Ned  Gartside 

Ned is a service designer at the Department for the Environment, Food and Rural Affairs. Half his time he spends on a project to create a new digital system to track waste and resources through the UK and empower the move to a more circular economy. The other half he works on ‘sustainable design’ matters, and has led the development of a set of principles for the design and delivery of greener services. He used to work in agencies and startups. He gets quite excited about systems thinking, sustainability in general and running up and down hills.

How might we design and deliver greener services?

Summary: There’s a climate crisis and organisations big and small are setting all kinds of targets for reducing carbon and environmental footprints more broadly. How can we design and deliver services that serve this agenda and serve to reduce our environmental impacts?

The talk will divide into 3 main parts:

  • The ‘promise and peril’ of digital: how digital can simultaneously enable and endanger our shift to sustainable future
  • What are the typical environmental impact hotspots of a (digital) service?
  • What do we need to start delivering greener services on our projects (covering standards, measurement and training)?

Thank you. Thanks so much everyone for coming to this talk. Welcome. Yes, I am Ned, I'm a service designer or was at DEFRA, I'm also now doing a bit of greener services stuff which is what I'm going to talk about today. How am I to start delivering greener services is the question and it's about a 40‑minute talk so I'll try and get through it. There might be a couple of minutes for questions, but there also might not be. So probably, if you don't get chance to ask a question, do come and speak to me afterwards. If you would like to. So, let's get to it. I think maybe before we begin, we'll try to answer that question. It's worth spending a couple of minutes thinking about what is the problem that we are trying to solve on a high level. Most prominently when we talk about this stuff, we are probably inclined to think about climate change initially and climate change, as I expect we are all pretty aware of now, is developing as one of the main risks confronting us in the middle of the 21st century, and we have got a UN treaty in Paris in 2015 and national legislation in many countries, including the UK, which details commitment to reducing greenhouse gas emissions, usually measured in terms of carbon dioxide, to net zero, by 2050‑ish. So, the UN treaty says miss century,  the UK is committed to 2050. Then there are many organisations when talking about a third. Many are committed to reducing the carbon footprint which is good. When I talk about sustainability, a term that gets much overused and abused, it's important though I think that we are not just talking about carbon footprints, it's really important to avoid carbon tunnel vision on this topic. I think the best definition, the definitive definition for many is, from the UN report, the common future report in 1987 which defines sustainable development and not sustainability, but it's fair to treat it as a synonym really, as meeting the needs of the present without compromising the ability of future generations. It's broad in terms of the dimensions, economy, society and environment. It's not all about the environment. Actually, those three dimensions are reflected in the sustainable development goals which were created, I can't remember when, but a few years ago. Many countries, including the UK, are committed to trying to realise by 2030. So we see climate action here at number 13. Then we have got the environmental aspects pointed to in various points, you know, water, life on land, and then, of course, the economic and social dimensions are reflected in many of the other sustainable development goals.

That is good. So if we want to talk about things beyond carbon, I think it's quite helpful to be able to point to the SDGs, as they are often called, and the fact that the UK is committed to them.

So just before we begin, you might be wondering, why on earth is this guy talking about greener services then? Why we have to introduce another term here? This is something I agonised over a bit, but I chose the term greener services to try to emphasise the environmental sustainability aspect amongst those three dimensions we saw with the UN definition without the focus necessarily being all on low‑carbon and acknowledging that most legislation and reporting organisation levels tends to focus on carbon so the exclusion of most other impacts.
So to the question, how might we design and deliver greener services? I'm going to try to answer that question by asking three more questions. The first is, as you can see: Promise and peril. How might digital simultaneously enable and endanger the transition to a sustainable future? This is important to be on the same page as the nature of digital and the potential, but also the problems it presents from an environmental impact perspective. The second is zooming into the level of services, so where might a services environmental impact hotspot likely be? Thirdly, trying to get more practical, so what do we need to actually start delivering some of this stuff? So the talk is going to divide into the three sections. So, straight to it. Number one, how might digital enable and endanger our transition to a sustainable future? I think the first thing to say is that, digital or the digital world, ICT, sometimes labelled the Internet, has a direct environmental impact. And that comes from the whole life cycle of the whole ICT system, if you like. So the live use phase, we have end users on their devices through to the networks and data centres and infrastructure sits behind that. The other phases of the life sickle so the mining of metal and resources, the manufacturing and distribution processes and disposal at the end of life of these devices, whether talking about service or data centres and end user devices.
These, I put these in two red boxes here. So we have got the live impact when the whole ICT system is being used, and then the embodied impact such as the other four stages of the life cycle. If we are talking about carbon, again, I'll come back to carbon, but these are things which increase carbon footprints, effectively, because they're associated with distribution and mining and that sort of stuff. So we look at the direct effects of the ICT sector. Then we are looking at pretty massive footprints. So it's now, there is a Swedish study or research group that publishes the figures. The most recent point to a total carbon footprint of 763 mega-tonnes. I don't know what that is or looks like, but sounds like a lot and it is a lot. If we were trying to get to net zero, or if we wanted the ICT sector, in terms of its direct effects to get to net zero, we'd need that to come down sharply. It's not. And perhaps the fact that it's not rising sharply is, it's heartening in some ways, especially when we consider the traffic on the Internet's increased exponentially over the last ten years and the numbers of users have increased, because various efficiencies in the networks and data centres have allowed that not to increase wildly.

I was going to say, this is often estimated about 3‑4% of total global carbon footprint which is larger than an entire airline sector now. But, you may be unsurprised, given the previous diagrams, that there's a second level with ICT and digital, so we have the direct effect, and the enabling effects. And, what does an enabling effect do? Well, this comes from the environmental impact when ICT is applied to other sectors of the economy. Unfortunately, here, there are a couple of effects which tend to increase the carbon footprint again which is not great. We have the obsolete effect, devices no longer being supported by software update so we have to chuck them and get a new one. And induction, whereas technology might stimulate the resources, so for example when printers were introduced, they stimulated increasing demand for paper over typewriters.

However, there is some room for optimism here, because there are a couple of enabling effects which perhaps we hope will have the effect of decreasing our overall carbon footprint. So there are a couple basically related to efficiency gains here, within systems, so often it's optimisation effects which you can see there's information about which system is used to improve efficiency. So smart metres in the home or office, for example. Helping people work out how to reduce energy use. Or substitution effects. This is the one I'll focus on briefly, with an example. Where products are replaced by digital equipments.

An example I think may be instructed to look at is one that we compare the impacts of an e‑book with a paper book. Unfortunately, because there aren't that many studies on this kind of stuff out there, but unfortunately in 2017, there was a really good, rigorous in‑depth study done by some academics who looked at Washington state in the North West of the United State and they did the book‑reading research on book‑reading activities. One thing they did is, they calculated the carbon footprint of an e‑book read on a kindle. Estimated at about 1kg of carbon footprint. They looked at reading a single paper book, so the same book as a paper volume, it was about double the carbon footprint. What that study showed us is the substitution effect in action. So a genuine efficiency gain, so that if all we were interested in and focused on was that one book, then we can say yes, this is amazing, digital is empowering, it's great efficiency and reducing the footprint of reading books, which sounds great.

However, there is a level three. Things get juicy at level three where we see systemic effects. What do we mean by that? You can see top right, this is where we are talking about behaviours, citizens or consumers, the human being behaviours and patterns of consumption and the effecting technology has on those. Unfortunately here, a couple of effects that might have a tendency to increase the carbon footprint. There's one that might have societal effects, where for example ICT might enable people to work from home, for example, leaving them to buy bigger houses further from cities, becoming more reliant on cars, so the overall carbon footprint and environmental footprint might go up significantly, potentially.

This one is crucial to understand, and I'll talk about it. What we call rebound effects. Actually it's improved efficiencies of these sorts lead to an overall increase in consumption because patterns of behaviours change. But there are some positive societal effects, so things like, the Internet, being able to share knowledge, work out how they can live lower impact lifestyles. We could point to positive societal effects there. If we come back to the paper book vs e‑book, we can have a look on level three. What behaviours do we actually see? The great thing about this study is it looks at behaviours over time. I think it was over a year. So, yes, I think the annual impact that this group that they studied had before they owned the kindle, so when they were just reading paper books, it's about, whatever that is, a bit less than 10kg of CO2. And then they looked at actually what happened when that same group the next year then adopted a kindle and started reading e‑books. We might hope if we saw this substitution, if they substituted those however many books those are, paper book, e‑books, that we'd see the carbon forgot print halved. Unfortunately, what they saw, which is pretty common, is actually, there is no clear substitution in effect. They saw that that group from they were studying, they bought a few less paper books, but actually, they were still buying paper books and their consumption of books is then on top. So it's in addition, so there's no overall decrease in consumption. So that's the rebound effect in action but actually, what we can point to as efficiency gains are not always borne out when we look at behaviours over time. Once we start to look, unfortunately we see rebound effects more or less everywhere. Another example, what about the impact of a digital photo taken on a smartphone, vs a photo developed on a 35mm film? I don't have any numbers but I'm assuming the impact of one digital photo would be significantly less. What we have got is some estimation of how many photos are now taken globally. So you can see it's relatively increased up to the year 2000 when digital photography comes in, and then particularly since about 2008, the iPhone, there's this massive exponential explosion in the number of photos taken. That has big environmental implications. So there's a study done in the UK by Tull Institution of Engineering Technology a few years ago and they looked at the carbon footprint on photos that many of us take and leave in the cloud, so unwanted photos. So they estimated about the average person in 2020 took about 900 digital photos a year. And only looking at the duplicated unwanted images in the cloud, could they account for about 10kg of carbon dioxide emissions for every adult in the UK in a year which is fairly scary number. And to put it in context, just because it's fun to be scared, right, there's a relatively recent study that tries to make carbon dioxide tangible in terms of the amount of glacial ice that would be melted globally, including Poles and the Himalayas and placings like that. So 10kg of carbon dioxide per adult would ultimately melt 6.9 tonnes of glacial ice through the life cycle of that carbon dioxide that it has in the atmosphere. Given parts of the world, many people are dependent on glacialising melt for rivers and growing food and all of that, that is a bit scary.

We might say, wait, can't the whole system be powered by renewables? Unfortunately, if we look globally, at the planet as a whole, the answer is no for now. While there is relatively rapid expansion of wind and hydro, we are not seeing that substituting for fossil fuel and other means of energy production as of yet. It's just the extra demand which is being effectively provided for by renewable energy.

So no, unfortunately. But what about AI? Is that going to save us? There are plenty of headlines suggesting it might do. This is a random selection that I saw the other day. There's a study by Microsoft and PwC talking about the economic growth and the greenhouse gas reductions and growth in jobs that come from AI potentially, or could do, and that it could cut emissions to help us realise the sustainable development goals in all sorts of various ways. On the flip side, of course, there's also a lot of doom and gloom. So particularly when it comes to the effect on data centres and the huge expansion in the number of servers and data centres we might need and the effect that would have on water consumption potentially and on the grid. So demand for electricity. So just a few weeks ago, 26th March, we had the Chief Executive for the National Grid in the UK talking about the strain that the National Grid foresees in the UK. The boom in AI and quantum computing meaning potentially in the way he sees it, a six‑fold increase potentially in demand from data centres. Actually when we look at the grid as a whole, we can see there's potentially real competition between, for example, charging electric cars. If more of us are going to have electric cars and there's going to be a spike in demand and the data centres and AI are doing the same. That is a real problem and the grid won't be able to cope, potentially.

That is all a bit worrying. And, yes, the brother of Tim, Mike Berners‑Lee does academic work in Lancaster university, his perspective's really interesting to. Be honest, when I see this stuff, I don't know what to do with it. I am just a designer. I say that, but that's nonsense. I'm a designer, human‑centred designer, interested in people more than anything, and when I start to try to delve into some of this stuff, I get a bit lost. Sometimes. Mike sums it up nicely,  we have got all the new technologies, blockchains and Internet things, artificial intelligence, and there are opportunities for these efficiency gains that we saw, substitutions and optimisations. Overall, he says there is no evidence to suggest that the savings outweigh the additional emissions of the technologies. Where does that leave us?

There is a philosopher who says efficiently without sufficiency is lost. And I think honestly, we as a society, are going to have to pretty soon are going to have to butt up against this question. The UN is started talking about it. In 2022, they started talking about sufficiency. Efficiency is really about, how do we, if we are going to deliver design and deliver something, people are going to use it, how can we reduce the resource consumption of that thing? Which could be a service or a chair.


Sufficiency is really connecting us back to questions of genuine sustainability to those sustainable development goals. As the UN finds it, this is about ensuring consumption of energy, land, water, and other natural resources, delivers genuine living standards while staying within planetary boundaries. To realise this, we are going to need some kind of legislation, some limits to consumption. But unless we consider that question, it's fair to say that efficiency on its own is not going to do what we need from an environmental perspective.

So very last slide in this section ‑ there's a really interesting study that was published in 2023 in France ‑ this looks at the nation of France specifically. And it gives a sense of where sufficiency, but also sustainable design might get us. So they put a real emphasis on the power of design to kind of impact the potential future carbon footprint of technology, of the digital world. So they have got a figure for 2020 in France, 17.2 mega-tonnes of CO2, then they have the scenarios that project to 2030. At the end, they have business as usual, pretty much carry on as we are, and they predict, excluding AI actually, so they do one where they include AI and the numbers are terrifying, but here they estimated, if we carry on without AI, 45% increase. If France carried on as it was, that is. Then we have sustainable design coming into the picture. Moderate sustainable design, maybe about 20%. Widespread sustainable design, 5% increase. But it's only if we introduce a bit of sobriety. And the thing they suggest will have the biggest impact is actually limiting the number of end‑user devices, so the number of new phones and laptops that can be sold in France, because actually, those, the life cycle impacts the end‑user devices, all of the mining and manufacturing actually at this point have a much bigger impact than those data centres overall.

When they talk about sustainable design, they talk about systematically incorporating environmental considerations in the design and development products and services. It's exactly what I think we need. They give a few examples. They talk about extending the life of devices, optimising video streams, and optimising website and Digital Services, coding and data traffic management, improving network hardware and energy performance. That's not conclusive but that's just what they are pointing to as areas where there can be a genuine impact through the way we design.

Going to go straight on to section 2. Where might services and environmental impact hotspots likely be? I think it can probably break down into three sections, or two in the main, then come back to the policy or strategy piece. So what about when a service is live and being used by people? I have thought about the possible user interactions here. So some actions users might take. Maybe they want to view a web page and they are looking at it through a desktop, reading email on mobile phone, calling up with a question, again on their mobile, or perhaps there's a real world interaction so.  In this case, it could be maybe the users has to drive somewhere for an appointment of some kind. Those are hypothetical potential steps. Once we have got those, we could think about what is the service infrastructure that supports and enables these steps in a service journey? And it's a very much simplified diagram. For these digital interactions, we have a whole system, digital infrastructure, with networks and data centres and that kind of stuff. Then for the real life or if someone calls up with a question, we have got potentially a member of service staff supporting the service once it's live.

So, yes, where do the environmental impacts come in that live service? Well, we can point to where the major ones are likely to be. Firstly, we have got the end‑user devices, consuming energy and there's going to be a carbon footprint. Energy use associated with data transfer one way or another. This gets complicated, but in simple terms, we have got things like the number in weight of web pages, how efficiently they have been built and coded, and we have also got the amount, another example, the amount of user data that might be stored by a service. And all that data has to be transferred between the user device and the data centre. We've also got the energy the networks and data centres use themselves, and the fact that they are always on in general terms. And also water use. Quite important in data centres especially in parts of the world where there's competition for water resources. And then on the right hand side, again, it's a bit obvious really maybe, but when we have support staff supporting the delivery of the service once it's live, we have got energy use and carbon footprint there potentially and perhaps greenhouse gas emissions. Probably ruining the photos. Better if I stand here! We might have direct greenhouse gas emissions, in that case, from the tail pipe of a car.

So that's a sort of high‑level summary of a live service. Just before we move on to the life cycle, it's worth saying that what they call grid emissions factor does matter. We looked at the globe earlier and saw that renewable energy's not yet substituting for fossil fuel generation, but actually when we look at individual countries, it might be. Apologies, this map comes from a fine website which you can play with, and you can see how green the grid is in different countries, it's a bit fun if you are a bit sad like me. But yes, apologies to anyone with any degree of colour blindness, but essentially, it's trying to show that on this particular time I looked 25th February at 8am this year, there's a big disparity in Europe between how green electricity is. We have Poland, naughty, where they have got... sorry, carbon intensity is 676 grammes per kilowatt hour of electricity generated. South central Sweden, where, it was only 13grams. Interestingly, the renewables are not wildly different, but the big difference in south central Sweden, they have got nuclear power, so it's nuclear and renewables. Why does this matter? If our whole system is in France, the end use devices are being used by the data centres and the whole shebang. It's something to bear in mind , because sometimes, we had conversations recently about moving data centres from the UK, not move them, but not use them in the UK and choose ones in France instead, and that could potentially be a quick, easy way to redo us the carbon footprint. Not really sure about whether that will work legally or whether it's scalable, but certainly something that has been pointed to. What about the full life cycle then? If we look across the phases of the life cycle that are up there, five phases, we have considered the use phase and we are talking about greenhouse gas emissions and the users and we can have a look across other phases. If we consider that physical service infrastructure which we saw in the live phase where you have got end user devices, networks, data centres. What we can do if we want to understand the total environmental footprint of a service is work out what our services share of that infrastructure. Effectively, that infrastructure's not likely to be used solely for our service so our service is sharing that with many other services, but we can do a calculation to work that out what the part of this a is attributable to our service. So if we look across the stages from raw material extraction to disposal, this is when we see those other arrays of environmental impacts. I touched on that at the start. But as well as greenhouse gas emissions and water use, we have all this other stuff, resource depletion, acidification, soil erosion, deforestation, et cetera. And then, alongside that, we could also if we want look at service development or management. I think the thing which is effectively, what are the impact of the teams? Us in some cases, designers, developers, managers, who are our impacts of that and then decommissioning? Those may seem small in comparison to the service when it's live but it's often considered in calculations that are currently made.

Then the other thing before I move on to the third section is, a policy or strategy bit maybe, in the private sector. So, this is really again, this question about, if we are asked to design and deliver a thing, which could be a service or a chair, a runway, we can only do so much to reduce the impact of that thing. So it comes back to the efficiency sufficiency thing. This is a new slide that I didn't finish last night, apologies. It looks a bit kind of ugly. I was just thinking about the double diamond and it really, you know, once we are over here in the right and clear about what we are being asked to do, we are being asked to build a shiny Digital Service, or whatever it is we are being asked to do, this is really, I think when we see the efficiency piece, from an environmental perspective, all we can try to do is reduce the consumption of resources within the boundaries of what we can control. Sometimes, it might feel not very much, it feels like decisions might have been made by architecture and suppliers in all sorts of ways, and actually, we can think about user journeys of them. But beyond that, we might feel like we can't effect some of the things. Really, this question about sufficiency, about whether the solution is the right one, has to come up front here, and I suppose Lou Downe talked about the pause, take a pause and reflect on what might be the broader implications of what we are doing societally and going forward over the long‑term. I suppose this is where certainly in Government it feels like it's really important, if we want to be able to point to our services and say this is genuine and sustainable, this is when it needs to happen, I suppose. Before that decision is made about what is the thing, we need to understand that problem from the point of view of sustainability. Now to the more practical bit. What do we immediate to start delivering greener services for real? Again, everything breaks down to threes. I hadn't realised that until now. Section 3 breaks into three again. I think at the very least, we need to have measurement that we can have confidence in that helps us understand what the impacts of our decisions are, we need to have principles or standard we can work towards which give us a sense of good practice. And which services can be assured against in some way. And training and guidance is important. It's all very well having all of that, but if we don't know what to do with it, we are a bit stuck.


Measurement ‑ why is that important? I said it. Someone once said, to measure is to know. So if we are not measuring, we are ultimately just guessing. There is a cool leadership positioning model that the author of this book, which is about carbonising digital, produced, and essentially, four Quadrants, which you can see, Eric here argues that most leaders in companies a sit in the top left hand box, and he talks about high motivation. So in general, the way he sees it, most people, most leaders, do want to do the right thing. They care about sustainability. They are not really able to do much about it because they do not have the data today back up their decision‑making. So where we want to get to he suggests is high. High motivation, high value and data. But we are not there yet. I would say strides are being made. Even in the last months. This is an example from the Civil Service, from someone I know at DwP.


Department for Work and Pensions that is, sorry. So what they have done is had a go at measuring the end‑to‑end carbon footprint of the Personal Independence Payment service, usually known as PIP, the main disability benefit. And you can perhaps see they have got some high level version at the end‑to‑end journey where they have a user accessing the page, making an application for it, being notified in various ways about the back end stuff, the applications process, and then at the end, the user will be notified whether or not it's been successful.

So yes, they kind of looked through this, end‑to‑end and come to 1kg per claim. That might not sound like a lot, but when we are talking about a service that is used by millions of people a year, that figure very quickly adds up. I should say, maybe there is also the fact that they are transparent, so there's a business of missing data there. The real impact might be slightly higher or lower, but yes, just a great can example of some of the innovative work that's being do into try to break this down.


Again, it's just from the carbon forgot print stuff, there is the aspiration to include water and other things. Last week, again in Government, example, there's a cross‑working group which has started to look at trying to provide a standardised way across Government and an action‑focused carbon footprint measurement. There’s a straw‑man proposal for a gold standard tiering system, so they might get rated in that way based on how rigorous and holistic the accounting measurement was. This was actually, there was a vote, and this was voted against, so it's been binned and there's the proposal to have more of a maturity model that points or generates an overall score for a service, and but also always points to continuous improvement. So always points to well, okay, this is good, but what is next, what is next ... that was seen as maybe the problem over gold, silver bronze. A company may get gold and think, awesome, job done. But that's not necessarily the thing. It necessarily didn't feel right to the audience at that time. Measurement is still a challenge in various ways. We are going to have to go a bit quick on the slides. Lack of available data. Some of it is still seen as proprietary data, difficult to get it from certain suppliers, like cloud‑hosting suppliers, for example. Still different methodologies, debate between academics on the right way to look at services, particularly because, as we pointed to earlier, there are the level 1, 2, 3 effects, so it's difficult to calculate those. Very few tools to help make calculations and the baseline is difficult. That means how you benchmark against what exists now. So, again, there's debate about methodologies there. If we want to understand substitutional effects, then that's particularly important. So, just quickly then, last five minutes, principles and standards. Great stuff going on, I think, on this outside government and inside government in the UK. The W3C, the web content accessibility guideline have has the web design sustainability group for about ten years or something. But last August they published a first draft of a set of web sustainability guidelines which is really cool. You can go on the website and have a look through.


They have masses of detail on all the guidelines. I would say, they're a bit more focused on websites than the services as a whole, as I suppose the name implies and it breaks down into four sections. Stuff on user experience design, web development, hosting and infrastructure, and systems and business strategy and product management. So that is really cool. Work has been going on in Government. So this is a slightly shameless plug for some of the stuff I've been doing. There is an overlap with the web sustainability guidelines in that some of the principles, there are ten of them for design and delivery, still very much a draft. Some do overlap. Talking about user journey design, software smart data. We are keen to emphasise the upfront piece. How do we connect to that bigger question of sustainability. So, this is actually developing into a bit of a beast, designed with a green policy. And we have been trying to do it in a collaborative open way, develop the principles, had a bunch of online workshops last year and in March, there was the impact, the transformation agency helped us hosts a day in London and we had about 70 people helping us essentially to work out whether the principles are working and are practical. So we gave them a challenge and see how they could green the service using the principles to help guide them. And just to say, the principles are not just ten high level principles, we have a giant spreadsheet where we have detail in on various aspects, which I won't go into. There are discussions in Government now about how an update standard and guidance can include environmental impacts and we can test out the principles and objectives. This is effectively the end. Training and guidance, I don't have a slide for that, which is a bit disappointing right at the end, but there's all kind of great stuff going on, like the School of Good Services that Lou and Sarah spoke about earlier, which is awesome. Stuff going on in Government to try to work across that. The thing I wanted to leave you with is a mini provocation, which is this question, which is really, you know, can a service be considered good unless we can be confident that it's green or sustainable? I don't know. But I'll leave you with that and say thank you very much. (Applause)