Customers on Hold While Staff Search Five Systems: The Cost of Scattered Answers

The customer on hold while someone searches

A customer phones a Durban distribution business to ask where her order is. The agent says, "Let me just check," and the customer hears typing, a muted line and a short silence. The agent opens the order system, finds a reference, switches to the warehouse spreadsheet, then to the courier website, and finally to a shared inbox to see whether anyone has emailed about a delay. Four minutes later she has an answer, and the agent has a headache.

Nothing in this story is a failure of effort. The agent is competent and the systems all work. The problem is that the answer to a single customer question is spread across several places, and a human being has to be the integration layer between them. Every question costs customer patience and paid staff time, and the cost repeats dozens of times a day.

This differs from the cost described in employees wasting time because information is stored in the wrong places, which looks at internal searching generally. Here the focus is the moment a customer is waiting on the line, in the chat window or at the counter.

Why fragmented information persists

Fragmentation is rarely designed. It accumulates. Sales chose a CRM, finance chose an accounting package, the warehouse built a spreadsheet, a support person started a shared mailbox, and a manager set up a WhatsApp group for urgent items. Each tool solved a real problem for the team that chose it. Together they mean that a complete picture of one customer exists nowhere.

Tools that search across sources can help, but they have limits. Microsoft's overview of Microsoft Search explains that users only see results they already have access to, and that search works across content in Microsoft 365 apps. That is useful, but it does not help when the answer sits in an accounting package, a courier portal or someone's personal spreadsheet. Search finds documents; it does not reconcile conflicting records.

Why businesses tolerate it

• Each system is "fine" on its own. No single tool is failing, so nobody owns the combined problem.
• Experienced staff compensate. Long-serving employees know where to look, so the cost is invisible until they leave.
• Customers rarely complain about a short wait. They simply form an impression and call less willingly next time.
• Integration sounds expensive. Owners assume connecting systems is a large project, so they accept the manual workaround.
• Handle time is not measured. Without data on how long answers take, the cost never reaches a budget discussion.

Operational symptoms

• staff saying "let me call you back" for questions that should take seconds;
• agents with six browser tabs and two spreadsheets open at all times;
• different staff giving different answers to the same question;
• customers asked to repeat information they gave last week;
• new staff taking months to learn where things are kept;
• escalations to a manager only because the manager knows which system to check.

Related patterns are covered in too many applications for one task and in departments keeping separate customer records. This article looks at the combined effect on service speed and on payroll.

How the pain becomes money

• Payroll cost of searching: every minute an employee spends looking for an answer is a minute of salary not spent on the work the role exists to do.
• Service cost: longer calls need more agents to handle the same volume, or leave customers waiting in the queue.
• Repeat contact: a customer who receives an incomplete answer calls again, doubling the cost of that enquiry.
• Error cost: answers assembled from partial records are more likely to be wrong, leading to credits, re-deliveries and lost goodwill.
• Revenue risk: slow, inconsistent service leads customers to try a competitor.

An illustrative cost model

This example is illustrative. Every figure is an assumption for a hypothetical business. Picture a service team of eight people at a regional distributor.

• Searching time: 60 enquiries × 3 minutes = 180 minutes (3 hours) a day × R180 × 230 = R124,200
• Repeat contacts: 10 × 6 minutes = 1 hour a day × R180 × 230 = R41,400
• Illustrative total: about R165,000 a year in payroll time

That figure covers only direct staff time. It excludes lost customers, credits issued because of wrong answers, and the manager time spent on escalations. For a small team it is the equivalent of roughly half a salary spent navigating between screens.

Secondary consequences

The human cost is real. Staff who spend their day hunting for information feel unproductive and stressed, and turnover in service roles is expensive. Customers who are placed on hold, or told they will be called back, lose confidence in the business. Where personal information is copied between spreadsheets and inboxes to answer questions, privacy risk also grows, because the more places customer data lives, the harder it is to protect and correct. South Africa's Protection of Personal Information Act places duties on organisations through a designated Information Officer, as the Information Regulator's POPIA page explains, and scattered customer data makes that accountability harder to meet.

When it becomes commercially urgent

• call or chat volumes are rising and the team cannot keep up;
• customer complaints mention waiting, repeating themselves or inconsistent answers;
• a long-serving employee who "knows where everything is" announces they are leaving;
• the business is about to add a channel such as web chat or WhatsApp support;
• load-shedding or connectivity problems force staff to work from fewer screens or devices.

What better looks like

The goal is that a person answering a customer sees the answer, not a search problem.

1. List the top questions. Most service teams handle a small number of question types: order status, invoice, delivery, price, account balance. Start with those.
2. Trace where each answer lives. For each question, write down every system and person needed to answer it.
3. Pick a single view. This might be a customer record in a CRM that pulls in order status, balance and delivery information from other systems.
4. Connect the sources that matter. Workflow tools such as those described in Microsoft's cloud flows overview can move data between services automatically, so staff stop copying it by hand.
5. Report on the result. Dashboards using a platform like Power BI, which is designed to connect and share data across an organisation, can show handle time and repeat contacts.
6. Keep a single source of truth. Decide which system owns each piece of data, and correct it there.

The Insights article the hidden cost of apps that do not share data explains why this is an operations question before it is a software question.

Practical next actions

1. Sit with an agent for an hour and note every system they open to answer ordinary questions.
2. Ask staff to log, for one week, enquiries that required more than one system and how long each took.
3. Rank the five most common questions by total time lost.
4. For the top question, design the single screen that would answer it.
5. Connect only the two systems that feed that screen before touching anything else.
6. Measure handle time again after a month and decide the next question to fix.

Bottom line

When customers wait because staff must search several systems, the business pays twice: once in salary for time spent searching, and again in goodwill from customers who notice the delay. The fix is rarely a large replacement project. It is usually a matter of deciding what customers most often ask, finding where the answers live, and bringing those few pieces of information into one place that every employee can see.

Sources and further reading

• Microsoft Search Overview (last verified 10 October 2026)
• Overview of cloud flows - Power Automate (last verified 10 October 2026)
• What is Power BI? - Power BI (last verified 10 October 2026)
• https://inforegulator.org.za/popia/ (last verified 10 October 2026)

Related Mellorca reading

• The Cost of Retyping Customer Details into Every System
• Thirty People, Five-Person Processes: The Cost of Systems That Did Not Grow
• Walking the Paperwork Around: The Cost of Moving People to Move Information