Iain Schmitt

---

The Strangest Carrier: Southern Linc

Any firm's contingency plans or redundant infrastructure says a lot about the business that they are in and what tradeoffs they are faced with. As an example, if your company's HR software vendor was inaccessible for a day because of an AWS availability zone failure, you probably have bigger problems that day. Beside the point that AWS customers are far likelier to bring an outage on themselves, the impact to a company of HR software being inaccessible for a work day is not an "everything grids to a halt" situation. Some offer letters will be sent out late and any pending changes in benefits wouldn't be acted on, but it isn't the end of the world. Lattice is the first company that came to mind here, and if Lattice engineering leadership is trying to prioritise between a) lift-and-shift capabilities to another public cloud vendor and b) new features that will differentiate the product in the marketplace, it's pretty clear that they should go for the latter.

My former employer, RELEX Solutions, is in a different situation. Their Forecasting & Replenishment solution is a retail supply chain SaaS offering that takes in retailer data about products, locations, sales, and other similar information. This data is used to calculate order proposals to make sure that retailers have enough items to satisfy demand but not too much as to introduce spoilage or other costs that come with surplus inventory. This means that if RELEX is down, orders aren't going out. In contrast to Lattice, RELEX Forecasting and Replenishment is a load-bearing part of the business of their customers and this really shaped what Site Reliability Engineering work was like there. Even if a primary RELEX data centre is unavailable, RELEX customers are still relying on the service to route trucks. Because of this, we had the ability to fail over to warm reserve data centres. I quite enjoyed this level of criticality: our services were not matters of life-and-death, but customers really did rely on RELEX to be able to run their business: I had a certain amount of pride in this.

While warm reserve data centres are a pretty significant contingency plan, firms responsible for critical services in the physical world have a much higher bar to clear. Rather than keeping a handful of data centres available, they generally have more ground to defend and higher stakes when things go south. Any major airport employs a small army of air traffic controllers, security personnel, and firefighters to respond to any number of aviation disasters. Hospitals plan and train for any number of mass casualty events from natural disasters to dangerous epidemics. What I hadn't thought of until very recently was 'do utility companies have a backup plan if wireless networks are unavailable', and this takes us to one of the more impressive company contingency plans that I've heard of.

Southern Company is a gas and utility company best known for it's subsidiaries Georgia Power, Alabama Power, and Mississippi Power serving over 8 million customers.¹ As the regulated electricity monopoly in its service areas, it is responsible for transmission from power plant to the service drop into homes and businesses. If Southern Company was completely reliant on the big three wireless carriers they would be taking a great risk during hurricane season: communicating with crews restoring power to customers would be hampered by interruptions in someone else's network. This is apparently too great of a risk for Southern Company, because they operate an independent LTE network covering over 122,000 square miles, a network which almost completely overlaps with their utility service area.

One of the main tenets of telecommunications unit economics is that building networks has tremendous fixed costs, but the marginal cost of a new customer is miniscule. Southern Company built their LTE network at great expense, necessitating winning spectrum auctions, constructing towers, and setting up a backhaul network connecting cell sites to the network core. Once all is said and done, they were left with a network that designed for the needs of one company's critical infrastructure but almost certainly has excess capacity to spare. Unsurprisingly from this framing, Southern Company has done exactly what would make economic sense and sells this excess wireless capacity. Southern Linc is a wireless carrier that is much less famous than its corporate siblings underneath the Southern Company umbrella. Even if you live in Southern Linc's service you've likely never heard of this carrier give that its other offerings include "mission-critical push-to-talk" and fleet truck tracking: not exactly a concern for consumers and even most businesses. ² The "Southern Linc at a Glance" page reads: "for more than 25 years, Southern Linc has been the wireless network provider built from the ground up for utilities, government and business".³

While there is nothing I'd rather do more than compare Southern Linc's subscriber growth and financial performance against the big three, heartbreakingly the Southern Company quarterly reports do not break down this information. As a disclaimer I am a CPA in exactly zero jurisdictions, but it looks like this is spelled out in the Financial Accounting Standards Board rules, rules that require any company segment responsible for more than 10% of parent revenue, P&L, or assets to be reported separately.⁴ Unsurprisingly, Southern Linc does not meet this 10% criteria. What we can see is that Southern Linc falls into the "Other Revenues" bucket, which brought in $283 million in Q3 of 2024 for $820 million year-to-date. As a small consolation prize, in that same 10Q the company reported a $50 million contribution to year-to-date revenue by what it described as "unregulated sales at Georgia Power". I'm sure "unregulated sales" are one of the interesting asterisks in the regulated utility monopoly model in most of the United States. ⁵

Given that Southern Linc is comparatively more obscure than the big three, the information available about it is more of a PR flavour than ideal, but much of it is still useful. According to a 2022 interview with telecommunications trade press Urgent Communications, the then Southern Linc CEO Tami Barron explained that the carrier had redundant network cores in Atlanta and Birmingham with two generators and 14 days of fuel on-site. As far as cell sites are concerned, “We have cell-site redundancy—from a last-mile [standpoint]—at probably 80% of our cell sites. Interestingly, at every cell site, clearly we have battery backup to about eight hours. But we also have on-site fuel-cell [sic] or [power] generation to the tune of five days.” Barron also explained that much of Southern Linc's backhaul is on fibre owned by Southern Company. I'm less surprised to hear about a utilities company own fibre than its own LTE spectrum, but in what seems to be a Southern Company tradition they resell their excess fibre capacity through their Southern Telecom subsidiary. ⁶

Southern Company isn't the only electrical utility company operating a private LTE network; Xcel Energy operates such a network across 8 states by leasing spectrum owned by Anterix. ⁷ But Southern Company is the only utility that I am aware of which operates a wireless carrier selling to external customers on its own network. The utilities industry is relatively unique in how critical it is, the technical capabilities required, and the geographic range that it entails. Few other industries would justify such an investment, but the fact that Southern Linc is so relatively unique likely implies that there isn't an overwhelming case for such expansive investments in wireless communications for most utilities. If I were to speculate, the carrier's existence can be credited to historical path dependence and factors unique to the deep south.