The best outbound campaign Brex ever ran (75% demo rate, 75% demo to close)

Brex launched in June 2018 with a simple pitch: the first corporate card for startups. Soon after, with about 30 employees and close to zero revenue, it ran what Sam Blond, its chief revenue officer at the time, calls "the most successful outbound campaign I've ever seen." Blond shared the details on LinkedIn in January 2023, and every campaign number in this case study comes from his account.

The list came from a single Pitchbook report: seed through Series B startups in the Bay Area that had raised a round in the previous six months. That gave Brex about 300 companies. It bought 300 bottles of Veuve Clicquot at roughly $50 each and had TaskRabbit deliver them to the founders, each with a handwritten note from the CEO.

The note said something close to this: "Congrats on your recent fundraise! We know how hard it is to build a startup and we're rooting for you." That was all. No product, no offer. Shortly after the bottles arrived, the CEO followed up with an email asking whether the founder would be open to a demo.

Seventy-five percent said yes, or had the right person at their company take the demo. Seventy-five percent of those demos became customers. That is 225 demos and 169 new customers from a list of 300. Total spend was about $19,000: $15,000 for the Champagne, $2,000 for the handwritten notes, and $2,000 for delivery.

Put the funnel on one line: 300 founders, 225 demos, 169 customers. More than half of the companies on the list became customers. Divide the $19,000 by each stage and the economics look unreal for outbound: about $63 per company reached, $84 per demo, and $112 per customer.

For contrast, imagine a cold email sequence sent to the same 300 founders. If it booked meetings with 3% of them, most outbound teams would call that a solid result, and it would produce nine demos. At an optimistic 10%, it would produce 30. The Champagne campaign produced 225. The gap does not come from better copywriting. It comes from choosing every part of the campaign, from the list to the sender, to make saying yes easy.

The cost per customer matters more than the headline rates. A corporate card earns a small cut of every purchase, so a startup that grows spends more on the card every year. Against that, $112 to win a customer is close to free. Your product's economics decide what a campaign like this can cost, and the plan below starts with that math, before you buy a single gift.

Most outbound fails on timing before it fails on copy. Brex did not target founders who might need a card someday. It targeted companies that had just raised, which is when a startup opens accounts, hires, buys software, and starts spending in earnest. The need was not hypothetical. It was on that month's to-do list.

The trigger also lined up with what made the product different. At launch, Brex qualified startups on their funding rather than credit history, did not require a personal guarantee, and set the spending limit as a share of the cash in the bank. A company that had just closed a round had more cash than ever, so it qualified for a bigger limit than ever. A traditional bank, by contrast, asked the founder to personally guarantee the card. The campaign reached each founder at the moment Brex's advantage was largest.

The list was small, and every name on it was a fit. About 300 companies, in one region, at stages Brex could serve, each with a fresh reason to buy. Nobody on it was a long shot. That is what made the spend per account affordable: when every account deserves the attention, you can be generous with each one.

Champagne is what people open when they close a round. The gift was about the founder's moment, not about Brex, and that is the difference between a gift and swag. A mug with your logo on it is an ad people can drink from. A bottle that arrives the week they are celebrating is a gesture, and people remember who made it.

The note asked for nothing. It congratulated the founder and said the sender knew how hard building a startup is. That line was believable because it was true: Brex's founders had built their first company as teenagers and were in the middle of the same grind. A pitch in the note would have turned the gift into a bribe. Leaving it out let the gift do its one job, which was to earn the right to ask later.

It came from the CEO, to a founder. A note from a sales rep is a sales touch. A note from another founder, followed by an email from that same founder, reads as one company leader reaching out to another. Blond adds that the handwriting can be automated, and services that write with a real pen make that practical at a few hundred notes, so this step does not have to cost the CEO a week of evenings.

A physical object cuts through. A bottle on a desk gets noticed, talked about, and often opened with the team, which puts the note in front of more people than just the founder. When the follow-up email arrived, it came from a name the founder already connected with something generous. It was no longer a cold email.

The gift came first and the ask came second. People feel a pull to return a favor, and agreeing to a short demo is an easy way to return one. The order matters. "Take a demo and we will send you a bottle" attracts people who want the bottle. A bottle with no strings, followed later by an ask, attracts people who are curious about the company that sent it. The ask also had a way out: founders could hand the demo to the right person on their team, and those handoffs counted toward the 75%.

The campaign did not run in a vacuum. In a second post, Blond describes the launch around it. Brex announced the product and all of its funding at once, which drew wide press coverage. The same day, it covered San Francisco in billboards reading "The First Corporate Card For Startups." It sponsored the podcasts its buyers listened to for weeks afterward and made itself hard to miss at the conferences they attended.

His own example makes the point. A founder might see the funding news on social media, hear a Brex ad on a podcast on the walk to the office, pass several Brex billboards, and then find a bottle of Champagne from Brex waiting at work. For many founders, the bottle was the most personal of several touches in the same few weeks, not the first time they heard the name. A gift from a company nobody has heard of, sent into a quiet market, should expect lower numbers. The gift works best as the sharp end of a launch, not as the whole launch.

The close rate belongs to the product. A 75% demo-to-close rate means Brex solved an urgent problem that banks would not, and made it easy to say yes. TechCrunch reported at launch that a startup could get a virtual card in minutes, where a bank took weeks to approve one and offered a lower limit. A bottle can get you into the room. It cannot close a product the buyer does not need.

Treat the rates as a ceiling. They are one person's rounded recollection of a campaign run in a strong year for startup funding, in a city packed with the exact buyer. If your version gets a third of the demo rate and half the close rate, it may still be the best channel you run. Plan your budget on numbers like that, and let a better result be a pleasant surprise.

Start with the trigger, not the gift. A trigger is a public event that makes your product suddenly more useful to a specific company. For Brex, it was a fresh round. Good triggers share three traits. They are public, so you can find them. They are recent, so the need is live. And they are worth congratulating, so a gift feels natural rather than strange.

Common triggers, and the products they suit: a funding round, for finance, payroll, recruiting, and office tools. A new executive, such as a first VP of Sales or a new head of security, for the tools that person will be judged on. A burst of job posts, for the software those hires will need on day one. A product launch or a move into a new country, for infrastructure, localization, and support tools. A new office, for furniture, IT, and workplace services. Skip triggers that are bad news. Nobody wants a gift the week of a layoff or an outage.

The fastest way to find yours is to ask your last ten customers what changed in the month before they bought. The answers cluster quickly: "we had just raised," "we hired our first CFO," "we signed our first enterprise customer." The research agent can write that interview guide and pull the pattern out of the transcripts. If the trigger you find is not visible from the outside, keep looking. A trigger you cannot see is a trigger you cannot act on.

Pull every company that hit your trigger within a set window. Brex filtered a Pitchbook report by stage, region, and a six-month window. For funding, Crunchbase, Dealroom, and news coverage all work, and in the US, Form D filings with the SEC show many raises, sometimes before they are announced. For new executives and hiring, use LinkedIn and job boards. For launches, use Product Hunt, Hacker News, and press releases. Then filter hard for fit: stage, size, region, and whatever else predicts a closed deal for you.

Aim for a few hundred companies. Brex had about 300. That is big enough to matter and small enough that the most senior person at your company can send every follow-up personally. Check each row by hand before you spend anything. Is the company still operating? Is the founder still there? Is the news real and recent? Is there an office address? Every bad row is a wasted gift and, worse, an odd experience for someone who might have become a customer.

Choose the recipient on purpose. At a startup, that is the founder or CEO, because they make the decision and because a note from one founder to another lands. At a larger company, it is the leader who owns the problem and the budget. Send one gift per company. Five bottles to five people at the same company starts to look like a campaign, which is exactly what this should not look like.

Before you choose a gift, decide what you can spend. Multiply three numbers: the share of accounts you expect to take a demo, the share of demos you expect to close, and a new customer's gross profit in their first year. The result is what one account on your list is worth, on average, in year one. Your all-in cost per account, meaning the gift, the note, delivery, and the time involved, should sit well under it.

Brex's rates give 0.75 × 0.75 = 0.56 customers per account. At about $63 per account, the campaign paid for itself once a customer was worth more than about $112 in first-year gross profit, a low bar for a growing startup running its spending through a corporate card. Now try deliberately cautious numbers for a typical B2B product: a 20% demo rate and a 25% close rate mean 5% of accounts become customers. If a customer brings $8,000 of gross profit in year one, each account is worth $400, and a $150 gift package still pays back more than twice over.

Then budget for the cost that is not on any invoice: time. Brex's 225 demos, at 30 minutes each, add up to more than 110 hours of calls for a company of about 30 people, before any onboarding. If your version works, demo capacity will run out before gifts do. Send in batches of 25 to 50, so the follow-ups and demos stay prompt and so you can stop if the first batches tell you something is wrong.

Blond's advice is to come up with the equivalent that is on brand for your company. A useful test has three parts. The gift should be something the recipient might buy themselves to mark the moment. It should feel generous without feeling like a bribe. And it should connect, even loosely, to what you do. Champagne passed all three for Brex: it marked the round, it cost about $50, and a card built for startups had every reason to celebrate a startup's raise.

Some equivalents. A sales tool congratulating a company on its first VP of Sales could send a brass bell for the sales floor, to ring on the first deal the new team closes. A localization platform congratulating a company on its launch in Japan could send a box of Japanese snacks for the launch team. A developer tool could send dinner to the engineers the night a big release goes live. A workplace company could send a large plant to a newly opened office, where it will outlive every bottle. Each one is about the recipient's moment, and each one quietly shows what the sender understands.

Avoid the obvious traps. Branded swag is about you, not them. Gift cards feel like payment. Anything too expensive raises eyebrows and may break the recipient's gift policy. If you send alcohol, have a non-alcoholic option ready, because plenty of people do not drink, and check the rules for delivering alcohol where you are sending it. Many large companies cap what employees can accept, public sector employees often face strict limits, and some regulated industries have rules of their own, so check before anything ships.

The note should take five seconds to read. Handwrite it, name the event, wish them well, and stop. For example: "Congrats on closing your Series A, Maya. We know how hard it is to build a company, and we're rooting for you." Sign it with your first name and your company, and leave out everything else: no product, no offer, no QR code. The only thing the note should make the reader feel is noticed.

Real ink matters. People can tell a script font from a pen, and a fake handwritten note is worse than an honest typed one. If one person cannot write them all, use a service that writes with a real pen. Check every name and company name twice. Misspelling someone's name on a congratulations card undoes the whole gesture.

The follow-up comes from the same person who signed the note, in plain text, and four sentences are enough: "Hope the bottle made it to the office in one piece, and congrats again on the round. We built [product] for teams right where you are: [the problem the trigger creates, in one line]. Would you be open to a 20-minute demo next week? If someone else on your team owns this, I'm happy to show them instead." The last sentence matters. It gives a busy founder an easy yes, and at Brex, those handoffs counted toward the 75%.

Hand delivery is worth paying for when you can get it. Brex spent about $2,000 on TaskRabbit deliveries across the Bay Area, so the bottles arrived intact, on time, and in person. Local couriers do the same job today. Deliver to an office, never to a home address bought from a data broker. For remote companies, use a gifting service that asks the recipient to confirm where the gift should go. You lose some of the surprise, but you stay on the right side of the line between thoughtful and invasive.

Make sure the email lands after the gift. Confirm delivery, then send the follow-up within a day or two, while the bottle is still in the fridge. If there is no reply, send one short nudge a few days later and one more the following week, then stop. Three emails after a gift is persistence. Six is a sequence, and it cancels out the gesture.

Track it like any other channel, with one row per company: delivered, replied, demo booked, demo held, closed, revenue, and total cost. After the first batch, compare cost per demo and cost per customer with your cold email numbers. Expect the gift campaign to win easily on rates and lose on cost per touch. Cost per customer is the number that settles it.

Asking first. "Take a meeting and we will send you a gift" buys meetings with people who want the gift. Brex's order, gift first and ask later, buys goodwill with people who might want the product. Because nothing depends on the meeting, it also sits more comfortably with most gift policies.

Sending without a trigger. A bottle that shows up for no reason is confusing, and confusion does not convert. The congratulations is what makes the gift make sense. Without a real event to celebrate, you are sending expensive swag.

Pitching in the note. The moment the card mentions your product, the gift becomes an ad, and the reader starts working out what it was really for. Save the pitch for the email, where it is expected.

Scaling past the point of care. At 300 accounts, a CEO can send every follow-up and read every reply. At 3,000, the notes go generic, the follow-ups go templated, and the campaign turns into a direct mail program with worse economics. If it works, run it again next quarter on the next set of companies that hit the trigger, rather than making one batch ten times bigger.

Following up slowly, or not at all. A gift without a follow-up is a kind gesture and nothing more. A follow-up two weeks later arrives after the bottle is gone and the moment has passed. Draft and approve the email before the first gift ships.

Blond closes his post with a question worth sitting with: "When was the last time you received a bottle of decent Champagne and a handwritten note out of the blue?" For most buyers, the honest answer is never. That is the opening. It is also why so few teams take it: the list, the gifts, the pens, the couriers, and the follow-ups are a lot of moving parts for a team that already has a number to hit.

So we run it for you. The outbound agent finds the companies that hit your trigger and researches each one. We choose an on-brand gift with you, write every note with a real pen, and sign it in your name. The gift is delivered in person wherever a local courier can reach the office, with tracked shipping everywhere else. The follow-up goes out once delivery is confirmed, and every reply and booking lands in your pipeline.

You approve the list, the gift, the note, and the email before anything ships. Each campaign is quoted up front, and gifts and delivery are billed at cost. Tell us your trigger, and we'll run this play for you.