Following on from our first environmental impact study last year, we entered 2020 with optimistic plans to reduce our carbon footprint. It turns out that nature had some plans of its own.
Wildbit has been a remote company for 20 years, but the COVID-19 pandemic still impacted how we operate. While our day-to-day operations stayed mostly the same, travel plummeted this year as we canceled our annual team retreat and conferences moved online.
For 2020, we again focused on measuring our three main areas of emissions: infrastructure, travel, and operations. We have a little more experience measuring our emissions this year, but there are still some significant difficulties in calculating the impact of a remote SaaS business.
Infrastructure emissions
Our engineering team continued to move services from our colocated data center to AWS in 2020. This has helped reduce our estimated infrastructure footprint from 61.4 tCO2e in 2019 to 58.4 tCO2e in 2020.
tCO2e means metric tonnes of Carbon Dioxide Equivalent. This unit incorporates the global warming potential of a range of polluting gases and converts the amount of each gas into the equivalent amount of Carbon Dioxide, giving us a single figure to track emissions by.
Cloud providers like AWS operate data centers that run far more efficiently than most colocated operations. However, we lose the ability to monitor the energy consumption of our hardware with these providers, making it incredibly difficult to report emissions caused by our usage. We haven’t found a way to estimate emissions from our AWS usage yet. For Google Cloud, we’ve applied research done by the team at Etsy to estimate emissions from our usage and billing data.
| 2019 (tCO2e) | 2020 (tCO2e) | |
| Co-located data center | 60.2 | 57.8 |
| Google Cloud | N/A | 0.001* |
| AWS | N/A | N/A |
| Other Marketing websites estimated emissions based on page views | 1.2 | 0.6 |
| Totals | 61.4 | 58.4 |
* Google cloud runs on 100% renewable energy, so we’re only accounting for emissions associated with transmission and distribution losses.
Travel emissions
After starting the year with exciting plans for a blowout team retreat to celebrate Wildbit’s 20th birthday, COVID-19 brought all of our plans to an abrupt halt. With the exception of a leads team retreat in early January, we canceled almost all our travel plans in 2020.
Our overall travel footprint in 2020 was 3.4 tCO2e compared to 20.5 tCO2e in 2019, an 83% reduction. We’ve made the call not to hold an annual retreat in 2021 either, so we expect this number to remain low in 2021.
Flights
With no retreat, we made just 19 flights in 2020. Traveling a total of 17,936 miles compared to the 100,148 miles we flew in 2019. Our carbon footprint went from 16.2 tCO2e in 2019 to 2.8 tCO2e in 2020, a huge 83% reduction.
Ground transport
Taxis, trains, and bus travel accounted for 530 miles and 0.1 tCO2e in 2020 compared to 1,376 miles and 0.5 tCO2e in 2019. This covers travel associated with company trips and conferences. Some team members use local co-working spaces but we don't currently track their commutes.
The percentage of ground travel using public transport increased from 13% in 2019 to 45% in 2020. Pre-pandemic we had set out to encourage the use of public transport over Uber and private taxis to help reduce our travel footprint. Given the low level of travel in 2020, it’s difficult to say whether those efforts actually contributed to the increase in public transport usage or not. It’s something we want to continue working on once travel increases again though.
Hotel stays
With the cancellation of our annual retreat and in-person conferences, hotel stays were drastically reduced. Going from 184 nights in 2019 to 21 nights in 2020. Emission associated with hotel stays reduced from 3.7 tCO2e to 0.5 tCO2e.
| 2019 | 2020 | |||
| Miles / Nights | Emissions (tCO2e) | Miles / Nights | Emissions (tCO2e) | |
| Air Transport | 100,148 | 16.210 | 17,936 | 2.766 |
| Private Transport Ubers, taxis, private-hire buses, etc. | 1,194 | 0.525 | 290 | 0.098 |
| Public Transport Trains, trams, buses, etc. | 182 | 0.026 | 240 | 0.039 |
| Hotel Stays | 184 | 3.735 | 21 | 0.456 |
| Total | 101,524 / 184 | 20.5 | 18,466 / 21 | 3.4 |
Operations emissions
In 2019 we decided to downsize our HQ in Philadelphia and move to a dedicated area in a co-working space. That worked out pretty well for a while, but once the pandemic hit and we were all bound to our homes the office didn’t get much use.
As the co-working space has shared utility meters, we lost our ability to accurately measure our electricity and gas usage after the move. Instead, we estimated our usage based on figures released by the US Energy Information Administration (electricity data, gas data). Based on these estimates, our electricity footprint came in at 1 tCO2e and our gas footprint at 2.3 tCO2e in 2020.
That’s not the full picture though. The tenants we sub-let our old HQ to are in the travel industry and were badly impacted when the pandemic hit. They moved out in June and we took on responsibility for the building again. Accounting for the electricity and gas usage to keep the old office ticking over while it was unoccupied pushes the footprint of our total operations for 2020 up to 14.2 tCO2e, an increase of ~20%.
| 2019 (tCO2e) | 2020 (tCO2e) | |
| Philadelphia HQ | 12 | 10.9 |
| Dedicated co-working space | N/A | 3.3 |
| Total | 12 | 14.2 |
We’ve since decided to go fully remote and vacated the co-working space when our lease expired in March 2021.
Wildbit’s 2020 footprint
Looking back on an unusual year, our 2020 estimate footprint came out at 76 tCO2e. 20% lower than our footprint in 2019. That’s about the same as the emissions from burning a tanker full of gasoline.
| 2019 | 2020 | |
| Infrastructure | 61.4 | 58.4 |
| Travel | 20.5 | 3.4 |
| Operations | 11.9 | 14.2 |
| Total | 93.8 tCO2e | 76 tCO2e |
We worked with Gold Standard again this year to purchase carbon credits from a solar power project in Bhadla, India to offset our footprint. While purchasing carbon offsets is not a long-term solution to the climate crisis, it can help some of the communities most at risk from the effects of climate change.
Our next big challenge is to find a way to estimate the energy usage of our fully remote team and factor that into our footprint. Last year we tried running a survey with the team to collect information about their energy use but found it difficult to get hold of accurate numbers due to the wide range of energy sources and billing intervals used across our team.
We’re currently looking into ways we can estimate energy usage based on location-based averages. We’ll be sure to share our approach once we’ve figured it out. If you’ve tackled this for your own team, we’d love to hear from you. Send us an email or give us a shout on Twitter!