
The Middleman Always Comes Back
Part 1 of 3: Disintermediation and the fight for the front door
The Kettle
Picture a woman in Parramatta whose kettle has just died. She doesn't open a browser. She says to her phone, "Replace the kettle." That's the whole instruction.
The agent that hears her already knows a great deal. It knows the last kettle lasted three years and the one before it lasted eleven. It knows she hates the ones with the plastic spout, that she has a budget in her head she has never said out loud, and that she is home on Friday and not on Thursday. It knows she once returned a toaster to a company that made her print a label, and it has quietly marked that company down. It searches a dozen merchants, compares, rejects the cheap one with the suspicious reviews, and buys a stainless one from a seller she has never heard of. A parcel turns up on Friday.
At no point did she visit Amazon. Amazon may well have packed and shipped the thing. She never saw its storefront, never scrolled its results, never clicked a sponsored listing. Amazon did the work and somebody else got the customer.
Now here is the part that should bother you. Amazon is the company that was supposed to be the end of the middleman. It spent twenty years eating shopkeepers, wholesalers and booksellers on the promise that the customer could go straight to the source. And here comes something that stands between Amazon and its own customer.
The pattern is older than Amazon, older than the internet, and older than the word for it. I'll argue that every wave of disintermediation ever promised has ended the same way. The middleman didn't die. He moved house.
What the Word Actually Means
People use "disintermediation" as though it means one thing. It means at least four, and confusing them is how clever people talk themselves into stupid predictions.
The first is removal: the producer sells directly to the buyer and the intermediary is simply gone. This is the version everybody imagines and the one that almost never happens in full.
The second is unbundling: the intermediary's job is split into pieces and handed to several new actors. The old travel agent searched, advised, booked, held your money and sorted out the mess when the flight was cancelled. Those functions didn't vanish. They scattered.
The third is reintermediation: a new gatekeeper arrives and turns out to be more powerful than the old one.
The fourth is the one that matters for this essay. I'll call it displacing the interface: the old intermediary still fulfils the transaction, but it no longer owns discovery, choice or loyalty. It becomes plumbing.
Here is the analytical point most commentary skips. An intermediary is not just friction with a salary. It does real work: search, trust, aggregation, financing, distribution, standards, insurance, dispute resolution, and the very scarce business of getting a buyer's attention. Remove the intermediary and all of that work still has to be done by somebody.
I call this the conservation of intermediation. The work doesn't disappear. It changes address. The only open question is who now lives at the new address and what they charge for the privilege.
Banks Did It First
The word did not begin with the internet. It began with money.
As I understand it, from the 1930s US regulation capped the interest banks and savings institutions could pay on deposits (the rule was called Regulation Q). That was fine while market rates sat below the cap. When rates climbed past it in the late 1960s and 1970s, savers noticed they could earn more by pulling their money out of the bank and putting it into Treasury bills, and then into money market funds, which I think first appeared in the early 1970s. Money went around the bank. That was disintermediation, and bankers used the word as a swear word.
Look at what happened next. Banks did not die. They invented new products, lobbied for the caps to be lifted, and bought or built the very funds that were draining them. The money market fund became a profitable intermediary in its own right, with fees, a brand, and a customer who now trusted a fund manager instead of a teller. The saver cut out one middleman and acquired another.
Hold on to that template, because it never changes. A rule, price or structure makes the incumbent intermediary too constraining. Customers route around it. The routing itself becomes a business. The business grows fees, discretion and a regulatory department.
The word, notice, never meant "technology makes middlemen vanish". It meant capital moving around a regulated gatekeeper, and the gatekeeper adapting. The tech-utopian version came later and left out the ending.
The Catalogue and the Counter
The impulse to cut out the middleman is much older than the word. Montgomery Ward put out a catalogue in the early 1870s, and by the 1890s Sears, Roebuck was mailing a book the size of a brick to farm families who had one general store within a day's ride. The pitch was pure disintermediation: buy from the source and keep what the storekeeper would have pocketed. The local retailer screamed about it, and I'm told that in some towns the catalogues were burned in public.
Singer did the opposite and arrived at the same place. From the middle of the nineteenth century it built its own shops and sold sewing machines on instalment credit, skipping the independent dealer altogether. That sounds like a clean win until you look at what Singer had to build to do it: shopfronts, salesmen, a repossession department, a credit book and a repair network. It had not removed the intermediary. It had hired one, then put him on the payroll and called him a division.
This is the counterpoint every direct-to-consumer evangelist forgets. When a manufacturer cuts out the distributor, it inherits the distributor's job. Inventory risk, customer service, returns, freight, marketing and credit all land on its desk. Plenty of manufacturers have gone direct, discovered how much of the middleman's margin was payment for real work, and quietly gone back to wholesale. (Hello Nike!)
The middleman had not vanished. His work had changed address, and the new tenant was often worse at it.
Next in the series: the web promised every producer a direct line to every customer. What happened instead is the story of Amazon.
