This pillar of the maturity model covers ....
How we deliver services to users and allow them to personalise their experience provides opportunities to reduce energy impact. For example, does it make sense to deliver an asset to the users device if it is never used? Do we really need autoplay enabled on the video or should we trust the user to click 'play' should they wish to view it?
| Measure | Description | Score: 1 | Score: 3 | Score: 5 |
|---|---|---|---|---|
| Eliminating waste | By definition, removing what is not used makes the solution more efficient e.g. removing unused assets or minifying files. | When we modify an existing service if there is obvious waste or improvements to be made, we'll try to make this without increasing the scope of the change. This happens on a best-effort basis. | As part of our internal processes we have mechanisms to identify waste and raise it for consideration e.g. an unminified file, a sourced library that isn't used, an image which is included but not referenced. We utilise third-party tools to inspect our service and provide advice on energy efficient defaults. | As part of our internal processes we have mechanisms we use to identify waste and raise it for consideration e.g. an unminified file, a sourced library that isn't used, an image which is included but not referenced. We use third-party tools to inspect our service and provide advice on energy efficient defaults. The culture within our team is centred on reducing waste. |
| Optimising media types | How do you trade-off performance, experience and quality? For example, do you make use of high quality video and colour photography, or do you opt for monochrome images and interactive web animations. | We prioritise only the richness of the user experience. We opt for the highest quality images, videos and experiences without considering their impact. | We try to balance richness/quality with user experience pragmatism. We might generate assets in varying levels of quality and choose different defaults relative to the circumstances e.g. lower quality video or images when utilising a mobile phone or adoption of the WEBP image format. | We generate assets in varying levels of quality enabling us to choose different defaults relative to the circumstances e.g. lower quality video or images when using a mobile phone. |
| Adjuting user-experience | The provision of either context driven or user-based controls to alter the behaviour of the service e.g. disabling rich content like autoplay or tweaking the colour scheme. | Users can tweak their experience in basic ways like switching languages, but little focus on energy saving measures. | We optimise the default experience and provide some basic options that could be used to reduce energy usage e.g. changing colour scheme. | We optimise all modes of user engagement and dynamically select the most energy efficient mode relative to how the user is accessing the service e.g. if the user is abroad we may disable videos. |
By designing and optimising for the devices that will consume our service, we can lower the energy consumption and extend the hardware lifecycles. For example, when designing for low-powered or older devices, you might choose to turn off features that may use a lot of power.
| Measure | Description | Score: 1 | Score: 3 | Score: 5 |
|---|---|---|---|---|
| Device Optimisation | The more you can tailor your software to a specific device, the more chances you will have to overcome its limitations and prevent it from becoming obsolete. | We focus on browser compatibility for only the most modern web browsers. We don't worry about devices and use abstractions to avoid this pain. | Beyond testing web browser compatibility, we do test across a range of devices to ensure backwards compatibility. We actively profile how our service is consumed on those devices and optimise for reduced energy consumption. | We have explicitly profiled and optimised for consumption by the most-used devices. We may even link this to the energy efficient defaults mentioned previously e.g. understanding what colour themes work best to save energy on different smartphones relative to display technology. |
| Service delivery model | How do you get your software into the hands of your end users e.g. native app or website. | We do not target specific device capabilities and simply develop generic web-services which we know can be widely consumed across platforms. | Depending on the device, we try to make a native option for a user to consume our service. Doing this allows us to better use device capabilities like energy saving modes and background refreshes. | Where possible, we provide native mechanisms to consume our service so we can use the native capabilities of devices e.g. making use of GPU acceleration where it will reduce processing time and energy consumption. |
Users can also help by using software in an environmentally friendly way. You can teach them how to do this and reward them for their actions. For example, you can ask them if they need to run their task right away with a high carbon footprint of x CO2eq or if they can wait for three hours when the carbon impact will be lower at y CO2eq.
| Measure | Description | Score: 1 | Score: 3 | Score: 5 |
|---|---|---|---|---|
| Carbon Impact | How do you represent the carbon impact of your software to end users using ratings like the Software Carbon Intensity (SCI) score or carbon equivalences? | We don't do anything explicit. | We have calculated the Software Carbon Intensity for delivering a functional unit within the service. | We have calculated the Software Carbon Intensity for delivering a functional unit within the service and looked at the end-user impact of consuming the service end-to-end. |
| Carbon Awareness | How do you try to change user behaviour with this enhanced awareness of impact? | We don't do anything explicit. | Having calculated the carbon impact, we provide carbon equivalences which users can better understand as well as guidance on how to mitigate some of the impact. | Having calculated the end-to-end carbon impact, we provide carbon equivalences which users can better understand as well as guidance on how to mitigate some of the impact. In some scenarios we recommend user-actions which can minimise carbon impact. |