Growing Your Business May Soon Require More Power

Visa did not cut 2,600 jobs because revenue slowed. The company grew quarterly revenue 14% to $11.6B, according to Yahoo Finance, while U.S. payments volume rose 10% and Visa Direct transactions climbed 21%. The layoffs came anyway. That is the real signal business leaders should focus on this morning.

CNBC reported the reductions equal roughly 7% of Visa’s workforce and are concentrated in technology and product operations. CEO Ryan McInerney told employees the company needed to “continue evolving how we work” and said new operating systems were accelerating that shift. But the most important detail came from Visa’s earnings commentary covered by PYMNTS: product teams that previously had 10 or more people are now being reorganized into “agentic squads” of two to four.

That sequence matters. Visa did not cut first and hope productivity improved later. It changed the structure of work first. The company deployed more than 150 internally built automation applications across engineering and customer service, then reorganized teams around those systems, then removed overlapping roles. Visa said feature-development time fell more than 65%. That is the operating mechanism almost every wire story missed.

The winning companies right now are not merely “using automation.” They are redesigning accountability around it. Visa is reinvesting savings into affluent consumer products, cross-border payments, stablecoin infrastructure, and business-to-business money movement. In other words, labor savings are funding growth categories, not simply padding margins. That is why the company can reduce headcount while still talking aggressively about expansion.

If you run a services firm, software company, agency, or operations-heavy business, this lands directly on your desk in two places. First, vendors will push harder on renewals because they are under pressure to show productivity gains to investors. Second, smaller competitors that organize work around fewer people and tighter systems will suddenly move faster than firms still routing decisions through large teams.

Watch the next 90 days of enterprise software earnings calls carefully. The key number is no longer seat growth. It is revenue per employee and development-cycle speed. Companies reporting flat headcount with faster release schedules will attract capital and price more aggressively into Q4.

The opening this week is operational, not technical. Pull your five largest recurring software or service contracts and ask one direct question before renewal discussions begin: what work did the vendor automate internally in the last 12 months, and where are those savings showing up in your pricing? Vendors that cannot answer clearly are likely protecting bloated cost structures. The firms already moving are renegotiating before October budget cycles lock in next year's rates.

The strange details in the eBay harassment case made headlines years ago. The important business lesson landed this week. eBay and former executives agreed to pay roughly $56M to settle claims tied to a 2019 harassment campaign against Ina and David Steiner, the publishers behind EcommerceBytes.

According to TechCrunch and Ars Technica, former eBay employees and executives used fake social-media accounts, mailed threatening packages, attempted to track the couple’s vehicle, and coordinated intimidation tactics after critical reporting about the company. Seven former employees ultimately pleaded guilty to criminal charges between 2022 and 2024. Former security chief James Baugh received a prison sentence of nearly five years.

The number that matters for operators is not the settlement alone. It is the structure of the payout. eBay itself paid $46.15M. Former CEO Devin Wenig personally contributed another $2M, while additional executive payments and nonprofit commitments pushed the settlement near $56M. The agreement also required eBay to issue a public statement specifically addressing company culture and executive behavior at the time.

That sequence is becoming more common in governance disputes. Regulators and plaintiffs are no longer separating “bad employee conduct” from executive oversight systems. Internal messages matter. BBC reported that Wenig previously sent a message reading “Take her down” regarding one of the Steiners. His representatives later argued the statement referred to public-relations strategy, not harassment. But once leadership language appears connected to misconduct, the company loses the ability to frame the problem as isolated.

The companies handling this environment best are building escalation systems before conflict starts. Large retailers, financial firms, and healthcare operators increasingly route executive complaints about journalists, creators, customers, or employees through legal and compliance review instead of informal internal channels. That sounds bureaucratic until you compare it with eight-figure settlements and years of reputational damage.

For smaller businesses, the lesson is surprisingly practical. Many founder-led companies still treat Slack threads, texts, or executive DMs as private pressure-release valves. Courts increasingly treat them as discoverable evidence of organizational culture. The legal risk now extends far beyond discrimination or hiring disputes. Vendor retaliation, customer harassment, online reputation fights, and employee intimidation all create the same exposure pattern.

The move this week is simple and cheap. Pull the last six months of executive-channel communications around customers, competitors, critics, or former employees. If your leadership team casually jokes about retaliation, public humiliation, or “teaching someone a lesson,” fix that language now and formalize escalation rules before the next dispute arrives. The operators already doing this will spend the next decade avoiding settlements their competitors never saw coming.

The biggest infrastructure constraint in business right now is no longer office space or financing. It is electricity. PJM Interconnection, the grid operator covering 13 states from Ohio to Virginia, warned this week that some large facilities could face temporary power cuts starting in mid-2027 if the grid becomes overstressed.

According to Bisnow and TechCrunch, the curtailment plan would initially apply to facilities using 50 megawatts or more of electricity, primarily large data centers. Customers would receive notice before cuts and could receive compensation for interruptions, similar to long-running industrial demand-response programs. But the larger signal is what forced PJM into this position. The organization is staring at a nearly 7-gigawatt power shortage after June 2028 despite already running a record $16.4B auction for new generation capacity.

The economics are already moving. Monitoring Analytics reported PJM power-supply costs have risen more than 60% as demand accelerated. PJM now estimates it will need to raise roughly $30B from customers to maintain reliability, and residential bills could rise by about $70 per month by early 2028. Across the Atlantic, UK regulator Ofgem proposed refundable deposits between £237,500 and £712,500 per megawatt for projects seeking grid connections. A one-gigawatt project could require hundreds of millions upfront just to hold a place in line.

The mechanism here is critical. Utilities are shifting from “first come, first served” access toward “prove you can support the grid” access. PJM wants large new users to bring their own generation or contract for new supply. Ofgem wants developers to put real money at risk before occupying grid capacity. In both cases, speculative expansion is getting filtered out by infrastructure costs.

The winners are operators who secured long-term power arrangements before queues tightened. Large data-center groups like Vantage and hyperscale cloud operators already built sites with diesel backup systems because demand-response payments partially offset the expense. Manufacturers and logistics firms with combined heat-and-power systems or fixed-rate energy contracts are suddenly in a stronger negotiating position than businesses still relying on floating utility exposure.

Even if you never build a data center, this matters to your next lease, warehouse expansion, manufacturing line, or office consolidation. Landlords are beginning to ask whether tenants require upgraded substations, generator support, or guaranteed power availability before signing large commercial deals. That changes build timelines, insurance assumptions, and occupancy costs.

The opening is still early. Before year-end planning starts, ask every landlord, utility partner, and site-development contact one direct question: how much unused electrical capacity remains at your primary locations, and what is the timeline for upgrades? Operators that secure power certainty now will move faster than competitors stuck waiting in permitting and interconnection queues by 2027.

The FCC expanded its Covered List this week to block new imports, sales, and marketing of certain foreign-made connected robots and power inverters. Most headlines framed this as another geopolitical dispute. The operational reality is more immediate: warehouse automation, solar installations, and equipment procurement timelines just became more complicated for thousands of U.S. businesses.

According to The Verge, BBC, and PYMNTS, the ban covers “advanced robotic devices” capable of movement, navigation, obstacle avoidance, and network connectivity. The FCC definition is broad enough that future robotic vacuum products could require waivers. The restrictions also apply to connected power inverters commonly used in solar systems and data-center infrastructure.

The mechanism matters more than the politics. Existing approved products are not affected. New models are. That means distributors, contractors, warehouse operators, and facilities managers now face a transition period where replacement cycles and future procurement plans may diverge sharply from current inventories. Businesses that assumed they could simply reorder upgraded versions next year may discover those products require waivers, new certifications, or entirely different suppliers.

This is particularly important because China dominates several categories involved in the ruling. Reuters, cited by PYMNTS, noted China has more than 140 humanoid-robotics companies and also leads global power-inverter production through firms like Sungrow and Huawei. The FCC specifically cited remote-connectivity concerns tied to critical infrastructure and mobile devices.

The businesses already adapting are procurement-heavy operators that diversified sourcing after earlier restrictions on routers and drones. Several industrial distributors and logistics operators spent the past two years qualifying secondary suppliers in South Korea, Japan, Taiwan, and the U.S. before formal bans arrived. Those companies now have purchasing flexibility while competitors scramble to restart approval processes.

For normal operators, this is less about humanoid robots and more about replacement cycles. Warehouses adding automation, manufacturers upgrading connected equipment, and commercial buildings expanding solar infrastructure may all encounter revised lead times and price changes by early 2027. The hidden cost is not the equipment itself. It is the delay created when procurement teams discover approved vendors no longer qualify.

The move this week is defensive and practical. Pull every planned 2027 equipment purchase involving connected industrial hardware, robotics, solar infrastructure, or power management systems. Then ask suppliers one question in writing: are any future product models dependent on FCC waivers or overseas approvals? Businesses clarifying that now will avoid getting trapped between expiring equipment and unavailable replacements next year.