When Customer Service Goes Public – What Has Gone Wrong?

Another day, another LinkedIn post calling out a company for poor service.

The latest one I saw concerned Etihad. Over the past few years I seem to have seen increasingly more people using LinkedIn to publicly complain about airlines, banks, telecoms companies, delivery companies and just about every other major service business.

One name in the UK appears particularly regularly … Evri.

Coming from a logistics and service background, these posts always make me think the same thing.

What has gone wrong?

Not necessarily with the original service. Things go wrong. Parcels get delayed. Flights get disrupted. Systems fail. People make mistakes.

Nobody running a substantial service business can realistically promise that nothing will ever go wrong.

What they should be able to promise is that when it does go wrong, they respond well.

And perhaps that is where the bigger problem now lies.

The second failure

For me, one of the great strengths of a good service company has always been its ability to recover.

The customer experiences a problem. Somebody takes ownership. Communication is clear. The problem is resolved. Perhaps an apology is given. The customer feels somebody actually cared.

Sometimes a well-handled problem can even strengthen the relationship.

There is actually a recognised concept in customer-service research called the service recovery paradox: under the right circumstances, customers whose problem has been handled exceptionally well can become more loyal than customers who never experienced the problem in the first place. Recent research continues to find evidence that effective recovery can rebuild satisfaction, engagement and loyalty.

That makes complete sense to me.

The real test of a service company isn't necessarily what happens when everything works. It's what happens when it doesn't.

Unfortunately, this seems increasingly to be where companies are failing. Evri is an interesting example

Ofcom's research into UK parcel deliveries makes interesting reading.

In its 2025 survey, 68% of parcel recipients said they had experienced some sort of delivery issue during the previous six months. Yet overall satisfaction with parcel companies remained reasonably healthy at 78%.

Then look at complaint handling.

Across the companies surveyed, only 46% of customers were satisfied with how their contact or complaint was handled. Evri was considerably lower at around 30%, with dissatisfaction approaching 40%.

That distinction is important.

Customers can accept that occasionally a parcel goes missing.

What they find much harder to accept is not being able to find somebody who will sort it out.

And Evri is certainly not alone. Recent Ofcom figures have also shown significant complaint levels within telecoms, while UK financial-services companies reported 1.74 million complaints during the second half of 2025.

So what has changed?

Have we automated customer service too far?

Automation has transformed service businesses.

Mostly for the better.

We can track parcels, change bookings, check bank accounts and obtain information without speaking to anyone. AI will take this considerably further. Salesforce research suggests service teams expect half of customer-service cases to be handled by AI by 2027.

I see nothing wrong with that.

Use technology to deal with the routine.

The danger comes when companies also try to make automation deal with the non-routine.

The parcel that isn't where the tracking system says it is.

The booking that doesn't fit one of the chatbot options.

The refund that has become stuck.

The customer whose situation simply doesn't fit the script.

At that point automation can quickly change from convenience to obstruction.

We have all probably experienced the circular chatbot, the automated email answering a question we didn't ask, or the telephone system determined to stop us reaching an actual human being.

Technology isn't the problem. Using technology principally to prevent expensive human interaction probably is.

PwC's 2025 customer-experience research provides an interesting warning. Almost three in ten consumers said they had stopped using a brand because of poor customer experience, while 58% said they were only somewhat comfortable or not comfortable using AI to engage with companies.

Efficiency is valuable.

But efficiency that loses customers isn't particularly efficient.

Or is this simply about cost?

That leads to a more uncomfortable question.

Have some large organisations calculated that reducing customer-service costs is worth losing a percentage of customers?

Large contact centres are expensive. Experienced customer-service people are expensive. Giving employees authority to resolve problems can be expensive.

Automation is cheaper.

If you serve millions of customers, perhaps losing a few thousand frustrated ones appears manageable on a spreadsheet.

Margins are protected. Productivity improves. Shareholders are happy.

Possibly. But there is one thing that calculation may underestimate.

Those customers now have an audience. The complaint is no longer private

Twenty years ago, an unhappy customer might have complained to the company, told their family and perhaps mentioned the experience to a few colleagues. Today they can tell 5,000 people before breakfast.

LinkedIn has become particularly interesting in this respect because complaints appear alongside professional reputations. Someone with thousands of connections posts about an airline, courier or bank and suddenly the complaint is being read by customers, employees, suppliers and senior managers.

Research into social-media complaints suggests that going public is often not the customer's first action at all. It can be a secondary stage after other attempts to resolve the problem have failed.

That changes how I look at many of these LinkedIn posts.

Perhaps the post isn't actually the complaint.

Perhaps the post is evidence that the complaint process already failed.

And once it becomes public, the company's response matters to far more people than the original customer. Recent research found that the first public response to a social-media complaint can influence how observers themselves perceive the brand.

The company is no longer just recovering one customer. It is performing its customer service in front of an audience.

Has “customer first” changed? This is the question I keep coming back to.

For decades companies talked endlessly about putting the customer first. Perhaps many still genuinely do.

The latest UK Customer Satisfaction Index actually rose to 78.3 in July 2026, so it would be unfair to suggest customer service everywhere is collapsing.

But I do wonder whether something more subtle has happened.

In the pursuit of scale, automation, productivity, cost reduction and margin, have some organisations optimised the easy 95% of customer interactions while making the difficult 5% increasingly painful?

If so, they may be saving money in exactly the wrong place. Because those 5% are the occasions when customers discover what the company is really like.

Good service isn't about never making a mistake. It never was. It is about what you do next.

And if your customer finally has to turn to LinkedIn simply to get somebody's attention, I would suggest the original service failure is no longer your biggest problem.

Your service recovery has failed too.