The most hated customers

Customer Service Is Not a Department — It Is the Whole Organisation

I'll never forget the small room – fifteen people on phones, all chatting to customers and helping them with their enquiries. I scanned the room and saw a large whiteboard in the corner. At the top, scribbled in blue pen, were the words "Our Most Hated Customers."

Below the heading was a remarkable thing indeed: a list of twenty or so customers, ranked, with rather derogatory wording next to each name. Oh how they joked about the very people who keep the business alive – paying customers.

You see, every organisation has two brands. The first is the one created by the marketing and leadership teams and agencies, and carefully curated into slide decks, posters on walls and glossy web copy – usually with a playbook for new starters to match. The second is the one you meet when your flight is cancelled, or you phone up to get unstuck with a product. Only one of them is real.

I've collected examples of genuinely poor service over the years – not as complaint, but as study. They're instructive precisely because they're so consistent. A well-known herb and supplement shop that ran its busiest lunch periods with the fewest staff on the floor, letting queues build while bored customers drifted out: frustration rising, and money walking out of the door.

A tyre company whose branded vans are driven so aggressively on the roads around Winchester that meeting one feels like a scene from Mad Max. A train company that answered a letter of genuine praise for a guard's kindness with a standard complaint template and a £100 voucher "for the inconvenience" – I still have that letter, and it reveals exactly how many complaints they were processing and how little they were reading any of them.

Forty-two phone calls before I received the internet service I was already paying for. Twenty more to cancel it, a year later.

None of these are frontline failures. The people on the front line, most of them, genuinely want to do a good job. They're design failures – decisions taken far from the customer, whose consequences the customer absorbs entirely, and the frontline staff often absorb too. When service is poor, it's almost always a signal about the design of the thing: the quality of the hiring, the shape of the processes, the clarity of purpose, the behaviour that gets tolerated, and how far the system supports or sabotages the people inside it. It also tells you, rather precisely, how much you – the customer – are actually valued.

The useful question, then, isn't "how do we fix customer service?" It's "where does service actually get made?" – and the answer is almost never the call centre or front line. It's made in a handful of decisions taken long before the phone ever rings, or the customer is stood at the counter, most of them by people who'll never take a call themselves.

It starts, unglamorously, with hiring. You can't have a good customer experience sitting on top of a miserable employee one, and you can't have a good employee experience if you hire the cheapest people you can find and then treat their judgement as an inconvenience.

I once worked with a service manager who did exactly that – hired for price, measured how fast calls were resolved rather than how well, and was genuinely puzzled that satisfaction kept sliding. She wasn't hiring badly by accident; she was hiring to satisfy the wrong number – salary cost. Pick the wrong number to chase and everything downstream bends to it. And your overall costs typically go up as well.

Then there's the product itself, which is really the best customer service there is – a thing so clear and reliable that people rarely need help at all. Every unnecessary support contact is a design note insight. At one company, when I actually looked at where the calls were coming from, a large slice were simply people locked out of a platform that had no way to reset a password. The support manager's instinct was the usual one: hire more people to answer the calls faster. We added a self-serve reset instead, and ninety-five percent of those calls vanished. The data was sitting there the whole time. It nearly always is – the only question is whether you study it and remove the cause, or staff up to keep answering it forever.

Which is why it's worth stapling yourself to a customer request and following it the whole way through – not as the process modeller has drawn it, but as it actually travels. Where does it sit and wait? Where does the information fall down? Where does it pass through six pairs of hands before anything happens? Where does it wait for Max, who only works Wednesdays and Fridays? Those journeys tell you things no dashboard will, because a dashboard reports are often internal; compliance and targets and measures – but the customer only ever experiences resolutions in a time frame, they don't care what sits behind it.

And when they do finally reach a person, that person has to be allowed to help. A poster reading "the customer comes first" means nothing if the human facing the customer can't make a decision without escalating it two levels up. Care can't be centralised. The people closest to the problem need the authority to actually fix it, rather than take your details and promise a callback everyone in the office knows isn't going to happen.

The temptation, when all this is creaking, is to buy something – a new CRM, a portal, a clever bit of automation. But technology laid over a broken process usually just adds another layer to the friction rather than removing it; now you're stuck and navigating a menu through a screen. Better service almost always comes from cleaner processes, fewer handovers, and a little room for people to think – none of which shows up on a purchase order. Creativity and care, as it happens, are free.

Most real problems need more than one team to solve, too, and this is where they often go to idle – passed back and forth across a fence between departments who share a customer but no obvious collective goal. Work out who genuinely needs to be involved in fixing a customer's problem and build the cooperation around that reality, because a shared cross-functional goal is one of the most underused things a business has.

And then there's time. We've somehow decided that time spent with a customer is inefficiency, so we cap it and measure people on how quickly they can get someone off the phone – which reliably produces worse outcomes, more repeat contacts, and a higher bill overall, while telling the customer, unmistakably, that they're an interruption. Time with a customer was never waste. It's the most legible form of respect a business has.

None of it holds, of course, if the people at the top don't model it – which brings us back, unavoidably, to that whiteboard.

Stop hiding behind process. Stop closing interactions like transactions instead of tending them like relationships. And, for a start, stop keeping a league table of your most hated customers where the whole floor can read it and learn that this, apparently, is how we talk about the people who pay us.

Harry Beckwith left a useful discipline in Selling the Invisible, and I think it's an important point:

Assume your service is bad. It won't hurt. But it will force you to improve it.

Not out of gloom – out of humility. It keeps you looking, and it stops the small deteriorations compounding into the kind the press write articles about. Because customer service isn't just the department you can point at. It's a system, and it improves the way systems do: hire well, build products and services that work, cut down the friction whilst increasing the reward, trust the people doing the work, and study how the work actually flows.

Do that and service gets better – not as a process improvement campaign, not as an initiative to drive down costs, but as the natural by-product of building better work, and better places to do it.

Which is, in the end, where care becomes visible, and where the value accumulates. And if you do have whiteboards, don't list your most hated customers on it.