Latané Conant, Chief Marketing Officer at Parloa, did a Q&A with Greg Kihlström for The Agile Brand Guide’s Expert Mode series. This article is drawn from that conversation.
Marketing’s budget stops at the sale. Acquisition, brand, demand generation — that’s the ground a CMO gets measured on, and what happens after a customer calls with a broken order usually belongs to a COO or a VP of Support. Most of us have made peace with that line. It’s drawn on the org chart and nobody spends much energy fighting it.
Latané Conant thinks the fight over who owns service is the wrong fight entirely. Conant is Chief Marketing Officer at Parloa, the agentic AI company that raised $350 million at a $3 billion valuation in January and sells the software enterprises use to automate their contact centers. Her argument isn’t that CMOs should seize the phones. It’s that the conversations happening on those phones are the richest unused source of customer research in the building, and marketing is out there buying the same insight twice.
Everyone Owns It, So No One Does
Ask a marketing leader who owns customer experience and you’ll get a diplomatic answer about cross-functional partnership. Conant’s read on that diplomacy is less generous, and she’s blunt about the cost of leaving the question unsettled.
“The challenge with customer experience is that everyone owns it, so no one really owns it. That fragmented ownership can leave service disconnected from the rest of the customer journey, even though it has a direct impact on the brand.”
Her resolution is to separate accountability from authority. CMOs don’t run service and probably shouldn’t, but they remain accountable for what service does to the brand — and the practical move she describes is to bring “that dark part of the customer journey to light.” For anyone who followed her work at 6sense, the framing is familiar. She built a career on the argument that most B2B buying happens in the dark before a seller ever sees it. She’s now making the same claim about the half of the relationship that comes after the contract is signed.
What a marketing leader actually does with that is narrower than it sounds, and more doable. You don’t need headcount to transfer. You need read access to what customers say when they call, and a standing place for those findings to land — a monthly review with the head of service, a slot in the quarterly brand review, something with a calendar invite attached. Conant lists what’s sitting in those conversations: where customers found you, which competitors they weighed, what they value, where they hit friction. That’s a research panel you’re already paying for and mostly throwing away.
What You Should Be Able to See on a Dashboard
Parloa’s own Consumer Patience Index — a commissioned survey of 1,001 U.S. adults, which is worth reading with the sponsor in mind — found that the top-ranked pain point in American customer service isn’t hold time. It’s talking to a bot that doesn’t understand you, at 25.9%, ahead of long hold times at 22.8%. So the problem on the table isn’t too little automation. It’s automation that doesn’t work, which puts a vendor selling AI agents in an interesting position. Conant doesn’t dodge it, but she does reframe the question away from the technology and toward measurement.
“To understand that, you need to look at the full interaction. That means effort score and CSAT across every call so you can establish a baseline and see the trajectory over time. You also need to see where customers are hitting friction so you can surface it, inspect it and fix it, and how sentiment changes from the start of the conversation to the end.”
Three things in that answer are worth a CMO’s attention. The first is “across every call,” which is the part that used to be impossible — sampling a few hundred recordings a quarter gave you anecdotes, not a baseline. The second is sentiment trajectory rather than a single post-call score. A customer who starts annoyed and ends satisfied is a very different data point from one who travels the other direction, and a flat CSAT number hides both. The third is customer effort, which predicts churn better than satisfaction does and which almost nobody in marketing looks at.
Conant also points to the unglamorous mechanics — authentication, routing, handoffs. “Getting to a human cannot become another obstacle,” she says, and if a handoff happens, the context should travel with the customer so nobody has to start over. If you want one diagnostic to take into a meeting, ask whether a customer who escalates has to repeat what they already said. The answer is usually yes, and it’s usually nobody’s job to fix.
The Test That Doesn’t Cost Anything
Pressed on whether the Index is really a vendor telling marketers their bots are bad so it can sell them better bots, Conant’s response was flat: “The survey was not about bots. It was about service overall.” Fair enough, as far as it goes. What’s more useful is that she offered a way to check the claim without a purchase order.
“A company can test this within its own service organization without buying anything. Look at your own service interactions and ask where customers are getting stuck, repeating themselves, abandoning the experience or reaching back out because the issue was not resolved. That will tell you very quickly whether your automation is reducing effort or adding to it.”
That’s a real test, and the fourth item is the one to run first. Repeat contacts are already logged in most ticketing systems, and nobody has to agree on a definition of “friction” to count them. If a meaningful share of resolved tickets produces a second contact within a week, the resolution wasn’t one. The same survey found 53.6% of respondents admitting they’d actively circumvented a chatbot or IVR to force a transfer — 43.9% of them by yelling “human” or “person” into a phone. People shouting a single word at a machine until it gives up is a strange thing to find in a dataset, and it’s the clearest available measure of a system people have stopped believing in.
Worth noting what Conant doesn’t concede. Her answer treats the fix as better automation rather than more staffing, which is of course what Parloa sells. The counterweight sits in her own data: 85% of consumers said they’d stick with an automated system that resolves their issue nine times out of ten. Consumers don’t appear to be anti-automation. They’re anti-bad.
The Research You’re Already Paying For
The sharpest part of Conant’s argument is a budget argument, and it points at a line item most CMOs treat as untouchable.
“CMOs spend so much money on market research, win-loss analysis, competitive intelligence, focus groups and other programs designed to understand the customer. Those programs are still valuable, but they can be funded more strategically when CMOs are also using the deeper customer insights already sitting inside service touchpoints.”
“Funded more strategically” is doing a lot of work there, and it’s fair to say she stopped short of naming a number she’d cut. But the underlying claim holds up on its own. You’re commissioning a focus group to learn what customers already told your support line last Tuesday, in their own words, unprompted, at no incremental cost. The examples she gives are specific enough to be believable: service data at one Parloa customer traced a large share of call volume to a single battery issue, and at another it surfaced cross-sell openings nobody in marketing had seen.
Those two findings look nothing alike, which is the point. One is a product defect masquerading as a support cost. The other is pipeline. Neither would show up in a brand tracker.
For a CMO with no authority over the contact center and ninety days to show something, Conant’s sequence is to establish a true CSAT baseline, measure effort, and identify the friction points that recur. And when it comes time to make the case upstairs, she’d bring one number: customer lifetime value. It’s the metric that forces every function touching the customer into the same conversation, and the one a CEO can’t assign to a single department.
That’s the real move here. Not a reorganization, and not a budget fight over who owns the phones. Go listen to what your customers are already saying, then bring it to the people who’ve been paying for worse versions of the same information. Start with the repeat-contact rate. It’ll take an afternoon.
