This article is based on a Q&A with Fusion Risk Management Chief Marketing Officer Richard Barnett
Resilience belongs to other people. It’s a risk function, an IT function, a line item in the compliance budget. When the systems go down, marketing’s job starts — the holding statement, the sentiment monitoring, the apology campaign, the quarter spent buying back goodwill. Most of us have accepted that arrangement without much argument, and it doesn’t feel unreasonable. We don’t run the dispatch system.
Richard Barnett thinks the arrangement is more expensive than it looks. He’s Chief Marketing Officer at Fusion Risk Management, which sells enterprise resilience software, so his commercial interest is on the table from the start. But his argument isn’t really about software. It’s about timing. By the moment an incident begins, the decisions that determine how much revenue survives it have already been made — and marketing wasn’t in the room for a single one.
Restoring Every System Isn’t the Same as Restoring Revenue
Barnett’s complaint is with the category resilience gets filed under. Risk teams start from exposure. IT starts from systems. Those are the right places for those functions to begin, and he says so plainly. The problem is that nobody in that group is starting from the customer.
“The mistake is treating resilience as a recovery discipline instead of a business-performance discipline. Risk, IT, and compliance are essential, but they naturally begin with systems, controls, and obligations. The CMO begins with a different question: What promise did we make to the customer, and what happens to growth if the business cannot keep it?”
That difference shows up most sharply in the recovery sequence. When the order of restoration gets set without marketing in the conversation, it tends to be engineering-rational rather than revenue-rational — the internal inventory dashboard comes back before the order-status page, because one is upstream of the other. Barnett names three things marketing forfeits by being absent: the chance to say which journeys matter most before priorities are locked, the time it takes to reconstruct an audience and channel plan mid-incident, and credibility. The last one compounds. Every campaign creates an expectation, and repeated failures convert that spend into skepticism. The bill arrives as refunds, contact-center volume, abandoned transactions, higher customer acquisition cost, churn, slower renewals, and sales reps defending the company’s dependability instead of selling anything. Marketing then, in his words, “spends money rebuilding demand and confidence it already paid to create.”
Repetition Teaches Customers What to Expect
On July 28, the FAA issued a nationwide ground stop for American Airlines after what the airline described as a technology issue affecting connectivity for some of its systems. The stop went into effect around 6:30 p.m. ET and was lifted at 7:18. Under an hour. Cirium data cited by the Associated Press put on-time departures for the day at 37 percent, though severe East Coast storms were disrupting the same evening. American had also grounded flights over a technical failure on Christmas Eve 2024. Barnett declined to characterize the root causes as identical, which is the right call — but he doesn’t think customers are doing that arithmetic either.
“Repetition is usually more damaging than duration because it changes what customers expect from the brand. A single outage may be understood as an exception. Repeated outages—even when the technical causes differ—can teach customers to build unreliability into their choice. Once that happens, the operational issue becomes a commercial one.”
The damage arrives in waves, and only the first wave looks like an incident. Hours: service contacts, negative sentiment, abandoned transactions, traffic to the status page. Weeks: softer conversion among exposed customers, more expensive acquisition, weaker repeat purchase. A quarter or more: brand preference, customer lifetime value, pricing power, and a reliability objection that sales now has to clear on every call. Which is why Barnett wants a second number next to recovery time — decision time. How fast did the company identify which services and customers were hit? How fast did it size the financial exposure and settle on one accurate version of events? A 45-minute systems interruption can create a much longer customer disruption, he says, if the next several hours go to reconciling facts and chasing approvals.
A Seat Without Decisions Is Theater
Put the skeptic’s version of this to him — I don’t control infrastructure, I can’t fix a dispatch system, and a resilience committee is one more meeting — and he doesn’t push back. He agrees.
“I would agree with the premise: a CMO should not sit on a committee simply to add a customer logo to the governance chart. If marketing has no defined decisions, deliverables, or role in exercises, the seat is theater.”
What makes it not theater is a specific list of things marketing owns. Before an incident: identifying the promises that define the brand, mapping those promises to the operations that deliver them, segmenting which audiences need different treatment, and setting the escalation triggers and approval paths so nobody’s hunting for a legal sign-off at 9 p.m. During one: bringing customer intelligence into the room, translating operational facts into language a customer can act on, and feeding response data back to the operators. None of that is an infrastructure decision. They’re commercial decisions with operational dependencies, which is a different thing, and it’s where decision rights actually get contested. Barnett offers a test that’s easy to run and slightly uncomfortable to answer: if pulling the CMO out wouldn’t change the prioritization, the exercise design, the communication thresholds, or the customer-recovery plan, the role was never designed properly in the first place.
The Lesson From a Component That Cost Less Than a Penny
His evidence for this doesn’t come from Fusion. It comes from Supplyframe, where he was CMO during the pandemic and the global semiconductor shortage.
“Manufacturers were discovering that a constraint in a component costing only a few dollars—or sometimes less than a penny—could put an entire product launch and enormous downstream revenue at risk. At the same time, the traditional ways that electronics suppliers engaged customers—trade shows, field meetings, and relationship-driven information sharing—were disrupted almost overnight. Supply continuity and customer communication failed along the same fault line: critical intelligence existed, but it was not reaching the right decision-maker at the right time.”
Marketing’s contribution there wasn’t allocating chips. It was building a shared way to see the gap. Supplyframe’s own 2021 survey of 180 electronics distributors and suppliers scored the industry at an average of 2.4 out of 5 on a five-level digital customer engagement maturity model Barnett helped develop — a number that’s arguably more damning for being unremarkable. Nobody was in crisis. Everybody was just far enough behind that the intelligence couldn’t move. He also pushed hard on “shifting left,” pulling supply and risk signals into design decisions before a product was locked and a shortage turned into a launch failure. Engineering and procurement still made the sourcing calls. Marketing supplied the context: which launches mattered, what had been promised to whom, and which audiences needed to hear from someone before they heard from a competitor.
So the Monday-morning version of this isn’t a governance fight. Barnett’s advice is to skip the committee request entirely and book one working session with the COO, Chief Risk Officer, and CIO. Pick one credible disruption to a customer-facing service and walk it end to end. Which promise fails first, who’s exposed, who decides the recovery sequence, and what is marketing authorized to say and when. The deliverable is one page — priority audiences, escalation triggers, channel owners, approval paths, holding language, update cadence. Then test it in an exercise.
The useful part is that this can be scored well before anything breaks. Percentage of critical journeys mapped to operational dependencies. Minutes required in a drill to identify affected customers and size the exposure. Time to first accurate communication. Number of approval handoffs, which is usually the embarrassing one. “The goal is not for nothing to happen,” Barnett writes. “Disruptions will happen.” The goal is that when the next one lands, marketing already knows which promise is breaking and who needs to hear about it — instead of learning both from the same alert as everyone else.
