Have We Seen This Movie Before?

The Central Question

Johnson sets up his talk around a straightforward but generative question: is what is happening right now in the agency world — the upheaval driven by AI — actually unprecedented, or have we been here before?

He acknowledges that the fear is real. He references a talk given earlier in the same event by Karim Marucci, who shared that many of his enterprise clients in the latter part of the previous year and the early part of the current year were simply paralysed by the scale and pace of AI-driven change. Johnson uses that as a jumping-off point to argue that paralysis is understandable but not useful, and that history offers a more grounding perspective.

His talk is structured around two historical case studies. The first is the invention of the telephone and, more specifically, the 50-year period when a human operator was required to complete every single phone call. The second is the ice harvesting industry and its collapse in the face of mechanical refrigeration. Both, he argues, are mirrors for the moment the agency world is currently living through.

Case Study One: The Telephone and the Human in the Loop

Alexander Graham Bell and the Patent That Started It All Johnson begins with Alexander Graham Bell and the founding moment of the telephone industry. On February 14th, 1876 — Valentine's Day — Bell's representative filed a patent application titled "Improvements in Telegraphy." Johnson highlights this deliberately: the word "telephone" does not appear in the title of the application, and it does not appear anywhere in the body of the patent itself. On March 7th, Bell was awarded patent number 174,475, which Johnson notes has been called by many the most valuable patent in history.

The timing of that filing turned out to be critical. Bell's application was submitted in the morning. Later the same afternoon, a competitor named Elisha Gray — co-founder of Western Electric — filed something called a caveat, a formal notice that he had also invented a telephone. Because Bell's application arrived first, he held the right to defend the patent.

Thousands of lawsuits challenging Bell's patents followed over the years, but Bell's priority was repeatedly upheld.

Johnson adds several biographical details he found while researching that he describes with evident enjoyment. Bell's principal investor at the time was also the father of his fiancee, and the man had made it clear that Bell could not marry his daughter until the telephone was working. He also, Johnson notes with some amusement, would have preferred Bell focus on improving the telegraph rather than chasing the idea of transmitting the human voice over wire. The idea of voice transmission was at the time considered not just unlikely but in some quarters genuinely mystical or even demonic.

On March 10th — several days after the patent was awarded — Bell finally got the device working. On March 13th, he bombed an investor demonstration. His public moment came on June 25th at the Centennial Exposition, the event marking the 100th birthday of the United States. One of the judges evaluating demonstrations that day was the scientist who came to be known as Lord Kelvin. Bell also received a second patent in January of the following year covering improvements that allowed the transmitter and receiver to be combined in a single device. In May of that year, he finally monetised the invention, but by renting phones rather than selling them — a decision Johnson flags as strategically significant.

The Operator as User Interface Having set the scene, Johnson shifts to his central point about the telephone's first 50 years of existence. He shows an image of an early telephone device from the 1880s — not a Bell original but a Western Electric model from shortly after — and asks the audience a question: what is missing from the user interface?

The device has a microphone, an earpiece, bells at the top for incoming calls. What it does not have is any way to place a call. There is no dial.

That is because for the first 50 years of the telephone's existence, every call placed anywhere required a human being to physically complete the connection. That human being was the telephone operator.

The mechanism was a switchboard and a quarter-inch patch cable — Johnson mentions he photographed one on his desk that morning, since he keeps one as a guitar cable. When a subscriber picked up their phone, a light lit up on the operator's switchboard. The operator would plug a cable into the socket beneath that light, connecting her headset, and greet the caller. The early operators were known informally as the "Hello Girls" for this reason. The caller would then request a connection — for example, to the local florist — and the operator, who knew her community and its businesses, would complete that connection by plugging a second cable directly into the florist's line.

As telephone networks grew in scale and complexity, the operators needed access to large directories of information. Johnson describes them as an early form of Google. In urban exchanges, operators managed enormous switchboards and kept extensive records. For long-distance calls, one operator would connect a caller to an operator in the destination city, who would complete the local connection. This was the universal interface for the telephone for five decades.

The Scalability Problem and the End of the Operator By the 1930s and 1940s, the volume of calls being connected daily was reaching into the millions. The model of having a human being in the loop for every single call was not going to hold. Johnson is direct about this: you cannot scale a business that requires individual human intervention for every transaction when that transaction is happening millions of times a day.

He draws a parallel here. Operators, he says, were high-touch help. They were a source of expertise. They were also biased — their personal knowledge, relationships, and interests necessarily shaped whose line they plugged into first. And they were not scalable.

The push toward automation came from exactly the kind of problem that bias creates.

Johnson recounts the story of a man named Strauger, an undertaker who is credited with inventing an automatic telephone switch. The lore holds that Strauger discovered a friend's family had used a competing undertaker rather than him, and on investigation concluded that the local telephone operator — who in some versions of the story was married to his competitor — was routing calls for "the undertaker" to the wrong business. Strauger's solution was to take the human out of the loop entirely.

The challenge this created was that there was no speech recognition technology in the 1890s. If there was no operator to listen and respond, the subscriber needed some other way to communicate where they wanted their call to go. That problem eventually produced the rotary dial. The series of clicks generated by a rotary dial communicated with switching systems to route calls automatically. Johnson goes into a little of the technical history here

but stops himself before getting too deep, noting that he did fall into that particular rabbit hole while preparing the talk and found it highly enjoyable.

The result of this automation was that what had been, by some accounts, the largest single occupation for women in the United States became largely obsolete.

What Replaced the Operator Johnson then traces the evolution of how people found the information they needed to connect with businesses once the operator was no longer there to provide it. The sequence he describes: written lists kept beside the phone, then printed directories (white pages), then business directories where companies could advertise (yellow pages), then early directory websites like Yahoo, then early search engines like AltaVista, and finally Google.

And from there, he points out, came the entire industry that the people in the room have been part of. Building web experiences, websites, SEO — all of it was, at its root, solving the same problem the operator originally solved: how does someone who needs a product or service find and connect with the person who can provide it? The agency world, in this framing, stepped into the operator's role. High-touch help, a source of expertise, biased in various ways (the bias of search engine algorithms being something agencies have been wrestling with for years), and operating in a world where the question of scalability never fully went away.

And now there is a new player. First ChatGPT, and then a rapidly expanding ecosystem of AI tools and agents. These represent the next wave of automation of the same underlying problem.

Case Study Two: The Ice Harvesting Industry

A Business That Disappeared in a Generation The second case study Johnson presents is the ice harvesting industry. He describes it as something many people have not heard of, and the image he uses — a frozen pond — is his hook into the story.

From the 1830s through the 1890s, ice harvesting was one of the largest industries in New England and in various parts of the world. The business was essentially what it sounds like: workers would cut blocks of ice from frozen ponds, transport them, store them with insulation, and sell them. In 1886, the largest ice harvest in New England history produced something in the order of 25 million tons of ice.

By the 1920s, almost every ice harvesting business was gone.

The reason is equally straightforward: mechanical refrigeration. The technology to create cold artificially made the labour-intensive process of harvesting natural ice economically obsolete. Johnson notes that his father, who grew up in rural Oklahoma, still calls a

refrigerator an "icebox" — a linguistic holdover from the era when literal ice, delivered to the home, was what kept food cold.

What Ice Harvesting Companies Should Have Understood Johnson uses this as a thought experiment. Imagine you are an ice harvesting company owner in 1886. You have just had a record harvest. Business is good. And you are starting to hear rumours that someone has built a device that can create ice mechanically. What do you do?

His point is that many ice harvesting operators were probably blindsided, not because they lacked intelligence, but because their entire mental model of their business was about the physical work: the labour of cutting ice, the logistics of transport, the infrastructure of storage. They were not likely to be thinking about thermodynamics or the engineering of heat transfer. That was a different discipline entirely.

But Johnson argues that a smart strategic thinker in that position might have asked a different question: what are my customers actually paying for? Not what am I selling — but what are they buying?

The answer, in the case of ice, is not ice. The answer is cold. More specifically, the ability to preserve food, to keep milk from spoiling, to have a degree of control over temperature in their homes and businesses.

If an ice harvesting company owner had understood that the customer was buying cold — not ice — they might have pivoted to manufacturing refrigeration appliances. They might, as Johnson puts it, have become a household name and still be selling those devices today.

Imagine you are an ice harvesting company owner in 1886.

The Strategic Lesson for Agencies

What Were Clients Actually Buying?

Johnson turns this framing directly toward his audience. He asks them to wind back five years and think about the question: what were your customers buying?

The obvious answers, he says, would have been things like web development, SEO, conversion rate optimisation, paid traffic management, social media advertising and marketing, email marketing, inbound strategy, marketing technology. A long list of deliverables.

But those, he argues, were what agencies were selling. They were not necessarily what clients were buying.

When AI makes it possible for a business to build its own website, when SEO as a discipline is being reconsidered in a world where AI agents are increasingly the first point of contact for queries, when ad platforms are automating much of the targeting and creative work that agencies used to provide — what exactly is the client losing? What is it that they actually needed?

Johnson's list of possible answers: expertise that has been developed over a long period of time and can be applied more effectively when augmented by tools. A high level of service.

Strategy. Execution. The ability to think about the client's business in ways the client is too close to see clearly.

His point is not that any one of those is the definitive answer for every agency. His point is that the question needs to be asked, and that this moment of disruption is the right time to ask it.

The Call to Action Johnson's closing argument is about showing up. He says that this is a moment to make contact with clients — literally pick up the phone, or show up in person — and have a different kind of conversation.

He lays out a set of questions he wants agencies to be asking their clients directly: What problems are you trying to solve right now?

What keeps you awake at night?

Why would a particular AI tool or capability make a difference in your business, if you were to adopt it?

How would it affect your business specifically?

The goal of those questions is not just to gather information. It is to reestablish the agency as a strategic partner rather than a vendor of specific technical outputs. The act of asking intelligent strategic questions is itself a demonstration of value — it signals to the client that the agency is thinking about their future, not just executing a service.

He frames the ultimate purpose of agency work as making clients' businesses successful.

Whether the client measures success through e-commerce revenue, foot traffic to a physical location, or lead generation, the underlying purpose is unchanged. Whatever the delivery mechanism — websites, SEO, AI-driven visibility, something not yet invented — what the client has always needed is for their business to work. Johnson argues that agencies that hold onto that understanding, rather than attaching their identity to any particular tool or tactic, are the ones positioned to thrive through this transition as through every previous one.

Post-Talk Exchange

Following the presentation, the host Steph Hudson expressed that she had not previously known about Johnson's interest in history. Johnson responded warmly, describing the switchboard as an extraordinary piece of technology in its own right. Hudson noted that audience member Nancy had shared in the chat that her mother was a switchboard operator in the 1950s and 1960s. Another audience member, Zachary, had answered Johnson's "what is missing?" question about the early telephone with a suggestion that TikTok was missing — to general amusement. Johnson was invited to continue conversations in the networking lounge after the session, and the event moved on to the next speaker, Jimmy Rosen.

About the speaker David G. Johnson formerly Nexess / Liquid Web / Stellar WP

David G. Johnson has held leadership roles at Nexess, Liquid Web, and StellarWP, giving him a long view across hosting, infrastructure, and the WordPress product ecosystem.