Enterprise Sales for Founders: A GTM Playbook from TinyFish and Airwallex

TinyFish CEO & cofounder Sudheesh Nair and Airwallex VP of Enterprise for the Americas Philipp Reichardt on why the enterprise and mid-market split stopped working, how to get your first ten customers, and when to kill a deal. From our GTM Academy fireside.

Last week we filled a room in San Francisco with founders, almost all of them selling something right now. Sudheesh Nair, TinyFish's cofounder, has built two enterprise companies. Philipp Reichardt is VP of Enterprise for the Americas at Airwallex, and joined five and a half years ago to launch the business in the US. They took turns on the same questions, and about half the evening ran on the room's questions rather than the agenda.

Here is what they said, in the order you actually hit it.

Decide whether you want this at all.

Sudheesh opened by trying to talk the room out of it.

"If you like your life, don't."

His description of the process: a procurement team that does not want to do a deal, a technical buyer who is walled off from procurement, then security, then legal, each one worse than the last. Three months for an NDA before the real evaluation starts. And when it does start, they bring in three more vendors, because buying from one is not how enterprises buy.

The test is whether you can fund it. A long close means you can compete for six months and still lose. "Second place in sales is actually the first loser," he said. It is better to lose in a day than to lose in six months, and a business that cannot survive the second outcome is not ready for the first.

Enterprise and mid-market is the wrong split.

This was the reframe the rest of the night hung on. Sudheesh has stopped sorting customers by company size and started sorting them by how many people he has to touch to close.

High touch means many people, many times. Low touch means few. The reason the old labels broke is that AI has moved some very large companies into low touch. One of the largest grocers in the country arrived through self-serve exploration, no salesperson involved. Meanwhile a small testing startup needed a custom build and turned out to be the highest touch account on the books.

Why it matters beyond tidiness: low touch is easy come, easy go. Someone who switched to you on price switches away on price. High touch is where durable revenue lives, and durable revenue is what net retention is made of.

"There is no net retention if you are selling to a group that's churning 30 or 40% every quarter."

The math underneath is unforgiving. An enterprise seller costs 250 to 350 thousand, carried on a draw for two or three quarters before they produce. If cycles run five to seven months they can work 11 to 15 deals a year. If the deal size cannot carry that, the cost of sale puts you underwater no matter how good the product is.

You need a right to win, and proof of it.

Philipp's version of the readiness question is narrower. Do you have a unique insight, knowledge or expertise that gives you the right to win this specific customer? Can you say it in one sentence? And do you have proof?

"Just having an idea is great, but you need to actually have something to back you up."

Before Airwallex was an enterprise company it spent its time in conversations rather than pitches. What are you struggling with. How does this look from where you sit. Help us understand. Most people will not give you the time. Some will, either because you hit a nerve or because they like talking. Either way you learn, and you come back with the problem in their words rather than yours.

Get yourself off the centre of your own product.

Sudheesh's argument for enterprise is not the revenue. It is what it does to your judgement.

Founders build from the inside out. You talk to a few people, go to a conference, change your mind, and soon the tail is wagging the dog. Models ship capabilities that make your roadmap obsolete before it lands. You open LinkedIn and conclude that you are bad at this.

Selling to a large company forces you outward, because what they buy on is business value, which means understanding their customer's problem and not just theirs. His example is German cars. The engine is the best engine and the people who built it have no interest in cup holders, but a buyer will walk over a cup holder that does not fit. The differentiator you are proudest of shows up on day 75, on the freeway, when they finally press the accelerator.

It applies to unglamorous things too. Observability matters because someone inside that company stuck their neck out for you, and if your product fails silently they hear about it from their own customer.

"That is the worst feeling to have. And that is a feeling you have to feel."

The first ten come from your network. After that, go to your competitor's conference.

On the first ten, Sudheesh agrees with what our CRO Daisy Hoang said at the last GTM Academy: they come from people you already know. What he adds is that it does not matter much whether they stick. What you need from the first ten is a story, so that the eleventh is not being asked to jump into the water first.

The next fifty are a different exercise, and this was the most-repeated tactic of the night. Go to your biggest competitor's conference. If they are careless enough to sell you a booth, buy one.

The logic is that nobody flies themselves to these events. Your competitor has paid for the flights and the hotel rooms of a room full of qualified buyers. Partner conferences work the same way and are easier to get into. Sudheesh worked a booth at a partner event the week before, and a telecommunications company that TinyFish would never have reached came through it.

Better still, run the deal on the partner's paper. No NDA of your own. Their rep gets paid, which means they take you to the next account. And procurement is negotiating with a vendor they already have.

Then do the follow-up properly. A conference gives you hundreds of cards. "Don't send a BDR for the cold calling." Write real emails, make the touch specific, and make it worth their time.

Pick 25 accounts and expect to be ignored seven times.

Philipp's operating advice is to stop trying to talk to everyone. Pick 25 accounts. Research every single one before you write anything: who to talk to, what they do, what they might be struggling with, which conferences they attend, what they post.

"You can stand out by being human, because they're all getting blasted with AI bullshit."

Airwallex's own data says it takes seven contacts before someone replies. So six silences are the cost, not a verdict. Disqualify fast when you get a real no, take them off the list, add another, and keep the list at 25. Two or three outreaches a day gets you through it in a month.

Sudheesh's addition: right message, right person, right time, and only two of those are yours to control. Timing is a variable, so the job is to keep showing up without becoming the reason they say no.

Test your champion before you need one.

A founder in the room had lost three deals in pilots, one because the champion left. Philipp's answer was that a departing champion should not lose you a deal if the deal was run properly.

The test is simple and you run it early. Ask them to introduce you to their peer. Ask who their boss is. Ask to present together. A real champion does it. A friendly contact does not. And the moment they introduce you to someone else, you have a second relationship and the deal stops depending on one person's job security.

Give a range, anchor high, then find them the money.

When a prospect asks the price in the first demo, founders freeze, and the silence reads as evasion. Sudheesh's answer is to give a real range immediately, wide enough to be honest, then turn it into discovery. What are the requirements. What does it depend on. How much value does solving this create.

Anchor high. But the part founders skip is the second half: show them how they are going to pay for it. Enterprise budgets shift, they do not grow. So the useful question is what they stop buying when they start buying you.

"It's your job not just to sell the product, but to find the money for them."

Underpricing is its own trap. At a company before TinyFish, Sudheesh priced a deal aggressively low because the logo mattered more than the margin. The customer came back and said there was no way anyone could deliver that scope at that price, and suggested he go sell to someone smaller.

Philipp's version: don't come in too high, don't come in too low, and expect to get it wrong sometimes. What helps is not being single threaded, so you have enough people telling you enough things to know how price sensitive they really are.

Know which box your value sits in.

Asked how you actually calculate economic value, Philipp started with honesty: it is hard, and there is an industry of consultants who do only this. What makes it tractable is building the case with the customer rather than for them, in their language and their metrics. And ask questions that cannot be answered yes or no. Not "does this save you money" but "how much, and how do you measure it".

Sudheesh's framework is a two by two. Hard dollars or soft dollars, numerator or denominator. Hard dollars are on the books. Soft dollars are usually people. Numerator means you help them make money, denominator means you help them save it.

Numerator cases are the easy ones, because there is new money and you can take a share of it with no risk to them. Most deals are denominator cases, which means share shifting, which means you need to name the line item you replace.

Knowing which box you are in matters because you are not in the room when it gets decided. Your champion is, in front of a CFO, and if you have not armed them with the case they fold at the first question. Philipp's practical version: ask your champion what numbers their boss cares about, build the deck for them, and then ask to be in the meeting. You built it. They owe you that.

A maybe is worse than a no.

The longest answer of the night came from a founder who had been chasing one account for over a year.

"If you don't kill the deal, you're killing your company."

Large companies run on three to seven year plans, and the person you are talking to is running on a nine to five. Your timeline is the one that breaks.

But his own practice is not to walk quietly. At a dinner with a large Japanese distributor that had been stalling for months, he said it straight: this is the right fit, you have the resources to decide and we don't, every few weeks you ask for another thing we cannot afford to do. Give me a next step or a straight no. Both are fine.

What came back was an introduction. The company admitted it moves slowly, then named one executive who does not.

That introduction may still go cold. The point is that the outcome stopped being theirs to sit on. "Enjoy the no. It's not personal." A maybe is the one answer you cannot work with, so force it either way, and sometimes a forced no turns into a yes.

Hire when the funnel is already there.

Founders should personally sell the first ten to fifteen deals. Not because it is cheaper, but because the first product is wrong and the second one usually is too, and the only way to get the third one right is to have been in the room.

After that, hire operators rather than armchair quarterbacks, people who are still player and coach. At seven reps you hire a manager.

The constraint that matters: every time you hire a rep, building the top of the funnel is your job, not theirs. Anything they add is a bonus. If you cannot feed them, you are too early to hire them, and the problem to go fix is demand, not headcount.

Conviction has an angel and a devil.

Sudheesh closed on the thing underneath all of it. Conviction comes with two failure modes. Ego tells you that you are right, and drives you off a cliff. Insecurity tells you that you are not good enough, and you change direction every week.

"All of us have imposter syndrome. I'm sitting here thinking, why am I telling you all this stuff."

His case for doing it anyway is that the downside is not very bad. If you are anywhere in the world you want to be in San Francisco, and if you are in San Francisco you want to be at a startup, and if you are at a startup you want to be at an AI-native one. The worst case is unusually good. So swing, and check whether the conviction is coming from the work or from the ego.

This conversation was part of TinyFish's GTM Academy, a series for founders building AI-native companies. If you want in on the next one, join the community on Discord.