Baby Unicorns in a Sea of Horses

Sam Chandler didn’t come to customer success from the Bay Area or from account management. She has a film degree from the University of Georgia and started on the retail floor, running guest services at Target before moving into operations. Zendesk hired her because she’d tinkered with the product until she knew it cold, and people kept asking her to teach them how to use it. Ten years and a ladder of roles later, from CSM to Director of Customer Success for Startups, she’s now Director of Customer Success, Commercial at Kustomer. She’s built three scaled programs across two companies and has decided it’s her life’s work.

That outsider path is the tell. Chandler runs customer success like a marketing and community function, not a relationship business, and she builds everything off signals. It’s why AI landed the way it did for her: it lets her find the handful of accounts who would benefit from a differentiated experience across an enormous book. And it’s why she’s so insistent that AI sharpen judgment rather than replace it.

From the Retail Floor to Scaled Success

The moment that set her direction came early, as a CSM at Zendesk, when she kept hitting the question she now hears from her own team: how do I know any of this is working? She could pour activity into an account without guaranteeing the outcome. If effort didn’t predict results, she wanted to know what did.

Customer success had grown out of relationship account management, and that wasn’t where her head went. She started reading product and user-behavior research, then studied how social-media marketing scaled attention, and built her first frameworks around habitual engagement and the signals people actually give you. It reshaped who she hires, too.

“I look for the Lady Gagas. The ones who can take a whole room and make everybody feel like they’re getting individual attention, even when everyone’s in the room together.” — Sam Chandler, Director of Customer Success, Commercial at Kustomer

That instinct, scale intimacy instead of trading it away, is exactly what she went looking to AI to do.

What’s Actually Working

Chandler is wary of the grandiose AI stories people tell. Ask her about her own aha moment and she points somewhere small: trusting Gemini to help her craft a reply to an upset customer with a genuinely nuanced problem, on a day when she decided it wasn’t the best use of the brainpower she had left. Trust, she argues, is built from tiny incremental wins, the Google Maps and Netflix moments, not from building an app with Claude.

Those small wins compounded into her signature tool: a scoring model she calls the Growth Filter Index, built in part with Gemini. It takes the data points she learned to look for across her past scaled programs and scores every account in the portfolio, flagging who needs additional nurturing, who’s a strong candidate for AI products, and which quiet accounts are good brands worth connecting with.

“It used to take me months to sift through the data and become a human calculator. Now the algorithm scores the whole book and tells me who needs the differentiated experience.” — Sam Chandler, Director of Customer Success, Commercial at Kustomer

The results grounded her belief in the whole approach. When she first ran it, the accounts the model surfaced were running 47 points higher in net revenue retention than the rest of the book, and in the months since, that group has grown somewhere between nine and sixteen percent depending on how you cut it. She frames it plainly: finding the baby unicorns in a sea of horses, and giving them the attention they earn.

When the Model Hallucinates

Chandler’s relationship with AI is affectionate and deeply skeptical at the same time. She names her tools, argues with them, and has a running move when an answer smells wrong.

“I’ll say, be honest with me, did you just hallucinate that? I’m not going to be angry, I appreciate honesty. And it’ll say, I’m not going to lie, I did, and give me a completely different answer. It’s a constant reminder not to take anything at face value.” — Sam Chandler, Director of Customer Success, Commercial at Kustomer

The deeper lesson is about expectation gaps. Customers don’t extend AI the benefit of the doubt; a wrong answer isn’t “not what I expected,” it’s simply wrong and needs to be fixed. The fix, she’s learned across a lot of trial runs, is to start earlier, by asking what the customer actually wants the thing to do, and whether AI is even the right tool.

“AI is for probabilistic outcomes. Most people use it for deterministic ones, then ask why it didn’t give the right answer.” — Sam Chandler, Director of Customer Success, Commercial at Kustomer

If it isn’t the right tool, she says, save yourself the money. If it is, the real question is whether you’ll commit to the long tail of training, testing, and iteration, which runs longer than most people expect.

The Line She Won’t Cross

Her forward bet is discipline, not more automation. She’s doubling down on intention: knowing why you’re using AI, what habituated use actually looks like, which signals matter. She watches teams reach for tools like Clay and Apollo and blast their entire market, burning through their TAM for no return, and she watches inference costs climb because people reach for AI on work that never needed it. The tools she’s most interested in building are the intelligence and signal layers that help people use the right tool at the right time.

Underneath it is a hard line she draws between cognitive offloading and cognitive surrender, a distinction she picked up from a Wharton paper. Offloading the easy stuff is fine, the way calculators were. Surrendering the strategy is not.

“If AI is telling you your strategy, that’s a fragile foundation, and it’ll show sooner or later. If you surrender the thinking, they don’t need you there. You’re just a calculator they can buy cheaper.” — Sam Chandler, Director of Customer Success, Commercial at Kustomer

That’s also her advice to the CSMs and account managers watching their roles shift. The job is changing right now. Companies used to pay well for relationship-building alone, and that won’t be enough as signals get better and CROs absorb more of the org. The people who last will bring technical skill and run their own signal-based campaigns; she’s already watched scaled CS diverge into GTM operations and growth roles. The marketer’s instinct that got her here holds all the way through: find the few accounts that matter, and keep your own brain firmly in the loop.

The Sonora Project captures how CS leaders are navigating AI — the real experiments, practical wins, and hard-won lessons shaping the future of customer success.