Nobody buys from the vendor
Almost nobody who runs enterprise equipment bought it from the company that made it. They bought it from a . That reseller bought it from a . Somewhere inside the distributor there was an engineer who helped the reseller design it, size it and prove it worked, and who never met the customer at all.
That was my job for most of a decade — first from a reseller, then from two distributors in a row, supporting partners across Brazil. The role itself is described in the catalogue, and the path a product takes draws the whole chain. What I want to write here is the thing that decade taught me that I have not seen written down anywhere, and it took me years inside the job to notice it.
The property nothing else has
Every other seat in this industry has a preference, and it is structural rather than a character flaw.
The 's engineer wants their product to win, and should. The reseller's engineer wants the deal that closes and that they can support, which is usually the line they have most people trained on. The customer's engineer wants the thing they already know, or the thing that will not embarrass them, and both are rational. Even a consultant, the archetype of independence, is engaged and paid by somebody with an outcome in mind.
The distributor's engineer is different, and the difference is mechanical. A distributor carries competing vendors at the same time. In the years I did this, the same shelf held load balancers, firewalls from more than one house, wireless, and test equipment, and the distributor's revenue did not care which of them the reseller chose. My job was to help the reseller pick the one that would actually work, because the reseller coming back next quarter mattered more than any single line item.
That is an unusual place to stand. You get to hold two competing products, both of which you have been trained on, both of which you have broken in the lab, and give an honest answer about which fits — with nothing riding on the answer except being right the next time as well.
I did not appreciate this while I was doing it. From the inside it felt like being nobody's first choice: not the vendor, not the reseller, not the customer, just the person on the phone at the layer below. It took leaving to see that the layer below was the only one with a clean view.
What that view actually shows you
Three things you learn there and almost nowhere else.
Where products break in the field rather than in the datasheet. A vendor's engineer sees escalations that reached the vendor. A customer's engineer sees their own estate. The distributor's engineer sees dozens of deployments across dozens of resellers and every vertical, and the same three failure modes appear again and again until they stop looking like accidents and start looking like properties. Most of what I now teach about how these systems fail, I learned from the pattern rather than from any one incident.
Which product claims survive contact. In a proof of concept you find out in an afternoon which features exist in the way the slide implied. Doing that repeatedly, across competing products, for people who will be stuck with the answer, is the best calibration exercise the industry offers, and it is why I still read a datasheet as a set of claims to be checked rather than a description.
How much of the outcome is not the product. The same appliance succeeds in one customer and is ripped out at another, and the variable is nearly always sizing, integration or the honest limits nobody stated at the start. That is not a technical finding. It is the reason I stopped believing that recommending the best product was the job.
What it cost, and what it was worth
The honest ledger has a debit side. The distribution years were the least visible of my career: no customer knew my name, no vendor counted me as theirs, and the work leaves almost no public trace. When I later needed to show what I had done, the channel decade was the hardest part to evidence, because its output was other people's successful projects.
There is a version of the founders article argument here. That piece observes that a history told through the people who moved makes invisible the people who stayed. The channel is the same shape one level down: an industry described through vendors and customers makes invisible the layer that connects them, and that layer is where a very large share of the actual engineering judgement lives.
The credit side is that I came out of it with the one thing I could not have got anywhere else: the habit of assessing a product without needing it to be the answer. Every course I now teach is a vendor's official curriculum, and I deliver it as the vendor wrote it, because that is the job and the material is good. But the students always ask the other question — is this actually the right tool for what I have? — and the reason I can answer that honestly is ten years of not being paid to care which way it went.
If you are in that seat now
Two things, from someone who did not know either of them at the time.
You are accumulating something rare and it does not look like anything. The cross-vendor pattern recognition you are building by supporting five product lines at once is not available to the specialists you feel junior to. They know one thing deeply. You are learning what is true across things, which is a different and slower kind of knowledge, and it becomes visible about a decade in.
Write down what you see. The comparisons you make in a partner's meeting room are, in aggregate, the most honest technical assessment anyone in the chain produces, and they evaporate the moment the meeting ends. Nobody in distribution is asked to publish, so nobody does, and the seat with the best view is also the one that leaves the least record. That is the actual reason this post exists.