Every number in the Cost of Chaos Index 2026 is one of two things: a survey finding, or an estimate modelled from survey findings and public data. This post sets out which is which, and shows the working behind each estimate so anyone can check it or rerun it with their own assumptions.
The survey
MacroView commissioned independent consumer research agency Antenna to survey full-time Australian workers about email. Antenna surveyed 1,034 people in organisations with 100 or more employees, with fieldwork between 6 and 14 July 2026. [Antenna’s statement on weighting, representativeness and margin of error, in their wording.]
Survey findings are percentages or averages reported directly from the responses: 28 work emails received a day, 78% having experienced or witnessed a problem caused by a hard-to-find email or document, 2% with an automated system that files important emails for them, and so on. These are attributed to Antenna.
What the survey says happens next
78% of workers have experienced or witnessed a problem caused by an email or document that couldn’t be found. 28% have seen a follow-up missed as a result.
The confidence figures sit uneasily beside that. 73% are confident a colleague could find their important emails if they were away or had left. Only 13% are not confident. Those were separate questions, so the survey can’t say whether the confident people are the same ones who saw the problems. Taken across the workforce, though, people’s confidence in their own filing runs well ahead of what they report happening.
The post-leave findings show the same pattern under pressure. One in four workers say they deleted emails without reading them properly after returning from leave, and 17% archived emails unread. The email leaves the inbox without anyone deciding whether it mattered.
The five estimates behind the Cost of Chaos Index 2026
Each input is labelled by where it comes from.
- Survey finding · Antenna
- Public data · ABS
- Modelled · MacroView
- Assumption
Minutes per day
Antenna recorded time on email as ranges, so the survey gives the share of workers in each band. MacroView took the midpoint of each band and weighted it by that share, separately for each activity. The open top band has no midpoint and was set at around 150 minutes.
Both totals are MacroView estimates, not direct Antenna findings. They move with the value assumed for the top band. The 58 minutes is an estimated administration burden; the report does not claim every minute is avoidable or that it applies to all staff.
Hourly rate
The model values a worker's time using the ABS average weekly ordinary-time earnings for full-time adults, the most recent release when the model was built.
Newer data runs higher
The May 2026 ABS release, published in August, puts weekly earnings at $2,083.70. The model keeps the November 2025 figure it was built on.
On-costs are excluded
Ordinary-time earnings leave out superannuation, payroll tax, workers compensation and leave loading. The fully loaded cost of an hour is materially higher than $54.
Working days
Annual figures use 230 working days: the weekdays in a year, less public holidays and annual leave.
- Working days
- Public holidays (around 11)
- Annual leave (20)
Cost per employee
Avoidable minutes a day, multiplied by 90 cents and by 230 days. The report models three scenarios and gives the full 58 minutes as an upper reference point.
Want your own numbers? The Cost of Chaos calculator runs the same arithmetic on your headcount, minutes and hourly rate, with no sign-up.
Opportunity leakage
An illustrative scenario showing how missed follow-ups turn into lost gross profit. The inputs are assumptions chosen to show the mechanism. Move them to see how the result responds.
- Commercially relevant staff
Assumption - Share missing or delaying one follow-up a month
Assumption - Missed or delayed opportunities a yearStaff × share × 12 months
- Average gross profit per opportunity
Assumption - Share of the loss attributed to email chaos
Assumption - Potential annual leakage, report scenarioOpportunities × gross profit × attribution
What the survey found underneath
Share of workers who have witnessed each outcome.
What we deliberately didn’t claim
We did not multiply the per-employee figure across the Australian workforce to produce a national total. The survey covers organisations with 100 or more employees, and extrapolating beyond that population would overstate what the data supports.
We also did not combine the confidence finding (73% confident colleagues could find their emails) with the problem finding (78% have seen problems) to say the same people hold both views. They were separate questions.
