It was a pleasure to liaise with such a professional journalist during the app’s development - thank you, Rory.
I believe that this app provides the scientific basis for many other NPIs that may need to be deployed in the future. Proportionate use of data-oriented interventions, that put humans in charge of decision making, can make a big difference to society.
Freedom of speech is fundamental, but too often we obsess on the right to that freedom, and not on the responsibilities that accompany it, and the consequences of ignoring them. Thank you for reflecting on this.
It landed at by he moment of maximum fear when people were most worried about the disease and t he vaccine was still so distant. In retrospect it’s not surprising it became a lightening rod for people’s fears. If we had anticipated that, we’d have kept it’s Devlopment secret, but we were trying to be transparent about our plans.
My Lord, with respect, it is perhaps just as well then that you did NOT anticipate that.
As it was, as Rory pointed out very well, the government's initial plan for a centralised app was untenable, given public opinion. Without public awareness and protest, that app would have continued to be developed and released.
Consider the result - a centralised app developed in secret? That would, I am certain, have lead to a PR and operational disaster. Uptake would have been nearly zero, and if you'd tried to legally mandate it, you would probably have collapsed all public engagement with any government advice on the subject, with devastating results.
The main thing I saw as misguided with the final app was that it did not require people to scan out when leaving places, only to scan in on arrival.
I talked to many people who saw that as giving an unreasonable risk that they would get unwanted false positives from being linked with other people who visited the same place, but not at the same time. That was the major cause as far as I could see, for disengagement from the app.
I was part of both app teams - and agree that greater levels of privacy protection, in the end, resulted in stronger health outcomes (through greater levels of adoption). Sadly, to Lord Bethell’s point, I feel that there was too much pressure / hope placed on the app at the beginning of the pandemic (through government communications) which, at the time, rightly drove a sceptical media response. Far from being binary communication (it would never be appropriate to develop it “in secret”), we could have been more precise with the public communication - that this was the deployment of an unproven technology, but it has the potential (now proved) to be a useful tool in our arsenal.
In retrospect, I remain unconvinced by the efficacy of location-based alerts (manual or digital), so not sure the feature that you suggested in the response (David) would have made any difference to health outcomes. Happy to see scientific evidence which proves me wrong.
Hi Mark, I have no scientific evidence, but that is again not quite my point. As with my previous one about the two versions, I'm thinking more of public engagement and confidence in the product.
The project feels to have been focused solely on its technical goals, which as a developer myself of tools that people are told to use for compliance purposes, rather than actively wanting, I understand and can sympathise with. Compliance is critical.
But I also have an active role in engaging directly with the end users whose companies have purchased our tools, and have come to understand that an application should wherever possible also meet the expectation of the people who are going to actually use it. If a function is implied by other parts of the solution, either it needs to be there, or the reason for its absence must be obvious.
The sign-in was "sold" to the public as a key feature of the app, and one that people could intuitively understand the benefits of. This drove engagement, and people installed it on that basis. Given a sign in feature, the app needed either to clearly explain why a sign out was not necessary, or provide one. Omitting it undermined public confidence in the product.
Heck - it might even have made the location tracking a little more effective.
I don’t disagree with the points that you make, especially around making the sign-out feature (or lack of) more intuitively obvious as to why it’s not necessary. I don’t think this “feature” cost the app users though, so overall didn’t impact the efficacy of the intervention.
Why does this not surprise me ? Good news doesn’t sell newspapers and it’s so much easier for the media to knock than support. Concerns over data protection remain but the (now) undoubted benefits are clear.
I have to wonder if it’s at least partially due to the fact that reporting *anything* about the ongoing pandemic seems to be frowned upon these days.
Irrespective, it’s certainly a pity that the successes are being buried along with the failures.
That’s fascinating, Rory. Thank you. Let’s hope it now becomes more widely disseminated. You’ve left a big gap at the BBC, that’s for sure.
It was a pleasure to liaise with such a professional journalist during the app’s development - thank you, Rory.
I believe that this app provides the scientific basis for many other NPIs that may need to be deployed in the future. Proportionate use of data-oriented interventions, that put humans in charge of decision making, can make a big difference to society.
Freedom of speech is fundamental, but too often we obsess on the right to that freedom, and not on the responsibilities that accompany it, and the consequences of ignoring them. Thank you for reflecting on this.
It landed at by he moment of maximum fear when people were most worried about the disease and t he vaccine was still so distant. In retrospect it’s not surprising it became a lightening rod for people’s fears. If we had anticipated that, we’d have kept it’s Devlopment secret, but we were trying to be transparent about our plans.
My Lord, with respect, it is perhaps just as well then that you did NOT anticipate that.
As it was, as Rory pointed out very well, the government's initial plan for a centralised app was untenable, given public opinion. Without public awareness and protest, that app would have continued to be developed and released.
Consider the result - a centralised app developed in secret? That would, I am certain, have lead to a PR and operational disaster. Uptake would have been nearly zero, and if you'd tried to legally mandate it, you would probably have collapsed all public engagement with any government advice on the subject, with devastating results.
The main thing I saw as misguided with the final app was that it did not require people to scan out when leaving places, only to scan in on arrival.
I talked to many people who saw that as giving an unreasonable risk that they would get unwanted false positives from being linked with other people who visited the same place, but not at the same time. That was the major cause as far as I could see, for disengagement from the app.
I was part of both app teams - and agree that greater levels of privacy protection, in the end, resulted in stronger health outcomes (through greater levels of adoption). Sadly, to Lord Bethell’s point, I feel that there was too much pressure / hope placed on the app at the beginning of the pandemic (through government communications) which, at the time, rightly drove a sceptical media response. Far from being binary communication (it would never be appropriate to develop it “in secret”), we could have been more precise with the public communication - that this was the deployment of an unproven technology, but it has the potential (now proved) to be a useful tool in our arsenal.
In retrospect, I remain unconvinced by the efficacy of location-based alerts (manual or digital), so not sure the feature that you suggested in the response (David) would have made any difference to health outcomes. Happy to see scientific evidence which proves me wrong.
Hi Mark, I have no scientific evidence, but that is again not quite my point. As with my previous one about the two versions, I'm thinking more of public engagement and confidence in the product.
The project feels to have been focused solely on its technical goals, which as a developer myself of tools that people are told to use for compliance purposes, rather than actively wanting, I understand and can sympathise with. Compliance is critical.
But I also have an active role in engaging directly with the end users whose companies have purchased our tools, and have come to understand that an application should wherever possible also meet the expectation of the people who are going to actually use it. If a function is implied by other parts of the solution, either it needs to be there, or the reason for its absence must be obvious.
The sign-in was "sold" to the public as a key feature of the app, and one that people could intuitively understand the benefits of. This drove engagement, and people installed it on that basis. Given a sign in feature, the app needed either to clearly explain why a sign out was not necessary, or provide one. Omitting it undermined public confidence in the product.
Heck - it might even have made the location tracking a little more effective.
I don’t disagree with the points that you make, especially around making the sign-out feature (or lack of) more intuitively obvious as to why it’s not necessary. I don’t think this “feature” cost the app users though, so overall didn’t impact the efficacy of the intervention.
Why does this not surprise me ? Good news doesn’t sell newspapers and it’s so much easier for the media to knock than support. Concerns over data protection remain but the (now) undoubted benefits are clear.
Excellent and very informative, thank you. I had also missed the Full Fact clarification of the cost of the contact tracing.