an apology needs an action plan

Two companies apologized this month after failures that reached people in very different ways. Telstra's July 8 national mobile outage disrupted calls and data, including access to Triple Zero; a later Senate hearing counted 604 failed emergency calls.¹ Capita's work on the UK Civil Service Pension Scheme left retirees and bereaved families waiting after the company missed two recovery dates.²

As I try to teach my six-year-old, an apology needs an action plan that is quickly proven. I use the same standard when a company asks customers to trust it again: tell me what broke, name who owns the work, put a date on it and let the people who carried the failure see the result.

The research gives that home rule more structure. Nicole Gillespie and Graham Dietz describe organizational trust repair in four stages: an immediate response, a diagnosis of the causes, reforms across the system and an evaluation of whether those reforms worked. Their model reaches across leadership, process, culture and outside governance, which is why an apology belongs near the beginning of the work rather than at the end.³

Telstra had to respond at the speed of an emergency network. Capita has to work through months of pension cases, including people entering retirement without a quote and families trying to settle a benefit after someone has died. I would ask each company for a different plan and score both on the proof they provide.

Telstra put facts under the apology

Telstra CEO Vicki Brady apologized on July 13, said the company understood the cause and promised a full investigation. Four days later, her Senate evidence gave customers enough detail to begin checking the company's work.⁴

Maintenance had restarted a 15-year-old network time server in Melbourne, where an underlying configuration sent its clock back to 2006. The team had no documentation of a previous design change, and a manufacturer update flagged in 2022 and again in January 2026 had never been applied. Brady put the replacement cost at about $30,000 and accepted that the outage might have been avoided.⁴

At the peak, about 45 percent of calls and data sessions on Telstra's mobile network were affected, including 604 failed calls to Triple Zero. Telstra said it completed a welfare check for every call and knew of no life-threatening outcome, while service teams applied a separate emergency-call fix the next morning.⁴

The $30,000 replacement cost is hard to ignore alongside the missing maintenance note and updates left undone. Customers can ask whether similar equipment was checked; regulators can ask who approved the maintenance rules; and Telstra employees can ask whether the system gave their team the information needed to do the job safely.

Our trust model keeps those relationships separate because each person is being asked to rely on something different. The customer needs a dependable network, the Triple Zero caller needs Telstra to understand a public-safety duty, the regulator needs a clear account of control and disclosure, and the maintenance team needs a system that records changes and closes known risks.

Research on trust violations also separates competence from integrity because people read the response differently depending on what they believe failed. A broken server begins as a capability problem; the unapplied update and missing documentation invite a management question about known risks. I want the technical fix and the management controls because this case can be read through both doors.⁵

Brady appeared before the Senate, answered direct questions and commissioned an external expert to test Telstra's findings. I would keep the case open until that work is public, including the controls changed across related equipment and a clear response to people or businesses that can document a loss.

Capita was already behind its own dates

Capita took over administration of the Civil Service Pension Scheme in December 2025 after a two-year transition. The handoff included 86,000 cases in progress and 15,000 unread emails, yet Capita assured the government it was ready before missing both its April recovery milestone and the end-of-June date for restoring standard service.⁶

By the government's count, the unresolved backlog reached 120,000 cases and more than 6,700 people were waiting for retirement quotations after their planned retirement date. The July 13 update also listed hundreds of death-in-service and ill-health retirement cases, along with thousands of bereavement cases still waiting on Capita or missing information.⁷

By June 19, call waiting time had fallen to four minutes and older cases were receiving individual attention. The government added more than 140 people to the recovery effort and offered interest-free bridging loans to members in hardship, which tells me the operation had moved well beyond the service Capita promised to run.⁸

Complaint-handling research gives me three parts of the recovery to watch: what people receive, how the decision is reached and how they are treated during the process. The bridging loan addresses an immediate outcome, the case workflow determines whether the process is fair, and a person waiting after a death will judge every contact with the administrator. Progress in one part cannot cover a continuing failure in another.⁹

Capita apologized again in its July 9 trading update and said revised processes, technology and automation would help clear the backlog. The company put the 2026 profit reduction at £25 million to £40 million and the cash-flow hit at £35 million to £50 million; the government had already withheld £9.9 million, sent independent auditors into the scheme and begun appointing a remedial adviser at Capita's expense.¹⁰

Those financial charges give investors something they can price, while the pension remains the unfinished obligation to members and families. Parliament is due another update in October, and I would judge it from three seats: whether members are paid on time, whether the government still needs a rescue team, and whether investors can believe Capita's promises about the cost and credibility of delivery.

the market put a price on each failure

Telstra's shares closed down 2.96 percent on July 8 while the broader ASX 200 fell 0.21 percent. Trading volume rose to roughly 43 million shares from 25 million the day before; the stock gained 1.22 percent the next day and closed July 20 at A$5.07, back where it had closed before the outage.¹¹

I use that market move for a narrow question: did investors believe the outage would create lasting damage to earnings, claims or the business? Their answer appears to have been no within a few trading days. Customers and regulators are answering different questions, and the share price recovered before the full investigation was public.

Capita's shares closed down 13.54 percent on July 7 after the government disclosed the withheld payment and criticized the missed commitments. The stock fell another 20.65 percent on July 9 after Capita put a number on the profit and cash impact, leaving it 32 percent below its July 6 close; it remained about 27.5 percent below that base on July 17.¹²

Capita also reported £998 million in first-half contract value, its strongest first-half Public Service result since 2021, and adjusted revenue growth of 1.6 percent. Investors still attached the pension failure to cash flow, contract risk and a delayed path to positive free cash flow in 2027.¹³

Stock charts are terrible empathy maps. I read Telstra's recovery and Capita's larger decline as investor judgments about financial exposure, with no claim that either line tells us whether a failed emergency caller felt safe or a bereaved family received an answer.

I would score the two plans separately

Our research has been pushing us to make the score more specific. PwC reaches a similar conclusion by separating operational, accountability and digital trust, with each category supported by data, processes and controls. Telstra's time-server failure sits in the operational and digital categories before moving into accountability when the company explains the controls; Capita's backlog begins as an operational failure, while missed dates and assurances make it an accountability problem too.¹⁴

For Telstra, I want a precise incident timeline, the independent technical finding, a list of controls changed across related infrastructure and a clear account of compensation. Speed belongs in the score because emergency calls were involved.

For Capita, I want the age and type of every unresolved case, the share of new work completed within the service standard and the time it takes hardship support to reach someone without income. Put the oldest case beside the median so a better average cannot hide the person who has waited longest.

I also care about who is visible. Telstra's chief executive was overseas on leave during the outage, and the CFO handled public updates while operating teams restored service. During the incident I want the operator who can give useful facts; in the days after, I want the chief executive to own the obligation and submit the work to outside examination.¹⁵

Capita's leadership had enough time to make assurances and set dates, so each overdue case now measures those words. The independent audit and remedial adviser give someone outside Capita a hand in deciding when recovery is real.

the plan has to keep proving itself

For the next update, I would begin with the people still carrying the failure: unresolved emergency-call claims, pension cases past their promised date and the age of the oldest pension case. A smoother average should never erase that person.

Telstra should name the equipment reviewed, maintenance rules changed, undocumented configurations found and conclusion of the outside expert. Capita should show backlog reduction by case type, new work completed within the contract standard and findings from the independent auditors.

I want compensation and hardship support reported on their own because people will use those outcomes to judge the recovery. Telstra says it has contacted 8.8 million customers, with compensation tied to the effect and the claims people make; Capita's line should cover bridging loans, arrears, withheld payments and recovery costs borne by the company.¹⁶

Every promise also needs an owner and a date. Telstra owes the completed investigation and the changes that follow it, while Capita and the government owe Parliament the October update; any moved date should come with a public explanation.

On the next company update, I would skip the opening apology and go straight to the action plan. Telstra has begun putting its work under public examination, and Capita will be judged one overdue case at a time.

other reads on these failures

• ACCAN left the Telstra Senate hearing unconvinced. The consumer group wants enforceable reliability standards and a better answer on how another national outage will be prevented.
• The ABC's pre-hearing report is worth reading beside the Senate evidence. It covers consumer research that raised concerns about Triple Zero access and the regulator's challenge to how that research was conducted.
• The PCS union wants the pension work returned to the Civil Service. Members missing pensions are the center of its argument, because the financial penalties have not paid them.
• The Guardian spoke with affected pension members who were trying to cover rent and food after a bereavement, in some cases after months of asking for answers.

source notes

1. ABC News, July 17, 2026, reporting Telstra's Senate evidence on the server failure and 604 failed Triple Zero calls.
2. House of Commons Library and The Guardian, July 20, 2026, on the missed recovery dates and affected pension members.
3. Nicole Gillespie and Graham Dietz, “Trust Repair After an Organization-Level Failure,” Academy of Management Review 34(1), 2009. The authors propose a systemic four-stage repair process: immediate response, diagnosis, reforming interventions and evaluation.
4. ABC News, July 17, 2026, and Telstra's CEO statement, for the server age, configuration, updates, replacement cost, network effect and welfare checks.
5. Peter H. Kim, Donald L. Ferrin, Cecily D. Cooper and Kurt T. Dirks, “Removing the Shadow of Suspicion,” Journal of Applied Psychology 89(1), 2004. Their experiments found that trust-repair responses were evaluated differently for perceived competence and integrity violations.
6. House of Commons Library, on the transition, inherited work and missed milestones.
7. UK government recovery-plan update and House of Commons Library, for the backlog and case-type counts.
8. UK government recovery-plan update, for call waiting, added staff and bridging loans.
9. Stephen S. Tax, Stephen W. Brown and Murali Chandrashekaran, “Customer Evaluations of Service Complaint Experiences,” Journal of Marketing 62(2), 1998. The study evaluates complaint recovery through the outcome, procedure and interpersonal treatment.
10. Capita's July 9 trading update and House of Commons Library, for the profit, cash, withheld-payment and oversight figures.
11. ABC News market coverage and StockAnalysis daily history, for Telstra and ASX 200 closes and Telstra volume.
12. Reuters via MarketScreener, Reuters via Euronext and Investing.com UK daily history, for Capita's market moves.
13. Capita's July 9 trading update, for contract value, revenue growth and free-cash-flow guidance.
14. PwC, “The new architecture of trust,” July 15, 2025, on operational, accountability and digital trust and their supporting data, processes and controls.
15. ABC News, July 8, 2026, on Telstra's leadership and government-notification timeline.
16. Telstra's CEO statement, on customer contact and individualized compensation.

Previous: what owning the miss looks like