COVIDSafe downloads graph

Updates

4 May 2020: first list of signatory names released.
5 May 2020: See the coverage at ABC news and Sydney Ideas.
5 May 2020: We have now released the first hundred names.
6 May 2020: The uptake revived. We passed Singapore.
7 May 2020: We updated our analysis on the 40% target, based on an SMH article citing the PM who defined 40% in terms of number of individuals.
8 May 2020: restrictions are lifted by the PM, and uptake slows down.
15 May 2020: Downloads passed the 30% mark.
20 May 2020: Downloads increase but we do not know about net users.
21 May 2020: COVIDSafe has been used in VIC and NSW to trace the spread.

Graph tracking the COVIDSafe downloads

Up until 5th of May, we were not told how to translate the 40% uptake target in terms of number of users, nor how long we should expect to meet this important target. We did our own analysis below on these estimates. It is also important to note we were never which sort of model determined that 40% was enough.

Population of interest

The government has set a 40% target for taking up the COVIDSafe app. It is only meaningful to talk about such an uptake target relative to the population with technology that can access and use this app. Barring certain technical barriers that have emerged in different devices and operating systems, to keep this simple we assume that the population of interest are individuals with smartphones.

The Minister for Health, Greg Hunt agrees: “Obviously, we’ve been looking at the adult population and mobile phone users and in particular smartphone users, they’re the group that can download the phone.“

PM’s estimate

As at 5 May, a Sydney Morning Herald article cites the PM “aiming for 40 per cent of the country’s 16 million smartphone users“. We are not told the source of this estimate of 16 million users, but we suspect that this estimate might represent adult smartphone users of 18 years and above, drawn from a Deloitte survey as discussed below.

On the graph, we show the target of 6.4 million users (i.e. 0.4 × 16) as “PM’s estimate of 40% adult smartphone users”.

ABS estimate

The ABS does not provide estimates for individual use of smartphones, but in a 2016-2017 survey it estimated that 86.1% of Australian households had internet access, and of these households 91% had a smartphone. Therefore, about 78.35% of households had a smartphone at that time (i.e. 0.861 × 0.910).

The ABS projects that in 2020 there are 9,882,413 households and its population clock as at 3 May 2020 was ticking at about 25,677,050 Australians. This gives on average 2.6 persons per household (i.e. 25,677,050 / 9,882,413).

This means that there are 7,742,871 households with a smartphone (i.e. 0.7835 × 9,882,413), hence there are 20,131,465 individuals living in those households with a smartphone (i.e. 2.6 × 7,742,871). The 40% target translates to 8,052,586 individuals living in a households with a smartphone (i.e. 0.4 × 20,131,465).

We show this target as “ABS’s estimate of 40% individuals in households with smartphone”.

We understand that some households would have more smartphones than others, and that some households are larger in size than others. But we believe that there is merit in making household calculations because infectious diseases spread within household, and typically it is the entire household that must self-isolate to stop the spread.

Deloitte estimate

A 2019 Deloitte Mobile Consumer Survey reports 91% smartphone penetration in the Australian market using a sample of adults aged from 18 to 75 years old. According to the ABS, in 2018 there were 71.57% of 18 to 75 years olds. If we assume that this proportion remains about the same today, then it translates to 16,723,129 smartphone users between 18 and 75 years old (i.e. 25,677,050 × 0.7157 × 0.91). This looks close to PM’s interpretation of the population of smartphone users, with 40% being 6,689,252 users.

Deloitte | Access estimate

A Deloitte | Access Economics Mobile Nation 2019 report estimates that “89% of Australians own a smartphone”, using a sample from individuals aged 14 and over. According to the ABS, in 2018 there were 82.34% of 14 year olds and over. If we assume that this proportion remains about the same today, then it translates to 18,816,810 smartphone users of 14 years old and over (i.e. 25,677,050 × 0.8234 × 0.89). The 40% of this population is 7,526,724 users.

We report this estimate as “Deloitte | Access estimate of 40% smartphone users aged 14 and over”.

We find this estimate to be more representative that the one chosen by the PM, because it includes the most vulnerable population of 75+ year olds, as well as many young and active individuals aged from 14 to 18 years old.

Target timeline

The timeline target projection of meeting the target by the 16th of May is based on the following assumptions. First, based on the Singapore experience, failure to meet a critical mass within the first 3 weeks will result in the uptake rate stagnating and the app failing. We can revise the timeline target conditional on additional government efforts and incentives.

Second, the Australian government has already started easing some aspects of the isolation. Some states have eased more than others. Some opened up schools and others opened up parks. There is speculation from the media that by mid May we will even start opening up restaurants. If this happens and isolation measures begin to lifte by the 16th of May without meeting the 40% target, then we should consider this as a premature policy. This assessment is entirely in accord to the government’s own estimates of what needs to be done prior lifting isolation measures, and one of the preconditions is meeting the 40% target.

Australia data on downloads

The estimates are sourced from the Minister for Health statements and from newspapers, and they refer to number of downloads (not users):

Net users

We do not know the number of net users. We could not find this information in any official statement or analysis. The graph makes a naive assumption that there is a 0.5% daily loss. That is, for every 200 smartphone users installing the app, 1 of these users would uninstall the app by the end of the day. This assumption is entirely ad hoc but some loss must be assumed because of the technical hiccups discovered during the first 2 weeks of the launch and the low user experience (the app was rated 3 out of 5 stars in the app stores).

Singapore data

According to the Department of Statistics Singapore, in 2019 Singapore’s total population was 5,703.6 million, 86% of which had a smartphone.

Singapore’s TraceTogether app was launch as at 20 March 2020. We could only find two reliable estimates on downloads for TraceTogether:

  • As at 23rd of March, 3 days post the launch, Vivian Balakrishnan, Minister for Foreign Affairs, in a Facebook post he stated the app was installed by 620,000 individuals so far.
  • As at 1st of April, 13 days post the launch, the offical government website for TraceTogether stated that there were 1.1 million users.
  • As at 7th of May, 48 days post the launch, the offical government website for TraceTogether stated that there were 1.5 million users.
  • As at 1st of June, 74 days post the launch, the offical government website for TraceTogether stated that there were 1.9 million users.
%d bloggers like this: