontline. The right industry event, a speaking slot in front of a room full of buyers, or simply showing up in the niche Linkedin community, Slack group or Discord where your customers grumble to each other about the problem you solve, all of it puts you where the money is.
• Public relations. As with consumer, useful for building credibility and badly misunderstood. Worth pursuing once you've got something genuinely newsworthy and a couple of customer stories to point at.

If you want to go deeper than this, read Founding Sales by Pete Kazanjy. It's the closest thing the sales-led B2B world has to a bible for founders doing this for the first time, and it's far more thorough on the mechanics of early selling than I can be here.

For everything else, the shape of it is simple enough even if the doing is hard. Get your foundations straight, pick a painfully narrow buyer, win a few design partners by hand, and then turn every customer into the proof that wins the next. It will take longer than you want and involve more rejection than you'd like, and it's also entirely learnable by anyone willing to get on the calls and keep at it. So go and have the first five conversations this week. The product, and the polish, and the pipeline can all catch up from there.

And if you want to talk through your own early B2B traction, hit me up. I'm always happy to compare notes with founders doing the hard, slow, oddly satisfying work of selling to other businesses.