In 2015, ABC News published a fascinating piece where reporter Will Ockenden requested access to his own metadata. Australia’s new data retention laws had just come into effect, requiring telcos to store information such as call records, text logs and cell tower connections for two years.
Will received a dataset of his own activity, including phone calls, SMS records and mobile internet sessions. While the content of his calls and messages was not included, the dataset still contained thousands of data points revealing where and when his phone connected to nearby towers.
Curious about how much an outsider could learn from this, I recreated parts of the ABC team’s original analysis in Power BI. Using bubble maps, bar charts and filters for time of day, day of week and date range, I explored what assumptions could be made about Will’s daily life.
Where Are Will’s Active Spots?
By visualising the count of pings at each cell tower, it quickly became clear which locations Will frequented most. The towers in Balgowlah, Chippendale and Haymarket dominated his dataset.
Where Does Will Live?
Filtering to the hours between 23:00 and 05:00 (the time most people would usually be at home) reveals a strong concentration of pings at Balgowlah Hayes St. From this, it is reasonable to assume that this is Will’s residence.
Where Does Will Work?
Focusing on weekdays between 09:00 and 17:00 highlights two hotspots: Chippendale and Haymarket. As these locations are close together, it is possible that Chippendale represents Will’s workplace while Haymarket could be a regular lunch spot.
Where Does His Family Live?
Around Christmas 2014, all of Will’s phone activity was located in Hobart, Tasmania. A ping at Hobart Airport on 26 December suggests he travelled there to visit family before returning to Sydney.
Additional Insights
Beyond these four questions, the dataset lends itself to deeper exploration:
- Most contacted numbers: Bar charts of his top contacts hint at parents, close friends or work colleagues.
- Activity over time: Trends suggest changes in his routine, with a strong possibility that he moved from Redfern to Balgowlah in September 2014.
- Marginal histograms: Combining time-of-day activity with locations provides further validation of assumptions about where he lives and works.
What Does This Show Us?
Working through this exercise made it clear how much sensitive information can be inferred from metadata alone. Even without the content of calls, texts or internet activity, patterns emerge that allow outsiders to make surprisingly accurate assumptions about where someone lives, works, travels and who they communicate with most.
It is important to stress that metadata provides indications rather than definitive truths. However, the sheer scale of these records — tens of thousands of entries over a year — means the patterns are strong enough to build a detailed picture of a person’s private life.