AppsFlyer: The Customer Success Playbook Behind $500M ARR

Thirteen years ago, Ziv Peled had two job offers. One was better funded. One paid more. He took neither — he took the smaller one, at a company whose product he didn't fully understand even a month after joining.

That company was AppsFlyer. And the decision that looked irrational on paper turned out to be the opportunity of his life.

Today, Ziv is Chief Customer Officer, leading a 180-person CS organization across 20 offices. AppsFlyer has crossed $500M ARR. And the frameworks Ziv built for customer success along the way are among the most sophisticated in B2B SaaS.

The first signal that something was different came early. "The first milestone of realization was really how amazing the demand was," Ziv recalled. "I worked in many companies before and then I probably haven't seen inbound demand before." The demand was so strong that in the first two years, Oren — the CEO — was asking the team to decline roughly 60% of inbound leads. Not because the leads were bad. Because the company didn't have the bandwidth to serve them properly. No SDR function. No product team in year one. AppsFlyer chose quality over volume before most startups even consider that tradeoff exists.

In a recent episode of Unicorn Builders: CX, Ziv shared the frameworks and decisions that shaped one of B2B tech's most scaled CS operations — and what he'd tell every founder building the function today.

The Relationship Capital Framework

Most CS organizations measure success through renewals, NPS scores, and adoption metrics. Ziv thinks about something different: relationship capital.

The concept starts with scale. More than 50,000 users interact with AppsFlyer's product online each month — and over 100,000 when API users are included. These aren't just accounts. They're individuals building careers. "If I can help them better in their job and then grow in their career so they will work with me again and again," Ziv said. "And then the relationship capital goes to many different things."

The compounding effect is straightforward: users who grow in their careers move between companies. If a CSM has genuinely invested in their success, those users carry the relationship — and often the vendor — with them. Advocacy isn't a campaign. It's the natural output of a CS motion built around individual career growth, not just account health.

This framework scales because it doesn't depend on any single account outcome. It's a long-term bet on the density of relationships across an entire user base.

The Educator Positioning That Outlasted a Product Cycle

In 2020, Apple's iOS privacy changes disrupted the mobile advertising ecosystem that AppsFlyer operates within. For most vendors, it was a crisis. For AppsFlyer's CS team, it became a structural advantage.

"We got the first real opportunity to become the educator," Ziv explained. Customers didn't know what the changes meant for their measurement strategies. They came with questions. AppsFlyer's CSMs had answers — and over the next three years, that educator positioning became the team's primary source of trust and access.

The lesson isn't specific to privacy changes. The pattern is: when customers face genuine uncertainty about an external shift affecting their business, the vendor who shows up with clarity owns the relationship through the entire transition. "They come with some questions and then you are then leading the way in that process," Ziv said.

Today, Ziv sees AI creating the same window. Customers don't know what AI will do to their jobs, their roadmaps, or their teams. The CS organizations that develop and share a credible point of view now will earn the same disproportionate trust that AppsFlyer earned during the privacy era.

Why the CSM Meeting Is the Core Unit of Performance

Ziv is specific about where CSM performance is won or lost: the meeting. Not the QBR. Not the renewal call. The individual 30-minute conversation.

"The CSM that understands the objectives of the customer and is able to come to the meeting very prepared and to speak about what is really important and interesting for that customer on that 30 minutes, they also win the trust and the credit for that customer will want to see them next week and give them another 30 minutes."

The flywheel works in both directions. Show up unprepared and irrelevant, lose access. Lose access, fail the account.

AI has changed the economics of meeting prep in ways most CS teams haven't fully processed. Ziv's personal workflow is concrete: before any customer meeting, he asks Claude to generate the five questions he must ask. "It doesn't matter how better I am in 2026, Claude is more strategic than me." After the meeting, he brings the transcript back and gets a summary he describes as something he "could never add before."

The implication for CS leaders is significant. Domain expertise that used to take years to develop can now be compressed into 20 minutes of focused preparation. As Ziv put it, with the right model and prompt, a CSM can walk into a room and "not only be on par" — they may be "above some of their knowledge in some of the things."

What Founders Get Wrong About Hiring CS Leaders

When founders ask Ziv what to look for in a first CS hire, he gives a specific answer that has hardened over time: domain fluency in the customer's industry is non-negotiable.

"It must be someone that's fully connected to the industry or even the genre, the specific sub." His example is precise — a lawyer who becomes VP of CS at a legal software company has a structural credibility advantage that no amount of CS training replicates. The same principle applies in any vertical. A marketer who spent years running media mix models before joining AppsFlyer can speak to gaming customers at a level that immediately signals peer credibility.

That credibility cascades. The leader sets the standard, which shapes how managers hire ICs, which determines how the entire team shows up in customer conversations.

The Infrastructure Problem No One Is Solving Yet

Ziv's most forward-looking observation: the CRM was built for a different era, and CS leaders need to start thinking about what replaces it.

Today's Salesforce data architecture treats all accounts with the same structural logic — a fixed set of fields that a CSM fills in. But a strategic account that just crossed seven figures in ARR needs dynamic data that doesn't exist yet. "The data structure of Salesforce today is not servicing the need of the company today," Ziv said. AI will surface data points for specific accounts that no one has thought to create fields for yet.

His advice isn't to abandon existing CRMs. It's to recognize the gap and start building toward it. "The CRM — it's not for our era anymore." Founders and CS leaders who internalize that now will be better positioned when the tooling catches up.

The throughline across all of it: CS done right isn't a support function or a retention play. It's a compounding asset that gets more valuable with every career move, every market disruption, and every meeting where a CSM shows up knowing more than anyone expected.