Why we hired for customer success before we hired for sales.

We didn't hire a head of sales until month eighteen. We hired someone whose entire job was making our existing customers successful a good while before that. If you're used to how most startups sequence hiring, that probably sounds backwards. It was deliberate, and I'd make the same call again.

The hire we made first#

The conventional wisdom, especially in venture-backed startups, is that you hire for sales aggressively and early, because growth is the metric everyone's watching and sales is the most direct lever on it. We never had that particular pressure, for the obvious reason — no board, no investors asking about pipeline. But even without the pressure, I think we'd have made the same choice, because of what we'd already learned about our own product.

AIVA isn't a purchase you make once and forget. A customer who deploys it, sets it up badly, and never revisits it will have a mediocre experience and quietly churn, regardless of how good the underlying product is. A customer who deploys it, gets help tuning it in the first few weeks, and sees it actually work will renew, expand, and — critically, for a company with no marketing budget to speak of — tell other business owners about it. We didn't need more people finding that first group faster. We needed to stop losing the second one by accident.

Who we actually hired, and what she actually did#

The person was Meera Nair, and her title back then was closer to "whoever makes sure new customers don't quietly disappear" than anything on a conventional org chart. In practice that meant she was on the phone with new customers constantly — walking through their first week of calls, listening to what their AIVA was actually saying to real customers, and flagging when the tone or configuration wasn't matching what the business needed. A lot of that early work turned into the instinct that still shapes onboarding today: ask what a customer's caller is actually feeling before you touch the FAQ spreadsheet, because a technically correct call can still lose a business its customer if the register is wrong.

That work doesn't show up as a single deliverable. It shows up as accounts that were at real risk of quietly going inactive in month two, instead still active and expanding in month eight. Nobody hands you a report labeled "churn that didn't happen." You just notice, later, that the customer base from that period held together better than it had any statistical right to for a company nobody had heard of.

Breaking our own hiring rule for this one role#

We've written elsewhere about why we hire out of Rajkot rather than Bengaluru for almost every role on the team, and this is the one hire where we made an early, deliberate exception. Meera works remotely, based in Bengaluru rather than in the office with the rest of us, because the job genuinely needed it — closer to a cluster of our larger customers, and closer to the flights that role actually requires when an account needs an in-person visit. We didn't resist that exception on principle. The role's requirements pointed somewhere else, so we let the role win. It's a small detail, but it says something about the sequencing question this whole piece is about: we were willing to bend a hiring principle we otherwise hold to for the sake of getting customer success right, at a point when we weren't yet willing to hire a full sales function at all. The priority was never subtle, once you see where we made exceptions and where we didn't.

Why sales could wait and success couldn't#

Every early customer we had was disproportionately important, in a way that's easy to describe and hard to feel until you've lived it. We didn't have the volume to absorb a bad churn month quietly. A handful of unhappy customers wasn't a rounding error on a dashboard — it was a meaningful fraction of our entire revenue, and most of our credibility in a market where nobody had heard of us yet.

Sales, without a product experience solid enough to retain what it sells, just becomes an expensive way to churn customers faster. We'd have been paying to acquire people we weren't yet equipped to keep. Customer success first meant that by the time we did build a sales motion, we had a retention story worth selling — actual businesses who'd stuck around, actual results we didn't have to embellish. The logic isn't unique to us — research on retention economics has consistently found that keeping an existing customer is worth more than the acquisition metrics most early-stage companies optimize for, which matched what we were seeing in our own numbers even before we could put a name to it.

What it costs to get the order "wrong" this way#

I want to be honest that this wasn't free. Growth was visibly slower in year one than it could have been with an earlier, dedicated sales push. There were months I looked at a competitor's announcement and wondered if we were leaving obvious growth on the table by not chasing it harder. Some of that growth really was left on the table. I've made peace with not getting it back.

There's a subtler cost too: customer success work is unglamorous and mostly invisible from the outside. It doesn't produce an announcement. Nobody writes about the churn that didn't happen. It's one of the harder cases to make internally, when you're deciding where the next hire goes, because the return is a graph that stays flat instead of one that visibly moves. I've sat in enough of our own hiring conversations where the temptation to redirect that budget toward something with a more visible payoff was real, and we talked ourselves out of it more than once by looking at which accounts were actually renewing and asking honestly why.

The counterargument I take seriously#

The strongest version of the pushback isn't "you grew too slowly" — it's that this whole account might just be survivorship bias, a comfortable story bootstrapped founders tell after the fact because it happened to work out. I don't think that's entirely wrong to worry about, and I'd rather flag it myself than have someone else point it out. We don't have a clean counterfactual where we hired sales first instead and can compare the two outcomes. What we do have is a customer base, spread across fourteen countries now, where a large share of new accounts still arrive through referral rather than outreach — which is a fact I can at least point to, even if it doesn't fully settle the argument about what would have happened the other way.

Would we do it again#

Yes, without much hesitation. The customers we have today are disproportionately people who stayed because someone on our side made sure their first few weeks actually worked — not because a salesperson closed them well. That's a slower way to build a customer base. It's also, I think, the only way to build one that holds up once you stop pushing on it. It's the same instinct, in a different form, that shaped why we've stayed at eight people rather than scaling headcount ahead of proof — hire for what's actually fragile first, and let the more forgiving problems wait.

If you're a business owner reading this wondering whether that philosophy shows up in how we treat customers day to day rather than just in how we write about ourselves, the honest way to find out is to talk to us directly or try AIVA yourself and see who calls you back in week two.