Randeep Sidhu: Life and Death Decisions

Lessons From a Pandemic CPO

Randeep joined the UK government in July 2020 at short notice to build the NHS Covid 19 app in just 6 weeks after the first iteration was stopped. This was an unusually stressful time. The project was high extremely high-profile, urgent, and needed to play a part in potentially saving 1,000s of lives across the UK. No pressure.

Randeep shares the story of building the app along with some of the tools, techniques and processes that he employed to make this app succeed in some of the highest pressure circumstances he’s ever worked in. He shares his experiences across private and public healthcare and the lessons he learned in managing such a high profile project to offer some practical insights and about how humans work together, or don’t, to build products with real impact.

Covering everything from behavioural psychology to Marvel superheroes, you’ll leave with ideas about how you can have impact, and fun, in the face of chaos.

Slides

Find out more about BoS

Get details about our next conference, subscribe to our newsletter, and watch more of the great BoS Talks you hear so much about.

Transcript

I am speaking to you about some of my experiences, and they’re quite extreme, maybe not the same kinds of experiences as many of the audience will have every day, but my hope is they will be relevant, because when you have some experiences at the extreme, it helps explain some of the truths that we see in our industry. So a truth that is apparent to me. So when I started in health tech many years ago, someone said to me, who is a doctor? Randeep.

A product you build will kill someone. A decision you make will kill someone. What can you do to make sure your conscience is clear?

Now this wasn’t just hyperbole. Most of the decisions we make in building software and building products have an impact on our end users. Unfortunately, in healthcare, those can be life or death decisions, and all of you are like, Why the hell are we here?

So as I said, my announcement is this will be stuff that is hopefully relevant to the audience, but I will keep some time at the end for questions, and please grab me at the tables at the end after this session for stuff that’s more relevant. And it’s a bit of a kind of journey, unlike some of the other talks. So I’m going to be talking about my time working during the pandemic on the front lines of the UK, about having this kind of extreme stress in a daily, daily way.

But actually, lot of those experiences are fairly comparable to working in a startup. Yes, it was extreme, but they’re kind of quite similar.

Working Under Stress

So I’m going to talk about four things that I realized about working under stress, working in those kinds of environments. As I said, most of those will hopefully be universal, because I’m going to talk about behavior, I’m going to talk about psychology, I’m going to talk about some of those unhappy patterns that happen when you’re working in those environments, and that is relevant to lots of these rooms, people who are founders, people are building their own companies. And I’m going to try and be honest. I may not necessarily look good for that honesty, but I will try and be honest with what happened.

And the four realities I’ll talk about is under stress,

• What happened to me?
• What happens to people who are in senior leading positions when they are under stress?
• What happens to my teams when they’re feeling these kinds of stress in extreme situations?
• What happens to our users?

And summing up with what I realized about why creating change and impact is hard.

So me, Randeep, I have recently left a role as Chief Product Officer for a company called Reliance Health. They were headquartered in Nigeria. I was working in London, so I was remote – working and building healthcare across the emerging economies. Prior to that, I was at a company that is had had been headquartered in Austin, Texas, called Babylon. I built the NHS app, which I’ll speak to you about. I’ve worked for lots of nonprofits, sexual health, HIV, LGBT work, and I’ve advised people at the Alan Turing Institute, spoken at other big conferences, and I’m an LGBT leader.

The Story

So let me rewind you back to June 2020. Have a think, what were you guys doing? Where were you as the pandemic was kicking off?

I was working at this health tech company called Babylon, the largest primary care GP service in the UK, and I was building their Rwandan healthcare practice and undone. So while there, the government had spent three months, this is a UK Government, the National Health Service, building a covid app.

For those who aren’t familiar, it was a thing that would ping you on your phone to give you a notification if you’ve been exposed to covid. I’ll show you some screenshots of it in a second.

So the government spent three months building it as the start of the pandemic, and it failed catastrophically, publicly. It did not launch. And that was bad. As that happened, someone in the NHS gave me a call. It was locked down, I was at home, we were all at home. I was like, Fine, I’ll take a phone call. And that phone call was, “How would you build this differently around? It’s failed. It never launched. The government spent a lot of money doing this, but we may have missed something. Can you help us understand?”

So I was like, okay, had a chat, and I was talking to them about, out of the blue it was on a Friday night, and I started talking about the pandemic. I was like, Well, if you’re poor, give twice the death rate of covid. If you’re a person of color, at least in the UK, you had twice the death rate of covid. If you were black, you had four times the death rate. And I was speaking to them about the strategies that could be employed to make sure those inequalities could be dealt with. How do you help those who are most in need?

The first person I spoke to didn’t really have any answers, so they said, Oh, can you speak to someone else? And escalated me over Friday, Saturday, Sunday, I spoke to 5-6-7, different people, more senior, more senior, trying to get them to answer questions about how you could deal with this kind of inequality, to build a product that isn’t just technically accurate, but practically will help people. That resulted, I think I spoke to them first on the Friday, finished on Sunday, going, people are going to die. The team who are building this don’t have a clue. Tuesday evening, they call me back and say, Actually, we need you. So I told my employer on Wednesday morning. I left on Friday, I joined the government on Saturday morning. So less than a week from the first conversation, I was in government on the front line.

Now I know probably in America, employment law is different. In the UK, you do not leave a job in less than a week, that doesn’t happen. So they, basically the government called up my employer and asked for me to join. So the question is, why did I do it? Because many people told me I had that one night of people going, you shouldn’t be doing this. This is career suicide. If you are going to join government, you can’t build this kind of tech in government. It’s failed already. But the first team were bright, they were smart, they were capable, but it didn’t work. So why do you think you can do something different? So why did I say yes?

Back to the first slide I showed you, ultimately, so my conscience is clear, but if I’m being honest as I’m trying to be in this talk, there’s something else at play as well, which is interesting to unpack, because I think I could do something that others can’t. So it wasn’t just kind of public service. Was there some ego there? Was there some arrogance? Did I think I was some sort of savior? And by the way, I apologize up front for all the artwork I just had some fun with, with, with mid journey. You’ll see some more coming up. So I it. I said, fine.

First Area: What happened to me?

So of these four areas, I’m going to start talking about my behavior first. So I’m some sort of savior. My first reality is, I think I’m a superhero? So with this, I’m going to use one example, which is building languages within the app. So the app, this app was a thing you download in your smartphone. It was based on an API that Apple and Google developed that would interoperate between iOS and Android phones, and it had this green, the animated kind of heartbeat showing it was working, and it would ping you and tell you you’ve been exposed to coronavirus and you have to isolate or tell you how long for. And I chose to add a feature, which was a QR code check in, which was the only app in the world that did both contact notification and kind of checking in. And the algorithms we developed were working with the Alan Turing Institute, the lead body for AI in the UK. And those algorithms were shipped globally.

So if you have an Android or an iOS phone and you randomly got a notification saying, contact explosion notification in America, they were based on algorithms we developed in the UK with the Alan Turing Institute. So those algorithms are shipped globally, across iOS and Android. Happy to talk about that later at some point. So that’s the app. Dumpster fire I walked into.

So I walk in on Saturday, and given the three months that was wasted on the first app, we had to ship something in six weeks. I was like, let me get my job, dear people, my community is dying. Let me get on with it. So that’s, that’s the mission. I thought I was going in to do that was my superhero mentality. I’m in to build something people get out of my way.

And one of the first challenges, or one of the many challenges was languages. There are only two official languages in the UK that everything has to be translated into for government, English and Welsh. Now, if you’ve come from the UK, there’s not that many Welsh speakers, and most Welsh speakers speak English. And given the stats I showed you earlier, that didn’t feel like it makes sense. So I thought, Okay, here’s a here’s an easy win. We need to make sure the app works in languages that people you know speak in the UK. So I made sure the app was built in 12 languages covering covering over 98% of spoken language. So the app, the assets, the information.

And this was kind of unprecedented for the UK’s health service, because they didn’t typically do things in all these languages, and in terms of our thinking, for example, we didn’t have Hindi, Hindi as a kind of commonly spoken language in the UK. We didn’t have that in the app, because most people who spoke Hindi also spoke one of the other four South Asian languages, or English. We had a complaint from the Indian High Commission. Why do we not have Hindi? Because like most people who speak that also speak something else. But we did have Romanian, which a very small percentage of people speak in the UK. But those people are very, very insecure. They’re people who are meat packing, working in kind of insecure labor. So I was like, OK, that’s good.

The question is, working with my dev agency, how do we pick the language within the app? Simple. You can change it at the system level, or you can change it within our app. Very, really basic, really straightforward, the kinds of questions that get answered every day for people building software. So I was like, Well, you know what think about my mum? Think about people who maybe don’t have English as a first language. Their phone’s probably in English, but they use their apps and languages they’re more comfortable. They may be speaking on WhatsApp in Punjabi or other languages, but that app is different from the phone. So I said, Yeah, do you know what I want to have that app, being able to have the phone’s operating system in English, but have the app in whatever language you choose.

The agency pushed back. They’re like, No, no. It’s more effort for us. You want a language in it. Let’s just do it at system level. It’s easier for us if you do that. I don’t feel comfortable with that. That doesn’t feel like what our users need. So the challenge we had is our dev agency and the researchers were employed by the same company. So I was coming in, asking a bunch of researchers to do something which their paymasters were a bit uncomfortable with, because ultimately, if they said yes, they would have to build it, and we had to build at speed. So there’s a bit of a disincentive for doing what I wanted. They were trying to simplify it. And I’ve given the previous talk, maybe simplicity is better sometimes.

So I said, OK, do some research. I think it’s the right thing to do. You go and do some research to prove that I’m right. So they did, and then they had a big meeting with all the big wigs in the business coming around to present this information back. They were like great, Randy passed with some research to show that we need to make sure that the you know this thing he’s talking about isn’t important. And guess what the research with all the senior directors or the people who have seen you in government showing this research, do you know what that language search that Randy’s going on about? We don’t need it in the app. It’s happy to be at system level, Randy. It’s wrong. And it just felt weird, and it was quite annoying, and it felt quite angering. I was like, I don’t think these people are understanding what I’m trying to ask for.

So then I sort of stopped and went, let me ask some questions about your research, in this big meeting. Who did you speak to? Oh, we’ve done research. How did you find these people? Oh, you know, you use research agencies. We found that’s okay. How many of them actually spoke, you know, another language? Oh, everyone. But how many of them spoke English as a second language, not their first language? Did anyone you speak to have English as a second language? There was some silence and an oops. Everyone they spoke to might have spoken a second language, but English was their first language. Why would they need an app, a switch inside the app? They didn’t that was a problem for them. They weren’t native speakers like my mum. They weren’t immigrants, and at that point, huzzah, that’s why I’m here. This is a superhero that I am. I’m here to show this is what I’m here for. You see, I’m right. I am right. I’m not the hero that Gotham deserves. I’m the hero that Gotham needs. I have proven that I’m here as a superhero.

So my mission, of what I was trying to do as a leader, meant I acted like the villain, because I did not understand or recognize the power imbalance that was there. I was very senior. The team I was speaking to were less senior. So my flippant comment of, yes, granted, it was some anger and frustration there. Who this research based on? What. who you speaking to? Maybe if I’d realized, I could have taken them on the side and spoke to them one on one, but I did it in a public forum. I punched down in a way that’s probably very unfair, and ultimately, I was a bit of a dick. So this is where, when you think about, you know, the righteous mission that we’re all on the way to remember is your righteous mission doesn’t make you a hero, so be careful with how you act.

But that’s not all the story. Is it really? That doesn’t really come up to what happened, because the off me winning. Shall I say that battle had some consequences. The consequences. This presentation, even though I didn’t arrange the meeting, was public or visible to many people in the business, I had with my question and my anger undermined the work of these researchers, and not only that, I had made them feel stupid. I’d said you’re rubbish at what you do and you’re thick. Understandably. Me, that team hated me and didn’t just hate me. They proactively stopped me getting what I needed done for the rest of the time I was there. They hated me. And if you stop and think about it from their point of view, they might be right, because this column of whole you know, I think Monty Python said it best. I was not the Messiah. I was, unfortunately, a bit of a naughty boy, because when you think about building things under stress, anyone who’s a leader here will have a mission of what they’re trying to do. And the challenge is, with our myopic vision and our arrogance, we can sit there and think we’re changing things. We’re doing the right thing, but the reality is, when you’re under stress, you’re not likely to be that superhero. You think with stress, you’re most likely to act like a villain, and I’ll come back to this at the end. But what did that mean?

So that was my first kind of recognition and lesson from trying to build under stress.

Second Area: What happened to my team?

The second thing I move on to is about our team. So I wasn’t always a dick, but sometimes it happens. So under stress, chaos versus control, and the example I’ll talk about is having a product team or not. So when I walked in on the ground, we had to build something in six weeks, a research project for a product of this scale. This app was the fastest download downloaded app in UK history. It had, I think, 30 million downloads plus in the first weekend, very, very, very high downloads. If you had six weeks to build something that’s a research project, that’s a discovery process before you build we didn’t have time to do a discovery process. We had to build it. And as I stepped in, day one, I had no product team. We had 80 plus engineers and me. So all the people I was working around were incredibly stressed.

If you stop and think about health care, this is the original foundation document for the National Health Service, your new national health service, a way to make sure that everyone gets health care. Health care has to be universal. Technology isn’t always universal. So as I said before, WhatsApp is probably the most universal piece of tech, so we have to build something at scale and speed that’s universal, because otherwise people die. And my team, understandably, when we started hiring them, were really stressed. There was a lot riding on this for them, personally, professionally. And for me, it felt a lot like Groundhog Day as I got busy and kept working.

That is a genius. Genuinely, I’m quite impressed with that picture. Under stress, you are not at your best. We understand that, but unfortunately, neither is your team. And every day, as I was working, I keep getting the same things again and again. I wonder how many of this room have felt the same thing. So I walk in, and every day I’m repeating myself. I’m like, I just said this to you yesterday. Why am I explaining this to you again? Not just work based stuff, but what I’m about to do. So I’m not, oh yeah, this is my plan for the next week, and then the next day I have to explain it to someone else, and then a third person, and then a fourth person. I’m like, I spent every day just repeating the same information again and again to a various kind of group of people, and everyone had this strange intent of sinisterness. They would always assume the worst if I hadn’t explained something, their brains would always go to the worst case scenario of what I hadn’t explained. But why? Why do you think I’m somehow even maybe you know, the team thought I was a villain, fine, but the rest of them, why do you think I’m so sinister? And everything became emergency.

It wasn’t just you got the worst intent of my actions or the worst possible outcome. If that worst outcome happened, the world would end. Everything was an emergency, and that’s how they got in front of me. There’s an emergency. They have to speak to me to understand what my plan was for the product. And it got really tedious until I kind of stopped. At the end I explained maybe some of the strategies for how and then I was like, What the hell is happening to those around me, my team, the other seniors, and I started think, realizing it was adrenaline.

Stress causes adrenaline.

Adrenaline has two really interesting outcomes on your brain. It causes a level of anxiety and paranoia, not clinical, but we know this.

When you start getting stressed and it causes hyper vigilance. You’re looking for dragons, you’re looking for problems. And those two things is the reality of building under stress and building its speech. So instead of trying to tackle some of the things had on, I thought, what can I do to reduce the level of adrenaline in my teams? What is it I can do to make people feel emotionally more comfortable that will hopefully help me have more time to do my job? So I focused on lowering adrenaline, and I used a framework.

Now I don’t typically use lots of frameworks. They’re not I don’t think they’re always as useful. But in this instance, I had to build something fast. It had to be universal. So what’s the most universal framework that’s out there? What’s a framework that we think about that kind of helps as many people as we can. And I was like Maslow’s Hierarchy of Needs. I’m sure many people have come across this goes from the bottom of basic safety up to self actualization, being the most realized version of yourself.

So I thought, Okay, I need to create a structure that gives people reassurance, and I need to build something that helps people so let me try and apply Maslow’s Hierarchy to product, to the product I’m building.

Now apologies the next slide got slightly formatted weirdly, but hopefully you’ll be able to see it.

So I put Maslow’s Hierarchy down, and I went from bottom to top. What? Okay, so let’s use physical safety. So can I get the app? That’s the first question. Is it, how big is it? Can Is it free to download? Is it on my version of the operating system? Is it on my phone? Really basic. Think of it as a funnel. Can I actually get it?

Second question, is it dangerous for me? And if you stop and have a think about that, this app could tell you who’s around you. I mean, it’s all anonymous. It was all had no names. But if you were part of the LGBT community, as I am, and you checked into a gay bar, maybe someone could check your phone, and that might cause some problems for you. If you’re a survivor of domestic abuse and you had to go to a shelter, maybe your phone would tell your partner that’s where you were. If you’re an asylum seeker or got unsafe immigration status is something like this incredibly dangerous for you, and actually it was.

So for those two things, we actively worked on building techniques so for safety, we had a way you could delete your history and delete individual notifications for the physically getting the app. We worked with mobile phone operators and zero rated the data so you could use and download the app for free if you didn’t have the money to have data on your phone, or if you, for any other reasons, couldn’t so actually, it had a huge impact on how we built the product, and we moved up from there. Do I want the app? Practically, can I use the app back to languages? Back to using British Sign Language on accessibility? Does it do what I want? Is it accurate? Does it make me want to change my behavior, and does it make me happy? Start at the bottom, move to the top.

Now you’re like, OK, that’s a framework. It helped you make a good product. But how is this useful for your team? Because instantly, on one page, everyone knew what I was building, and if there was a problem, they knew where it ranked. Things at the bottom get fixed before things at the top, there’s always bugs, there’s always problems. And what it meant is, when there’s a problem that happened, instantly people knew where it was in my priority stack, and they could pre sort it before it got to me. A lot of the noise in the conversations that would happen didn’t happen because everyone was on the same page.

Is this a perfect framework? No. Is it a simple framework? Yes. Is it easy to communicate? I’ve literally communicated to all of you. Everyone can follow this now. So that meant when problems happened, everyone was really aligned on how to fix it quickly, and it meant that a lot of the problems got fixed before they even got to me, because someone would say, we’ve got four bugs, something at the bottom versus something at the top, we’ll fix the bottom one first. Brilliant.

It also had something quite interesting when it comes to accessibility health, trying to help those who are you know, have the least equal stakes when you think about building a product, most people focus on the top of the pyramid. But the bottom of the pyramid is accessibility, that’s getting in the hands of users. The top of the pyramid is kind of actionable change. Most people, when they think about product, say, Oh, we’re going to this hopey changey thing at the top. We’re going to make people think differently, be happy, do things differently. They don’t focus on how to get people into that room. And ultimately, it made it really quickly to align and really quick to fix. So in terms of working with a team that is under extreme stress, there’s other techniques I can come to if needed, but this was a really easy, really clear way of building products that help those who are most in need, and helping everyone be aligned about how to get there.

So not to use a really horrible phrase from the UK’s Brexit, but we gave back control through accessibility. We made sure that we signposted what we were doing. Remember, paranoia only happens when there’s a surprise that you know that you work you predicted. So that kind of sense of being proved right makes people more paranoid. So if you kind of give this visibility in the structure and some some level of reliability, people feel quite comfortable.

Second thing, we had a routine. There’s another routine I haven’t talked about, which was a kind of a weekly board where people would present. It gave a structure and a routine. People knew what my priorities were, and we reviewed them publicly. And finally, we forced a process, which meant people contributed. They didn’t just kind of complain from the sides, actively work on that and that let loads of other teams see other teams’ priorities, because there’s a simple framework. You can see how each team focused on those. And I will say a side note is we made sure people laughed. It’s an incredibly stressful environment. Decisions have huge impacts professionally. People are very scared. You have to make it okay to fail. You have to make it okay to be psychologically safe for something to go wrong. Because the reality is, things are always going to go wrong if people are scared and hide it, the impact of that is far worse, and they’ve seen that time and time again. So we made sure we had sometimes a light hearted environment, and we made sure it’s fun sometimes.

And I thought I’d share some of the amazing mistakes that happened in the process of building this just as a side note. So yes, it was a dumpster fire, but we had a consultancy who I won’t name, on the ground, and given we had 12 languages, I said, we need a CMS system, content management system. I mean, everyone in this room will know one of those is for some reason, this consultant, who was very senior, was like, no, no. Enterprise deployment of a CMS is too much. You’ve got six weeks. You can’t do it. Make this cloud solutions that can do this overnight. No, no, we don’t need a CMS, which resulted in some fun problems, because essentially, we had 12 languages that were being hard coded into the app and with people who didn’t know those languages. So unsurprisingly, we had some fun.

So one of the translations came back, and it was the Punjabi translation, and it was for a video. And I was watching the video going, I should be able to understand this, but I don’t, and it’s confusing to me, so I sent it to my mum. I thought, Mum, watch this video. Does this make sense to you? She was like, What the hell what the hell is this? I was like, this is the same agency, one that the NHS have used for many, many years. So this should be an agency that’s good, but two, they’ve translated our app. So Mum, look at this screen in the app. Does this make sense to you? So I send some screens to my mum. I get an urgent phone call from my sister, losing her mind, going, you cannot release this. This app is dangerous. What you mean?

So some of the screens are wrong. So a screen that says your coronavirus test is negative, it doesn’t say your test is negative. It says your test is worthless. It has no value. OK, small problem there. So I’m like this agency shit, and no one has checked their work, obviously, because they’ve been working for the NHS for many, many years, no one’s bothered checking because they didn’t have a diverse enough team to realize when the translations were wrong. So I then activate a WhatsApp group and get every single person of a different ethnicity onto what genuinely like. We’ve got this Indian WhatsApp, Indian aunties Whatsapp group, and just start throwing screens to them to get them to check if they worked. Not ideal. But what happens?

Some of the other side problems that came up. We had the Polish embassy in the UK contact us, because one of the screens where it said, Congratulations, you don’t have coronavirus, for some reason, just said, Good village. I just show is making this up.

But my favorite, favorite one, the Welsh Government, in official ministerial process, sent a letter letterheaded full process, kind of ministerial rubbish to the government to say one of the screens doesn’t say what it’s meant to be and just has the word moist.

Now it’s quite fun to try and understand where that screen was, which I never quite found out. But actually having that described on a formal like letterheaded paper from the government is quite entertaining. So like I said, we can laugh. You have to make sure your teams do too, because genuinely laughter reduces adrenaline. So I say this as a flippant comment, but something that’s serious. Make sure when your teams are stressed, you give some time for light hearted activity. So, fine.

We’ve talked about my behavior of sometimes being a dickhead, the team behavior of being under stress and needing to relax and not always having to be under that stress.

Third Area: What happened to our users?

Let’s think about our users. What happens when our users are stressed? Covid was quite unique, because everyone was stressed at that point. So we’ll think about chaos and control, and I’ll talk about an example of our QR code check in feature. So reminding you again, the world of our users, which is our world. Code was out, and Boris Johnson, our commander and clown in chief, was making a bit of a mockery of the whole process. Mark said, don’t talk about politics. I will quite clean talk about British politics, this fool should never be in government. And the problem was, people didn’t trust our government, necessarily. They didn’t think that what was happening around them was good or safe, and so they were stressed, they were scared, and they had the distinct pleasure of having people hate the app before it was ever launched. Publicly, there has been three months of building an app. Then they were going to spend another six weeks building a second one that I was in charge of. Everyone was in the newspapers. It’s a waste of money, it’s a stitch up. It’s cronyism and friendships and backhanded deals. And that was even before I put a line of code out. Before we launched anything, people hated it. Once it became live, people wanted to keep hating it. It’s a very strange environment to release a product. You’re not neutral. You’re having to deal with a kind of an anti backlog against this thing, this backlash.

And one of the things we built was a QR code check in feature. So to clarify, most people’s apps would just be a kind of a thing in your pocket that would pin going, you’ve got covid, or you might have covid. What I realized was you were having to go into these contact tracing met when you were going to venues, you’d have to write your name in a book so that, in case there was a problem, someone could call you and tell you, Oh, this venue had a coronavirus outbreak. I realized that having a QR code check in future will be really beneficial and psychologically beneficial to overcome some of the negativity we inherited. So why did people use it? Why did this thing become a bit of a special Trojan horse and make people use the app? Because I realized there was an unhappy dark pattern happening, which was we had a lot of negativity, and this thing in people’s pockets, which was equally negative.

And we thought about, how do you increase adoption when people are stressed? If people are stressed, making them more stressed is not good. It kind of sounds obvious, but think again, how many times when you’re trying to get someone to do something a user, do you just put more pressure on the system? Do you resort to threats? Do you resort to fear and panic to motivate people to do something that you want them to do, that they may not be interested in doing? Now, it’s a natural resource. When your kids do something wrong, you shout louder. That’s what we do. But it’s really important to stop and think what happens in an extreme situation like covid is equally applicable to most of our users. What we tried doing is something slightly different, instead of putting more pressure on them, give them more of a sense of control.

So if you think about this app, people were scared, and the app worked with just sending you a notification telling you you’ve got covid, or you might have covid, it’s like walking around with a grenade in your pocket, randomly, it’ll just blow up and then ruin your life, or ruin a couple of days when you’re in lockdown. But it’s not a good note. It’s not a good. Experience. It’s not things that people want. They’re not going to want to have this thing, and if it does blow up in their pocket, they’re not going to want to trust it. It’s a very negative way of interacting. So what you need to do is give them some sense of control back. So what we realized was having this sense of a QR code check in meant when you walked into a cafe, you could check in, it gave you protection immediately. It was a bit like buying an insurance policy just for going into the cafe. It’s a different kind of interaction. You check in, I had a nice little kind of animation with a tick and a smiley face and something that felt reassuring. So every time you check in, you get a dopamine hit. And that dopamine hit makes you think, yes, the world is scary. Yes, covid is out of control, but I’ve just bought myself a quick insurance policy. For the next hour, if anything happens in this cafe, I’m protected. And it sounds ridiculous, but it worked.

People are quite basic sometimes, and also it was much easier than some of the alternatives. If the alternative was sitting down, writing your name in a book, having to kind of go through that effort, and also that level of exposure, there was loads of situations where, you know, say, a woman was walking to a bar and someone behind them liked the look of her and took the number and started texting them. That happened repeatedly. So there’s people feeling uncomfortable giving their details out. This system was anonymous, it was secure and it was fast. If you were exposed to something, it would tell you immediately. So it wasn’t just easier to use. It reduced the friction of existing alternatives.

So with that in mind, to summarize how I think about trying to build a product when your users are stressed. I’m going to use an analogy of a gun, probably not the most politically relevant, but let’s just, follow me. If you’re thinking about speeding up a bullet, how would you do that? Psychologists often talk about this as an analogy, because if you think about using threats, threats is a bit like putting more gun power in the bullet. It feels natural. Gunpowder, the bullet goes faster. That’s a natural thing we do. And it kind of works to an extent. But the problem is, if you put too much power, the bullet just explodes before it even gets to its target. There’s too much stress. You can make the bullet more aerodynamic, you can reduce the friction, and that makes it go faster. And that’s the way I’d think about trying to build user products when they’re stressed.

Ultimately, what your users need is a sense of increased control and some reduced friction.

And the way and the strategy that we used worked. The QR code check in feature, this was the only app globally that had both was seen as one of the leading drivers of success for the app. I’m a scientist by training, and I never thought any of my work would ever get into the journal nature, but it did. So, this app was featured twice in nature, and the research proved it stopped a minimum of a million infections in the UK, and saved the most conservative estimate is 10,000 lives. It’s likely multiples of that because of how it worked. We essentially halved the number of people with covid in the hospital. So if you think about halving the number of people seriously ill with covid, you can think that number was far, far higher. So brilliant. We had a success. The app worked.

So now I’m going to change tack slightly, instead of just talking about stress on you as a leader, on your team and on your users, I’m going to think about having impact. So some of these experiences that I’ve had, I’ve always worked in high impact environments. Brilliant, great. Seems really nice to be able to have that kind of impact in your work. But the more you do it, the more you realize something’s not always right in the world.

Why is impact hard and how do villains take part?

So I’m going to go back to an example of user research. And I’m going to say as many downloads as we had, as successful as we were, there was a worry that our user numbers weren’t high enough. Everyone in this room will have that problem. Our user numbers aren’t high enough. I was a director inside the National Health Service. Never planned to be in government. Didn’t want to continue in government afterwards, but I was there, and I’ve got a job to do. And being senior, there was a kind of research so you’d be in a room, unfortunately, you’d be very monocultural in that room, because senior managers and directors are often of a certain kind of ethnicity, class, culture, that’s just a reality. And someone comes in and shows some research, they’re like, OK, team all the most senior people in this division. Let me show you some research of what’s happening. And this research said, Yeah, user numbers aren’t always the highest but what we’re seeing is black men are not using the app that much, and the Muslim community, they’ve got lower levels of adoption. That was presented this room and then the conversation started.

So as the only person of color in the whole call. It was quite interesting hearing the way that the conversation went. And these were people who were, unlike me, I just joined government. These were people who were kind of established there for a longer period of time. These are some of the quotes that I can show you that were ha that this is, I mean, the verbatim would be even worse. But these are some of the things I heard in the room while this was happening. Excuse me,

if these people don’t want to work with us, what can we do? It’s not our fault that these people are what too stupid to follow, too ignorant to listen. It’s not our fault as a business if they’re not following what we want. So I wait. I think people are just maybe speaking out of turn. They’re not thinking about what they’re saying. But obviously it’s not the best room to be in when this is happening. But I wait and I’m like, someone’s going to say something, someone’s going to do something. They do. Put my hand up, and as much as I was angry and frustrated, I kept my tongue and I just asked a question. So it’s really interesting thinking about what I’ve heard because, of course, it’s really understandable why some of these communities might not be engaging in the app, because at least in the UK, there’s some inequalities that you know, sometimes the government isn’t always best to these communities. So why would they trust something? You know, we would never stop giving overweight people health care. So why are we? Why are we ignoring these communities who aren’t why are we writing them off?

And then something interesting happened, the person who’d made some of these comments, I didn’t address it, there’s maybe 40 people in this room. Everyone else started speaking, so the person who made the comment didn’t actually respond. The rest of the room started defending both what he said and stopping me challenging what was happening, which I found incredibly interesting, so I asked a fair question. The room appeared to turn against me. So not in a kind of I wasn’t being dick, I was being cautious of what I was saying. Many people have been in that situation, we have to call something out, and so I did it in the most polite way.

So I’ll come back to what that my understanding of that conversation is. But what we moved to is thinking, okay, to cut the conversation short, I was like, Okay, well, let’s do some research. I have assumption that the way we’re thinking about the problem is not right, so I said, let’s do some research quickly. Who do users trust? Do they trust our covid app? Do they trust the division test and trace that was building the covid app. Do they trust the NHS of which this division was part of? Do they trust the government? Or do they trust Boris Johnson? Simple. The room I was speaking to was full of government employees. They thought users trusted everyone. I mean, yeah. I mean, kind of, it’s obvious to think that. So we did research. We asked general public, and people use the app, who do you trust? And majority of people didn’t trust the Prime Minister, unsurprisingly, they’re a bit on the fence about the government, but they trusted everything else. Because the NHS is a great institution, it helps so many people, it’s the kind of leading light of our country.

Then you ask these rejectors, those communities I spoke of, who do you trust? Unsurprisingly, they didn’t trust Boris Johnson and not unsurprisingly, they also didn’t trust the government. We asked, why? Over policing. Historic discrimination against the Muslim community, persecution, the things that then these communities have engaged with in government, there haven’t been positive experiences. So if the thing that’s tracking you is connected to government, why would you use it? And then that connection to government also put massive question marks over systems that they generally used to trust, like the NHS. It didn’t feel like it was part of the NHS. It felt like it was an arm of the government. I mean, of course, the NHS is built by the government, but it felt like there was this wall between them and this app somehow felt like it was now government overreach, some sense of control, some sense of persecution, identification, seeing who you’re hanging out with. So that wasn’t good.

So I used this research to change the conversation that was happening in that room effectively. So we started having different strategies for dealing with them at the end. I’ll show you one example of that strategy. So yes, that room was confusing. Yes, it was frustrating. There was a way we navigate it using data, but if I go back to that room, it’s a very interesting kind of way to think about achieving change. That room was angry at me somewhat. Why? What was my what was my sin? What had I done wrong? I was polite. I never accused anyone. Any person who’s had to have that really difficult conversation would know you watch your words and your tone and your demeanor so carefully, because you don’t want to be the angry brown man. You don’t want to be the person who’s calling out truth to power, but you have to do it in the right way. So yes, I was.

What did I do? I had a genuine, innocent question. The thing I realized I did wrong, I challenged the status quo. Now there’s a bit of audience participation here. I’m going to show two different groups of people. I would like you to decide which group you’d like to be part of.

• Group one. They need to be liked. They defend the status quo because they think things are good as they are. They’re reactive, They react when something needs to change, and they’re comfortable. If you can read that, they don’t look to improve the world, They’re happy with how it is. Group one.

• Group two. They don’t need to be liked. They challenge the status quo because they know today is not good enough. They are proactive, and they’re striving. Now I’m going to do a show of hands. I’m going to start with the blue group. Who would like to be in the blue group? Show of hands.

Who would like to be in the pink group? And for those who can’t see, that’s pretty much the whole audience. If there’s anybody who put their hand up for blue, I’ll chat to you later. But what’s interesting is I’m now going to bring this into something else. I’m a complete film nerd. But that being said, I think films and storytelling are a really interesting way to look into psyche of a country or a nation. So when you look at that, and you think about all the movies, all the things that we follow, the blues are the Marvel Superheroes, the pinks are the Marvel villains, and I’ll explain why. They’re very, very close, but they want different things.

Heroes don’t change the world. Superman, he could do whatever he wanted. Superman only acts when someone else causes a problem. Superman’s comfort with the status quo, the people who in our public discourse, in the stories we tell, in the mythologies we watch, the ones who are villains are the ones upsetting the status quo. Are the ones causing problems for people. Now it sounds strange, but they are very close and want different things. I’ll use one example from Black Panther – Killmonger, the villain of Black Panther, actually was right. He was calling out Wakanda for saying, You’re not doing enough to help black people across the world. He was right. However, his methods were wrong.

The thing that happens in our current culture, we’re so vigilant against people upsetting the status quo, against people calling out truth to power. What we do is we shut down people who are trying to create change. We vilify them, not just in our communities, not just in our boardrooms, but in public consciousness.

Villains are the ones who are kicking out change, because challenging a status quo creates a level of discomfort, and humans in communities are very uncomfortable with that. Me calling out that question, however politely, however non judgmentally, just me raising that question caused a room to turn against me because I was challenging the status quo and someone felt uncomfortable. Yeah, we get in group bias. Sometimes, if you’re a person of color doing it in a room with white people, that might be another difference. But it happens in any community, in any change. If you’re someone that wants to create change, genuine change, you have to understand that creating that change means challenging status quo, which creates discomfort, even if you’re doing it in the right way, people will vilify you for it.

If I go back to my first example of that research, what did I do wrong? Why was I a villain in that first example, I had that research project in that room when I, you know, called out the group for not using research that was, you know, with people who had English as a second language, punch down, I had a power imbalance which most of us in this room will have because we’re senior. I. Punched down and I acted like a dick. The reality is, when you call out things, when you challenge the status quo, people will vilify you. They will say you’re a villain.

The problem then, is, if someone says you’re a villain, you act like one. What you have to do. And I know this is a kind of weird and abstract problem, but I genuinely believe it change is hard, because challenging the status quo makes you seem like a villain. It invites people to act or think of you as being a person who’s a problem. Everyone wants to create this whole room, wants to challenge the status quo, but yet, when we try things change. You have to be uncomfortable with that level of discomfort. You have to know you’re right, because if you don’t, you then inadvertently end up acting in that way.

Make sure you punch up, not punch down, and think and act with empathy.

So I’m almost coming to an end, but I just wanted to summarize what I’ve shown you over the last hour. So lesson one, you’re not a hero, because your ego and your power makes you likely to be more of a villain than a hero. My lesson two, you and your team aren’t okay when working under stress, you genuinely aren’t. So think about being predictable. Think about empathy. Think about techniques that you can use to reduce adrenaline. Frameworks are good. Other things you can do as well your users when they’re stressed, there’s this balance between chaos and control. So what you need to do is make sure you increase their feeling of control and reduce that friction. And finally, villains can change the world, but you have to understand the dynamic that happens when what you do makes you seem like a villain. You just cannot act like that.

And that was the last slide, and then say, there’s another five or six minutes which I’m happy to take questions, or I’m just waiting for a signal from my back phone, no, but yes, in which case I will just finish and say, Thank you for your time.

Q&A

Mark Littlewood 

Some questions this first Q&A that we’ve got at BoS. So two things you need to know, I hate waiting for questions to happen.

Randeep Sidhu 

There’s a hand up there.

Mark Littlewood 

desperately impatient. So we’re going to get Mike five over there, and Mike six is here. Has anyone got a question? You put your hands up. I’ll make sure the mics are ready for you. Is that a bid or a question?

Audience Member 

Did you Did you notice an outside, an outsized adoption of those previously marginalized communities that were not initially adopting the platform?

Randeep Sidhu 

So you cannot overcome a structural inequality. We did see some higher adoption for this than other services. A lot of the communication did help. We did a lot of work with community leaders and trying to make sure that people could trust not just what we were doing, but they could check our homework and look at the receipts. We had higher adoption than expected, but it was not as high as the white communities, and that just is the situation. So we would want it to be higher, but we had great adoption and more than they expected, but it was not equal, unfortunately.

I’m sorry. Try that again, again. Yes, please. Try again. I can listen in this.

Audience Member 

How do you walk it back once you’ve already punched down?

Randeep Sidhu 

How or how do you walk it back when you’ve punched down? Acknowledging it. Ego. Ego is what we do. I mean, look, I’m not an expert. Everyone here is a human who can, like, think about it, but genuinely. And because that thing in the example I gave in that research project, I messed up. I did it in a way that I shouldn’t have done, and as soon as I realized it happened, I should have turned it back in the room to let them save face. The fact that I didn’t do that hurt me and hurt the project, because that team actively tried sabotaging what I was doing afterwards, which hurt everyone. So I think if I’d acknowledged it and dealt with their egos, it probably would have been better any Oh, yep.

Audience Member 

Mark, so do you think you fundamentally changed the view of those people in the room who are being defensive, because this is how we’ve done it, and this is how we’ll always do it, or do you think they’ve gone back to using the same Bangladeshi translators? And if so, how do you think we can make that change?

Randeep Sidhu 

So we have changed, because the NHS is no longer using that research agency. So the researcher, the translation agency, the room that kind of complained or kind of was resistant to change. I don’t know if anybody in that room. The problem is us in power have to acknowledge how we act. That’s what the purpose of this talk is. Because those underneath us see it and want the change. So a lot of the change I’m talking about didn’t just happen because of me. I created an umbrella. So soon as I said, this is that Maslow’s hierarchy. When I said that exists, the people underneath me who wanted to have better processes, more equity, more inclusion, they did the stuff because I gave them cover. So the reality is, having people in power who don’t really think about their impact of their actions is what stops those underneath us who might be able to help do those changes. So I don’t know if that room changed their point of view, but what I do know is the people beneath them, who I was working with, who will eventually get to those positions of power, they did change. So my hope is it kind of gets better over the future is my hope.

Audience Member 

I have a question about, sorry, I have a question about the safety, because there is always that balance between it’s okay to fail, right? But if you fail, like too many times, and it’s still okay, you just become too much comfortable, maybe, and you don’t try to succeed because you know you it’s simply like too okay to fail and nothing’s gonna happen, right? Like there is no consequence too. So what are your thoughts on this, like balance between comfort and psychological safety.

Randeep Sidhu 

I don’t think people are inherently wanting to fail. Humans have a desire to do good. Look at look at a bunch of school kids. If you say well done, they feel amazing. So this idea that if you don’t put pressure on people, they’re just going to want to fail, is probably misplaced. The reality is, psychological studies look at the kind of Japan airways. You look at all these examples, not making it psychologically safe to report doesn’t reduce errors. It increases risk and increases danger. People die when you don’t, kind of admit when you made a mistake. So you have to make it safe for people to make a mistake. But what’s interesting is what I’ve seen multiple times because I’ve worked in healthcare for a while. If people feel comfortable to own up when they make a mistake, they’re also more likely to ask for help, and then they feel less likely to make a mistake in the first place. So you create that environment, and people aren’t spending a lot of energy being stressed and worried, so they perform at their best, they’re less likely to cause those problems.

So in my opinion, you can put some barriers and some safety things in place, but genuinely expect the best of people, and they’ll generally deliver. But I’m conscious of time. I think we’re probably out, should we try the video again? Or are we up? Let me try, in case it works, and then I’ll jump off the stage. Oh, I hear something. No, try one more time. Come on. Come on. It’s a little PowerPoint slide that could and now the clicker stopped in, which case I’ll say thank you.

Mark Littlewood 

Okay, Randeep, thank you.

Randeep Sidhu

CPO, Reliance Health

Randeep is a Healthcare and AI specialist, most recently CPO for Reliance Health in Lagos, Nigeria, focused on creating healthcare for marginalised communities across Nigeria, Egypt and other Emerging Markets.

Randeep joined the NHS frontline to build and run the NHS COVID-19 app, which was proven to have saved over 10,000 lives and halved the number of Covid cases in hospital, while pioneering specific research to include the needs of LGBTQ+, refugees and PoC communities.

Previously, he’s advised The Alan Turing Institute – the UK’s National Institute for AI – in removing AI bias across healthcare and policing, been Product Director at Babylon Health, where he rolled out AI enabled healthcare to patients across markets like UK, USA and Rwanda.

He is currently a board member at Terrence Higgins Trust – a sexual health and HIV charity – and LVNDR Health – an LGBTQ+ health care startup.

Next Events

BoS Europe 2025 🇬🇧

🗓️ 31 March – 1 April 2025
📍 Cambridge, UK

Spend time with other smart people in a supportive community of SaaS & software entrepreneurs who want to build great products and companies.

Save your Spot

BoS USA 2025 🇺🇸

🗓️ To be announced soon
📍 Raleigh, NC

Learn how great software companies are built at an extraordinary conference run since 2007 to help you build long term, profitable, sustainable businesses.

More Details