Being your customer's thought partner

For about a decade, the phrase every customer success team reached for was “trusted advisor.” It sounds great in a QBR deck and even better in a performance review. The problem is that most people quietly believe it means something it doesn’t. They think a trusted advisor is the person in the room who has the answers, who has seen this movie before, and who can calmly guide the customer from where they are to where they should be.

That definition worked when the playbook was stable. You’d onboarded a hundred accounts like this one. You knew which features mattered, which rollout sequence reduced churn, and what month three usually looked like. Being the advisor meant being the one who’d already walked the path.

That world is gone, at least for the part of the job that matters most right now.

Nobody has the map

Walk into any customer conversation about AI and you’ll feel it immediately. The customer is excited and nervous in the same breath. They’ve been told by their own leadership to be further along than they are. They’ve read the same overheated posts you have. And underneath all of it, they’re looking for someone, anyone, to help them figure out what to do next, even if it’s just one workflow or one corner of how their team works.

Here’s the uncomfortable part. You don’t have the map either. Neither do I. The honest answer to “where does this all end up” is that nobody knows, and the people claiming otherwise are selling something.

This is exactly why “trusted advisor” buckles under the weight. You can’t advise from a position of certainty in a space where certainty doesn’t exist yet. If you wait until you’ve got it all figured out before you bring it to a customer, you’ll be waiting a long time, and you’ll be useless to them in the meantime.

Thought partner is the better job

The framing that’s worked for me is to drop “advisor” and pick up “thought partner.”

The difference isn’t cosmetic. An advisor is supposed to know. A thought partner is supposed to think alongside you. One of those is achievable in a moment where the ground keeps moving, and the other one is a setup for either paralysis or pretending.

Being a thought partner means walking into the room and saying some version of this out loud: I don’t know exactly where this goes. I’m navigating it too. But I’m paying close attention, I’m trying things, and here’s what I’m learning. Let’s make this a space where we can experiment, compare notes, and be honest about what’s working and what isn’t.

That sounds modest. It isn’t. For a customer who is anxious about falling behind and surrounded by vendors performing false confidence, a partner who will actually sit in the uncertainty with them is rare and valuable.

An advisor is supposed to have the answers. A thought partner is supposed to think alongside you. Only one of those is honest right now.

The shift worth making

What lets you be that partner

You can’t think alongside someone about a subject you’ve never engaged with. The reason this role is hard isn’t that the relationship skills changed. It’s that the relationship now depends on a fluency most of us are still building. So before any of this works with a customer, you have to do the unglamorous work on yourself.

For me it started embarrassingly simply. I committed to a few months of listening to AI podcasts on my commute, the same handful, week over week, not to find a single answer but to hear smart people develop their thinking over time. The compounding effect was real. After a while I could hold a conversation, form an opinion, and disagree with something without feeling like a fraud. That’s the whole game at the start: enough exposure to stop being scared of the topic.

I’d break the fluency into three things that feed each other.

The mistake I made early was treating these as a ladder you climb in order. They’re not. They happen at the same time and they accelerate each other. You learn faster when you’re building, you build better when you’ve formed a view, and you form sharper views when you’re actively learning. Treat it as a menu, not a staircase. Pick where you start based on what you’re most curious about that week.

The limiting factor is not the tool

The thing that surprised me most is where the ceiling actually sits. It’s not the model. The tools are more capable than our ability to ask them for useful things. The real constraint is creativity, the human kind, the part that asks “what’s the unexpected way this could solve the problem” or “where exactly should the handoff between the machine and me live.”

I watched someone solve a customer problem that I would never have thought to point software at. The customer wanted to identify businesses with two dumpsters out back, because that detail predicted something about their buying behavior, and then target them with a specific campaign. The tooling could do it. What I lacked wasn’t the tool. It was the imagination to frame the question that way in the first place.

That realization changed what I spend my non-working time worrying about. Staying in the conversation still matters, so I keep doing it. But I’ve started caring more about stretching the creative muscle, because that’s the input that’s actually scarce. The most practical advice I can give a customer success leader who feels stuck is almost annoying: go do something that has nothing to do with work, and protect the part of your brain that comes up with strange, good ideas. The fluency gets you in the door. The creativity is what makes you worth talking to once you’re inside.

What I tell leaders who feel behind

A lot of the leaders I talk to are in a hard spot. Senior leadership is telling them to be “AI first,” they don’t personally feel like early adopters, and they don’t have an obvious AI native person on the team to lean on. The paralysis is real and I have sympathy for it.

Two things help more than anything else.

First, lower the cost of trying. Give people a small budget to expense AI tools, the order of twenty dollars a month, and the explicit permission to play. Let them break things in a safe environment. Make it clear that the embarrassing failures, the agent that fires off an email it shouldn’t have, are part of the deal and not a fireable offense. The only failures worth fearing are the catastrophic ones, and you can usually design those out. Everything below that line is tuition.

Second, put your experimenters on a pedestal. The fastest way to move a team is to show them a peer with the same job, the same lack of a coding background, and the same daily workload who built something small that genuinely helps. Nothing reduces fear like watching someone next to you do the thing and live to tell about it. A little healthy FOMO is a feature, not a bug, as long as it fuels curiosity instead of freezing people.

The work that stays

Here’s what I keep coming back to. The relationship was never the part AI was going to take. Reading the room, knowing when to push and when to listen, sitting with a customer in a genuinely uncertain moment and being useful anyway. That’s the human work, and it’s becoming more valuable, not less.

What’s changed is the standard for showing up. You used to be able to coast on having walked the path before. Now the path is being drawn in real time, and the customer can tell the difference between someone performing certainty and someone actually thinking with them.

So stop trying to be the advisor with all the answers. Be the partner who’s genuinely in it, building the fluency, stretching the creativity, and willing to say “I don’t know yet, but let’s figure it out together.” In a moment when nobody has the map, that turns out to be the most useful thing you can be.