● Clutch 2026 Top Virtual Assistant Company · Top Medical Billing Company · Managed Virtual Medical Assistants

Bad Customer Service: 7 Causes and Fixes That Work

Published: Updated: By:

Key Points

  • Bad service rarely starts with a rude agent; at Telstra and Bell Canada, callers repeated one issue across three departments because front-line agents could not fix back-end errors.
  • Post-task surveys draw only a 4 to 7 percent response rate, according to a support-analytics vendor's white paper discussed on the Support Experience podcast in March 2026.
  • In HelpSquad's own February 2026 records, Slinger Bag saw response times improve by 90% and customer satisfaction rise 12% across more than 1,100 tickets in the first month.
Three things support managers believe. Myth or fact?
Call each one, then see how other readers called it.
1 Bad customer service is mostly caused by rude agents.
2 A remote agent can work inside a practice's EHR without storing patient data on a personal device.
3 For a complicated product, a good script is all a frontline agent needs.
A woman with long braids holds out a phone and gestures with frustration in a kitchen.

Quick Answer

Bad customer service usually comes from agents who cannot fix the problem in front of them, so customers get passed between departments and repeat their story. Fix access and escalation first.

A rude agent is rarely the cause. Handling billing disputes at Telstra and correcting sales errors at Bell Canada taught me that. Why? Because an agent who spots a back-end error but cannot correct it can only transfer the call. For healthcare clients, our agents work inside the practice's own systems, with a dedicated US-based account manager behind every account.

Did this answer your question?
A woman with long braids holds out a phone and gestures with frustration in a kitchen.

A frustrated caller gestures at her phone, the moment bad service begins when nobody can fix the problem.

Picture a patient calling about a bill that is wrong. The first agent is polite, but cannot open the billing record. The call goes to billing. Billing says the error came from scheduling. Scheduling asks the patient to explain it all again.

Nobody in that chain was rude. The patient still left.

That is the pattern I want to walk you through. Most advice on bad customer service starts with the agent: hire friendlier people, write a better script, send a survey. Here is a quicker test: pull a handful of recent complaints and count how many people each customer spoke to before anyone could change the record.

In this article you will find:

  • The seven causes of bad service, each paired with a fix you can start this month
  • What changes when agents work inside the practice's own EHR or practice management system rather than a separate tool
  • How remote access can be opened up without opening up patient data
  • Why satisfaction surveys miss the customers who quietly walk away
  • A short checklist for comparing outsourced support partners

I'll come back to that billing call throughout. Why? Because every fix here either shortens that chain or removes a link from it, and the first link worth examining is the agent who picked up the phone.

Bad customer service is any interaction that leaves a customer worse off than when they reached out: unresolved, repeated, or passed along. It matters because it stays invisible until the customer is gone.

Here is my main point, up front. Ask one question first: how many hand-offs does it take to fix a billing error? In my time handling billing disputes at Telstra and correcting sales errors at Bell Canada, the pattern I saw most was simpler and worse: a frustrated caller repeating the same issue across three different departments, because the person on the front line had no tool to fix a back end error. That is what drives churn.

The seven causes below all follow that thread:

  1. Agents who can see a problem but cannot fix it
  2. Escalation paths that bounce customers between departments
  3. Customers forced to repeat their story
  4. Scripts in place of real system knowledge
  5. Surveys that miss the customers who leave
  6. Broadcast channels standing in for direct contact
  7. Coverage gaps when nobody is there to answer

Each one has a fix. Most start in the same place: put the agent inside the same system the back office uses, whether that is Epic, athenahealth or Dentrix, with secure access to finish the job. Then repair the path a case takes when they can't.

Top 3 questions this article answers

One honest trade-off first: giving agents more access fixes more problems, but it also raises the security stakes. Both answers below take that seriously.

Which companies outsource healthcare customer service and support?

Small practices and multi-location groups alike outsource patient calls and chats. Our data shows about 149,000 phone calls and 267,000 chat and SMS conversations a month handled for healthcare practices.

Before you decide whether outsourcing will fix your service or make it worse, check four things:

  • Will the outside agents work inside your own systems, or in a separate tool?
  • Can they resolve the common request on first contact, or only take a message?
  • Who owns a case when it has to move to your back office?
  • Is there one named manager accountable when something breaks?

Who actually sends this work out? The volume tells the story. Averaged over the 12 months to February 2026, a small practice sent about 2,200 to 2,700 calls a month, while a multi-location primary care group sent about 6,500 to 7,000. Those are not call-center giants. They are ordinary offices where every unanswered call is a patient who may not try again.

Our coverage model reflects that mix. HelpSquad agents work from the US, Europe, Central America and South Africa, which gives bilingual coverage, and every healthcare client gets a dedicated US-based account manager with backup coverage included. It's worth noting why that last detail matters. A single accountable owner is what keeps a problem from bouncing between teams.

Conventional wisdom says bad service starts with a rude agent. My time handling billing disputes at Telstra and correcting sales errors at Bell Canada taught me otherwise. The research section below walks through these findings.

Why does the direct line matter so much? Because broadcast channels do not reach most customers. In 2018, a speaker from Social Media Examiner told The Agents of Change podcast that the company had 550,000 Facebook fans yet counted itself lucky to get 10,000 impressions per post, with a viral post reaching about 50,000. Many fans who had set the page to See First commented that a new announcement video was the first from the company they had seen in years.

It's worth noting what that means for a practice. A post about a billing change or new office hours will miss most of the patients it is meant for. Those patients still call. Some call after the bill has already gone out wrong, which is the hardest call of all to handle well.

So when you compare outsourcing companies, ask how they handle the direct channels: phone, chat and SMS. Who answers? What can that person change in your systems? A partner that only adds social posting or email campaigns leaves the front line exactly where it was, with the same caller and the same unanswered question. Ask to see an agent work a real case in your system before you sign.

Complexity raises the stakes. In 2020, a co-host of The Bad Crypto Show told The Liberty Advisor Show that opening a crypto wallet was still complicated, even on Coinbase, and that users still needed education on securing their crypto. Swap the wallet for an insurance benefit or a claim denial and the lesson holds. When the product is hard, customers need an agent who can explain and fix, not one who can only transfer.

Combining 3 sources points to one pattern: outsourcing works when the outside team is set up to finish the job. If you are still deciding which tasks a medical practice should outsource, start with the requests your front desk transfers most often. List them, then ask each provider which of those requests its agents could close without handing them back to your staff.

Tired of patients telling their story three times?

HelpSquad agents work inside your own EHR, under logged, MFA-protected sessions, with a dedicated US-based account manager and backup coverage behind them. Bilingual, 24/7 support starts at $8 per hour. Book an intro call, or phone 877-775-3667, and map where your escalations break.

Why don't satisfaction scores catch the customers who leave?

Most support teams judge themselves by a satisfaction score. Very few ask how many customers that score actually speaks for, or who stayed silent before walking away.

A March 2026 episode of the Support Experience podcast took on that question. The two speakers walked through a SupportLogic white paper that draws on a University of Groningen study, several Harvard Business Review analyses and industry data. The first number was the bluntest. Post-task surveys, one speaker said, achieve "an incredibly dismal 4 to 7 percent response rate." At 4 percent, a company handling 10,000 conversations ends up setting strategy on the opinions of 400 people.

Customer interactions 10,000
Survey responses at a 4% rate 400
How many customers does a score speak for? 10,000 interactions at the low end of the 4 to 7 percent post-task survey response rate cited on the Support Experience podcast, March 2026.

The short survey itself holds up well. According to the figures discussed on the episode, a single overall satisfaction question shows a 90% correlation with long multi-question surveys. The trouble is who answers. Customers who have given up rarely bother, the speakers argued: "Rather than leaving a bad score, they simply choose to say nothing and take their business elsewhere."

"An organization ends up measuring the happiness of the people who stayed, completely blind to the systemic issues that drove the others away."

Speaker on the Support Experience podcast, reviewing a SupportLogic white paper, 2026

So what pushes people out? The creators of the Customer Effort Score (CES) made a lopsided argument: service experiences earn a brand very little loyalty, but bad service drives disloyalty. The episode gave an example of how that sounds. When AI reads support transcripts for frustration, one of the signals it flags is a phrase like "I have explained this three times already." That sentence points to a broken process more than a rude person. The customer has been heard several times, and the problem still has not moved.

Effort has limits as a measure, though. When the Groningen researchers tested feedback metrics across industries for how well each one predicted whether customers stayed, "CES fared the absolute worst of all the metrics tested." The white paper's explanation is that "looking backward at friction is far too narrow of a lens." Net Promoter Score has its own weak spot, because it throws away every 8 out of 10.

If every customer rates youYour NPS is
8 out of 100
9 out of 10100

One speaker called relying on any single measure "statistically unwise" and said combining NPS and CSAT "dramatically improves how accurately a company can predict retention."

We hold our own figures to the same test. Slinger Bag, a Connexa Sports brand with more than 40,000 US customers, chose us for Level 1 support when its response and resolution times had become unacceptable. Across more than 1,100 tickets in the first month, response times improved by 90% and customer satisfaction rose 12%. The response time figure is the sturdier of the two, because the satisfaction gain comes from the same kind of survey questioned above. On the healthcare programs where we track it, patient satisfaction and NPS average about 9 out of 10. Those scores come only from the patients who chose to answer.

What we can control before launch is the handoff. During onboarding we write each practice's escalation paths into its playbook and test them in mock calls. A practice goes live only once it is satisfied with how those calls went. Most inquiries are resolved in the same interaction, and follow-up items are completed within 12 hours, depending on the clinic's workflows and provider availability.

Seen side by side, these findings describe one blind spot from two ends. By the CES creators' account, bad service is what drives customers away, and the frustration it causes shows up inside conversations, in sentences about explaining something three times. Meanwhile the score rests on the few hundred people who filled in a form. The complaint did not go missing. It stayed in the transcript, and you hear it best by reading several measures together.

  • Put your survey response rate next to every satisfaction score you report, so everyone can see how many customers the number speaks for.
  • Each month, read a sample of transcripts or call notes for customers saying they have already explained the problem, then trace each one back to the handoff where it started.
  • Report effort, satisfaction and NPS together, and track how many customers rate you 8 out of 10, since NPS throws those ratings away.
  • Ask any support partner, us included, how escalation paths are tested before launch and how quickly follow-up items close.
  • When a vendor reports a quick win, ask which figure came from operations and which came from a survey.

How we checked this

The research findings come from one podcast episode that reviews a SupportLogic white paper. We did not read the University of Groningen study or the Harvard Business Review analyses directly, and the transcript does not name the speakers. The white paper argues for AI review of every conversation, so it has a point of view. The Slinger Bag results, the satisfaction average, the onboarding steps and the follow-up times are our own figures from February 2026. Slinger Bag is a single consumer brand, not a healthcare practice. We sell outsourced support, so we have a stake in this answer. Still unknown: how much customer loss traces back to transfers specifically. None of our sources measured that.

  1. Support Experience podcast, episode transcript on feedback metrics, March 5, 2026.
  2. HelpSquad, internal data: Slinger Bag case study, onboarding process, speed to answer and patient satisfaction, February 2026.
A smiling support agent with curly hair wears a face mask and a headset.
A support agent who can fix the problem on the first call keeps callers from repeating their story.

Which healthcare BPO companies are best for call center services?

The best healthcare call center partners work inside your own EHR or practice management system. Ours do, so they fix the record a patient calls about instead of transferring.

That sounds like a technical detail. It isn't. When you compare providers, ask which of your systems their agents will log into on day one, and what they will be allowed to change there.

I have watched this play out from the inside. Handling billing disputes at Telstra, I saw frustrated callers repeat the same issue across three different departments, because the person who answered had no way to correct a back end error. Nobody was rude. Everyone was stuck. The customer left anyway.

What changes when the agent has access?

SituationAgent without system accessAgent working inside your system
Patient disputes a balanceTakes a message for the billing teamOpens the account and reviews the charge while the patient is still on the line
Wrong appointment on fileTransfers the call to the front deskCorrects the schedule and confirms the change
Insurance questionReads a generic scriptChecks what is recorded and explains it
Result for the customerRepeats the story to the next personExplains it once

So which systems matter? Across our healthcare clients, as of February 2026, agents work in Epic, Cerner (Oracle Health), athenahealth, eClinicalWorks, NextGen Healthcare, Dentrix, Open Dental, Kareo (Tebra), DrChrono and Practice Fusion. When you compare call center vendors, ask which of those their agents already use, and whether the agents work in your system or in a separate tool of their own.

Here is the tension. Access is what lets an agent fix a problem, and access is also what worries a compliance officer. Both concerns are valid. Our review of how remote access is set up shows it can be controlled tightly: agents work inside an isolated Microsoft virtual desktop, each has a unique login with multi-factor authentication on every session, data is encrypted at rest and in transit, copying, drive mapping, USB devices and printing are disabled, and sessions are logged and monitored. Patient information stays in the practice's own systems. The Business Associate Agreement is signed with HelpSquad as a company, never with individual agents.

Access alone is not enough, though. A 2024 episode of The Coffee Experts Club Podcast offers a useful contrast. A former Dutch Bros barista described taking apart the entire espresso machine during a graveyard shift to learn how each part connected. That is the difference between an agent who has a login and an agent who understands what each screen controls.

In practice, ask a vendor two questions. Will your agents work in our system? And will they be trained on how its parts connect, so a billing fix does not quietly break a scheduling record? The second answer tells you whether escalations will shrink, or simply move to a new desk.

What should small practices look for in medical billing services?

Look for a billing partner that resolves balance questions on the first contact and reviews every patient conversation, not only the few surveys patients bother to return.

Billing is where bad service hurts most, because money makes people anxious fast. Before you sign with any billing or support partner, confirm these four things:

  1. Agents can look up and explain a balance while the patient is still on the line.
  2. Every escalated case has one named owner until it is closed.
  3. The partner reviews conversations, not just survey scores.
  4. Coverage matches the hours your patients actually call and message.

Why put the second item so high? Because the escalation is where trust breaks. When I corrected sales errors at Bell Canada, the correction itself was rarely the hard part. The hard part was getting the case to someone with the access to make it, without the customer telling the story again. A clean handoff = KEY.

Why can't surveys tell you service is broken?

Most practices judge their service by satisfaction surveys. That is the blind spot. One support-analytics vendor's white paper claims post-task surveys draw only a 4 to 7 percent response rate. Using the vendor's own example, a company with 10,000 interactions and a 4% response rate ends up basing its strategy on the opinions of just 400 people.

Who is missing from that 400? Usually the patient who was bounced between departments. Ask whether a billing partner tracks patients who stop calling after a disputed charge. The same white paper cites a university study that tested several metrics for how well they predicted retention, and it found that Customer Effort Score performed worst of all the metrics tested. For a billing partner, also ask how quickly a disputed charge reaches the person who can correct it.

My own takeaway is simpler. Ask each billing service which number it reports to you every month, and why. Then ask to see last month's report before you sign.

How does this apply when you compare outsourcing partners?

The same test works whether you are shortlisting 24/7 live chat outsourcing for a small business with a budget near $10 an hour, or comparing the best virtual medical assistant companies. Ask each provider how they spot an unhappy customer who never fills in a survey. Ask who reads the transcripts. Ask what happens to a conversation that ends with a frustrated patient and no resolution.

If you are weighing managed virtual medical assistant services against a billing-only vendor, put those questions to both. The research section above walks through these findings.

Volume counts can mislead in the same way. In 2018, a speaker from Social Media Examiner told The Agents of Change podcast that Facebook delivered 4 to 6 times the views of YouTube for the company's long-form video, but retention on Facebook was far worse. In that 2018 account, YouTube retention had dropped to about 60% in the first seconds and then held at 50 to 60% to near the end. Raw views said go all in on Facebook. The retention data said the opposite, and the company pulled its long-form Facebook shows.

A billing service report works the same way. Calls answered is the view count. Balances resolved without a callback is the retention number. Why? Because a patient who hangs up still confused will call again or leave the balance unpaid, and both show up later as a collections problem rather than a service one.

Before you sign, ask for three things in writing: the share of billing questions closed on the first contact, how errors are escalated to the person who can fix them, and who owns the case until it is closed.

What is the one fix that matters most for bad customer service?

Give frontline agents the access and authority to resolve an issue on first contact, then build an escalation path where one person owns each case until it closes.

I would put that fix ahead of any new script or survey. When the agent can reach the record, the customer explains once. When the agent can't, no amount of courtesy saves the call.

Tone still counts, though. Speaking on The Coffee Experts Club Podcast in 2024, a former Dutch Bros barista said the chain taught them how small mannerisms set the tone for both the stand and the customer. Warmth opens the door. Access is what lets the agent walk through it.

Start this week by mapping one recent complaint from first call to resolution, and mark every point where the customer had to explain it again. Then pull your last few transferred cases, and at every handoff ask one question. Who could have closed this sooner, and what were they missing?

Frequently Asked Questions

What else do people ask about bad customer service?

What is the most common cause of bad customer service?

In my experience, it is an agent who can see the problem but cannot fix it. The agent lacks access to the system where the error lives, so the customer gets transferred. Rudeness is rarer than people assume.

Why do customers leave after being transferred?

Every transfer asks the customer to tell the story again. By the third department, the frustration is about the process, not the original issue. Many leave without complaining at all.

How do you fix a broken escalation process?

An escalation pipeline is the path a case takes from the front line to the team that can resolve it. Fix it by giving each escalated case one owner until it closes, and by passing full notes at every handoff. The customer should never restart from zero.

Can outsourced healthcare support be HIPAA compliant?

Yes. HelpSquad signs the Business Associate Agreement with the practice as a company, never through individual agents. Agents use unique logins, multi-factor authentication on every session, and encrypted, monitored sessions.

How much does outsourced customer support cost?

HelpSquad charges $8 to $13 per hour for full-time staffing. That is a fully managed service with HIPAA compliance, bilingual agents, transparent pricing and dedicated account management.

Do outsourced agents support more than one language?

Yes. HelpSquad has bilingual agents across the US, Europe, Central America and South Africa. That spread also helps cover hours a single local team cannot.

How do I get started?

Book an intro call at helpsquad.com/contact or phone 877-775-3667. Bring a list of the requests your team transfers most often; that is where the first conversation should start.

Written by

Maria Rush

Marketing Team Lead, HelpSquad

Maria De Jesus-Rush is Marketing Team Lead at HelpSquad, a healthcare business process outsourcing company, with a background in content development, digital marketing, and project management.

Connect on LinkedIn
Tags
  • customer-service
  • customer-experience
  • team-management
  • small-business
Let's talk

Let's talk about what your practice actually needs.

A 30-minute call. No sales pressure. We'll tell you honestly whether we're a fit.

877-775-3667 · info@helpsquad.com · Doylestown, PA