Designing an Overarching Framework to Calculate the "Sustainability Score"
This is the written overarching framework proposed through the collaboration of multiple teams, and includes specific team contributions. It has been adapted from it's original Google Doc format.
Node: Paris (HEC Paris - E42)
Participants:
- Leader:
- Wanying Li (YE on Discord, [email protected]): in charge of creating and designing the basics of the overarching framework. Please feel free to reach out (by email or via Discord) if you have ideas, comments, or anything that you would love to share!
- Programmers for the proof-of-concept tool:
- Antoine Gelloz ([email protected]): responsible for programming the backend of the proof-of-concept tool
- Maxime Organi ([email protected]): responsible for programming frontend (i.e. user input) of the proof-of-concept tool
- Karolina Gorna ([email protected]): responsible for designing the calculation involved in programming for the proof-of-concept tool
- Benjamin Carlier ([email protected]): responsible for implementation of calculation as layout by Karolina for the proof-of-concept tool
- Idea contributors:
- Galaad Preau ([email protected])
- Diego Eguia ([email protected])
- Hugo Hernandez ([email protected]
- Remote (NYC):
- Bryce Wong ([email protected]): overall framework design of potential use cases
This is a proposal to design an overarching framework to calculate ONE final sustainability score. We aim to create a platform -- encompassing many tools and sub-platforms -- that will report ONE final sustainability score, the “supply chain map” of the product, and various relevant indexes to the consumers. The “supply chain map” and the various indexes are used to calculate the sustainability score. The most important thing for the consumers is the sustainability score, but we will also report the other info, consumers can choose to look at them if they wish to learn more details.
The proposal was initially developed by the Paris team during the Open Climate Collabathon, with contributions from Bryce Wong. The Yale and Berlin teams later joined the efforts to implement some of the ideas proposed in this framework.
Two major things that are left to be determined after the collabathon event in Paris:
- 1.Selecting the most important relevant data and various indexes that we need to collect in order to calculate the final sustainability score
- 2.Figuring out a way to calculate the final score with all of the info we decided to collect
The Paris team has listed some ideas in the section titled “Our proposal” and will leave the discussion open to everyone. We encourage collaboration efforts to build upon what we have laid out!
We encourage everyone to add your thoughts and insights into this framework proposal that we have developed so far. Please feel free to edit, comment and start a discussion! We'd love to hear what you have to say! Even if it is just correcting a typo, expanding on parts that are unclear or not well-explained, or adding in research efforts that we are not aware of. We welcome all kinds of inputs from people from all kinds of backgrounds!
We want to create a platform to disclose comprehensive info about the sustainability of a product to help consumers to make an informed decision before making a purchase. More specifically, we want to design a framework to assign a score to measure the environmental sustainability of a product, similar to the idea of Nutri-score https://en.wikipedia.org/wiki/Nutri-score:
We would like to emphasize that this is a global vision and it involves the sharing of data, influencing policy makers and collaboration with existing NGOs to make it work. People can build and improve upon this framework, and create spin-off projects based from ideas proposed in this framework!
Providing a sustainability score and details on supply chain to the customers can encourage consumers to purchase sustainable products, which in turn motivates companies to be transparent and more sustainable. In the proposed Open Climate Accounting System from Martin Wainstein[1], “Climate Action” can be instigated by “Non-State Actors,” such as individuals and corporations. A key focus of this project is to encourage these non-state actors to drive a climate marketplace that is sustainable and environmentally-conscious. The goal for this proposal is to first motivate consumers to adjust their buying habits, thereby influencing corporations to adopt more eco-friendly policies. This is known as consumer activism, which has had a long history in the boycotts from ancient Greece to now[2]. Consumer activism has the potential to hold industries accountable for their environmental impact.
We have identified a few organizations who are trying to measure sustainability from various angles. Some of them focus on the indexes like carbon emission/environmental impact/social impact and some only focus on a very niche market:
- Focus on supply chain measurements for the fashion industry
- Food related
- Interest way of presenting information
- But their info does not include the supply chain
- Company-level data, not product level data
- Focus on Chinese suppliers
- A rare find on supply chain related measurements
- It collects carbon-related data from companies who volunteer to disclose such data.
- The end-user for this data is the investors because investors nowadays care about investing in green projects. But this data is not aimed for the consumers.
- It only discloses company-level data, not product-level data.
We want to go a step further, and propose an overarching framework that would potentially apply to all products and include a detailed perspective on supply chain, which we considered to be an important part of the sustainability measurement but that’s lacking right now.
We propose a framework that assigns a score to measure the environmental sustainability of a product. The idea is to aggregate information on the supply chain of a product (note that this is on the product level, not the company level), and various existing sustainability measures from the public databases; then we want to come up with a way to combine all of those information and compute one final sustainability score. [a][b][c][d]Our platform will report the supply chain map of the product, various relevant indexes and one final sustainability score[e][f].
We want to collect the following info in order to assign a sustainability score:
- 1.A supply chain map of the product
- 2.What is a supply chain?
- 3.A supply chain is the myriad of stakeholders, resources, and processes that connects a company with its suppliers. A supply chain ultimately traces how a product goes from its source to the consumer.[3]
- 4.Why?
- 5.
- 6.
- 7.Image from McKinsey&Company
- 8.Maybe a tree like the pic below to detail the major players in the production and distribution of a product:
- 9.
- 11.Example of a value chain for iPhone
- 1.Another potential diagram to show:
- 2.
- 3.The above stages in a supply chain were identified in an analysis of environmental impacts of German industries.
- 4.A diagram on value chain of a product
- 1.Determinants of production
- 2.Land
- 3.Natural resources
- 4.Labor
- 5.Knowledge
- 6.Skill/unskilled labor
- 7.Capital
- 8.Money
- 9.Equipment
- 10.Innovation and Entrepreneurship
- 11.Money
- 14.Climate impact
- 16.Environment & ecosystem impact
- 17.Global Warming (kg CO2 equivalent)
- 18.Water usage
- 19.Waste created during supply chain
- 20.Durability -- this is a measure about the planned obsolescence of a product
- 21.We want to come up with a way to measure how “durable” a product is because many companies nowadays have what is called planned obsolescence. For example, the lightbulb is planned and designed to have a life of 3 years even though it can easily be designed to last for 10 years. For companies who purposefully decrease the life expectancy of the product, they should receive a low score on this index.
- 22.Could a POS-like/loyalty program opt-in infrastructure furnish empirical lifecycle data? Based on actual, authenticated registration of events/states like “produced”, “purchased”, “re-sold”, “disposed”, “re-cycled”...
- 23.A rating system with consistent baselines for (as actual/empirically determined as possible) longevity. A+ = 20 years usable life, B- = 5 years, D = 1 year, etc. ?
- 24.A product-category, sector, or even product specific “circularity rating”. i.e. How mature are the industrial ecology linkages and processes - that would facilitate continued circulation of the (environmentally relevant) material content of the product?
- 25.Reusability -- this is a measure of the end-of-life of the product. Reusable products should get a higher score than recyclable ones, and disposable ones, much lower than these.
- 26.Responsiveness [z]and Transparency -- this is a measure of how transparent the company is with the public. Companies that respond to inquiries and suggestions from the public should receive higher scores.
- 27.This German analysis[aa] mentioned below also highlighted four key areas of environmental protection at each stage: greenhouse gas emissions, air pollution, water consumption, and land use. Here is an example of how they assessed industries by these metrics:
- 28.
[Just ideas - not actual suggested edits] (Bryce):
- We could use something similar to WWF’s scorecard criteria for supply chain assessment:
- It looks like they score companies on the percentage of total use of palm oil (this scorecard is specific to palm oil, as you know) that is covered by different supply chain options (https://palmoilscorecard.panda.org/2016/check-the-scores/supply-chain)
- So I guess if the company uses a supply source that is more sustainable, they get a higher score
- I think one way to adapt the WWF scorecard is to create sub-indexes
- For example, for waste created during a supply chain, order all the companies that we have data on in terms of how much waste they create, then assign each company a decile ranking (are they in the top 10% - most sustainable - or the next 10%, etc.)
- Then each company can get a certain number of points for each metric - say, 10 points if they’re in the most sustainable 10% for waste created during a supply chain, 1 point if they’re in the bottom 10%
- And then you add up all the points and the higher the points, the higher the index for a company
- I don’t know, that system might be too complex - just putting it down so we can build a better solution off of it
Just a suggestion on this part: we could look into the CITI rankings for the supply chain. They evaluate the supply chain of a great of companies in China, with a quite efficient scoreboard. I’m working on it, but here’s the link http://wwwen.ipe.org.cn/GreenSupplyChain/userguide/CITI%20Evaluation%20Guideline.pdf
Looking into the IPE idea (Bryce):
- It looks like this score is partially based on how much a brand is transparent and how much it pushes its suppliers to be transparent and compliant with regulations (thus, it’s a score about integrity, not just physical environmental impact).
- This might not be a bad idea, especially if we’re thinking of proposing user-inputted and whistleblower-esque techniques
- i.e. ask volunteers to request this info from the brands, and if they don’t respond, that’s data in and of itself
- I’ll add a potential Responsiveness and Transparency score to the above variables for the index
We would like to display the final results (the final score, the supply chain map, the relevant indexes) via the means of:
In reality, the companies themselves know whether they are sustainable or not, and they know that in details, but they won't disclose this info to the public because they care more about the profits, and disclosure of such info would hurt their profit margins.
The major questions we need to address are how do we get the relevant sustainability info about the supply chain of a product?
We come up with several supporting mechanisms for the overarching framework:
- To gather data on the product supply chain:
- People have done extensive research on supply chain of major products (ex. iPhone, diet coke). We would like to collaborate with those researchers to collect data on those major products. To make this work:
- We need to have a system in place to encourage the flow of such information. This can become a spin-off project in collaboration with the teams working on prompts that aim to bring together isolated data from various sustainability research and databases.
- We also need people to aggregate such information and standardize them. Could be another spin-off project.
- For smaller products that don’t have any existing analysis on their supply chain:
- Our platform will show that their supply chain map is “locked” which could result in a lower sustainability score for this product.
- Then we might be able to use company-level supply chain data (if the company-level data exist) to approximate the product-level data.
- One other feature we would like to include is giving the opportunity to the consumers/users to point out the parts of the supply chain they would like to know about. Including a “like” and/or crowdfunding system to finance the necessary research is something we would like to implement.
- If it is really difficult to map out the supply chain and there is no way for us to collect any relevant info, one of the ideas is to provide a platform for “whistleblowers” to submit information anonymously so that we can know the red flags in the company’s supply chain.
- Then, we need a mechanism to validate those anonymous data and make sure they are not false info, and to avoid people who submit data for the purpose of sabotaging other companies.
- Once we start to collect more and more data from the whistleblowers on this platform, we can strive to leverage the power of data and collaborate with NGOs to “pressure” the companies (and manufacturers in the product supply chain) to voluntarily disclose their supply chain information.
- To gather data on various indexes:
- There are several existing databases that reports data on various indexes and measures of sustainability. For example:
- CDP
- ESG ratings
- Above suggestions:
- Something similar to the WWF Scorecard (for palm oil) system
- Something similar to the Chinese IPE system
- The other idea to collect data related to sustainability is to use machine learning measures to translate customer reviews on e-commerce platform to
- Use web crawlers to collect data
- Parse the Amazon reviews with machine learning tool by identifying keywords related to different aspects of sustainability such as “the product lasts…”
- The next stage is to figure out a way to translate those anecdotal details to quantitative values. Maybe we need to create a dictionary to enable such translation?
- A place for “wisdom of the crowd” / crowd-sourced distributed intelligence? Collaborative Filtering (https://en.wikipedia.org/wiki/Collaborative_filtering) - both concerning the selection of / relative trust and validity of data-sources/metrics (as discussed extensively in this doc) - as well as, in the gathering/ordering of the specific data points - such as occurs in cases like folding-at-home sort of distributed solving for valuable results.
- A place for distinguishing the distinct “personas” of users of this tool conceived by the project? Like in wikipedia - lots of people read, few people edit. Could this tool craft views and features suited to: “Trusted Data Contributor”, “Point of Purchase User”, “Registered Subject Matter Expert”, etc. ? This could enhance the useability, adoption and sustainability of the (human/social/institutional/industrial/data) ecosystems supporting the tool itself.[ai][aj][ak]
The following are hypothesized case studies to illustrate how the above sustainability score can be enacted for different products.
- Still need help working on this! But, of course, figuring out the framework is essential to complete the use case. (Bryce)
The Paris team develops a tool to calculate the sustainability factor by using a very simplified approach. It serves as a demonstration of the overarching framework.
One of the major challenges during the development of the tool is that we, as students, do not have free access to public databases. Therefore, we decided to crowdsource the data to estimate the various indexes we mentioned in the section titled “Our proposal.”
The plan is to ask the end users to input a few data such as where did you buy the product, where it is produced, is the material reusable, and what is the quality of the product etc. For example, we can estimate the carbon footprint index by finding the distance between where does the user buy the product and where the product is produced.
After collecting those data and translating them to various indexes, we can calculate the final sustainability score using a weighted average of those various indexes and output this sustainability score to the end user.
Variable | Scale | Conversion (0-5) | Weight in category (%) | Category | Category Weight
(%) |
CSR Hub score | 0-76 | (76-x)*5/76 | 100 | Company | 40-a/3-b/3 |
Distance origin | 0-3 | x*5/3 | 100 | Carbon footptint | 30-a/3-b/3 |
Reusable/Recyclable/None
Quality
Product lifetime (months) | 0-2 0-5 y/(mean category) | x*5/2 - x*5/(max of all) | 40 20 40 | Durability & End of Life | 30-a/3-b/3 |
Energy efficiency (A to F) | 0-5 | - |  | Energy efficiency | a = 10 - |
Use of consumable | 0-2 | x*5/2 |  | Waste | b = 10 - |
CSR Hub Social | 0-76 | (76-x)*5/76 | 100 | Social | - |
The CSR Hub Social is not taken into account for the calculation of the score, it is an indicator for the consumer.
The equation to calculate the final score with this method is established as follows :
Parameters used :
- Distance is calculated on a scale from 0 to 3 based on the graph below
Here is a screen shot of the user interface which collects basic product data:
- The Paris Team
- Leader
- Wanying Li (YE on Discord, [email protected]): in charge of creating and designing the basics of the overarching framework. Please feel free to reach out (by email or via Discord) if you have ideas, comments, or anything that you would love to share!
- Programmers for the proof-of-concept tool:
- Antoine Gelloz ([email protected]): responsible for programming the backend of the proof-of-concept tool
- Maxime Organi ([email protected]): responsible for programming frontend (i.e. user input) of the proof-of-concept tool
- Karolina Gorna ([email protected]): responsible for designing the calculation involved in programming for the proof-of-concept tool
- Benjamin Carlier ([email protected]): responsible for implementation of calculation as layout by Karolina for the proof-of-concept toolIdea contributors:
- Remote
- (NYC) Bryce Wong ([email protected]): contributor to the overall design of the framework, as well as potential use cases
- Cameron Sajedi ([email protected]): browser plugin prototyping
- The Yale Team
- Chintan
- James Gong ([email protected])
- Miriam Huerta (rojoredapple on Discord, [email protected]): responsible for programming the browser extension plug-in
- The Berlin Team (idea and task submitters @CoMakery but parallel working force to this one)
- Gürkem (Gurkem), UX/UI/Mockups
- Lefteris (lefterav), Backend/Databases
- Josh (joshuaoverbuy), Front-End/Browser-Extension
- Alex (Alex.Han), Research
- Dominik, Supply Chain Advisor
The Paris team would like to thank the climate policy specialist Darius Seiller from The SASI Co. for helping us to have a better understanding of climate policy and to help us consolidate and develop our ideas!
We also want to thank the student association HEC dataminds for hosting this event at HEC Paris. They create a comfortable and very supportive environment for the Paris team to collaborate on this project over the weekend!
At last, we want to thank the Yale OpenLab for initiating the Collabathon event for everyone around the globe to collaborate on solving climate issues and making the world a better place!
[5] Jungmichel, Norbert, Christina Schampel and Daniel Weiss (2017): Atlas on Environmental Impacts - Supply Chains – Environmental Impacts and Hot Spots in the Supply Chain. Berlin/Hamburg: adelphi/Systain.
[a]As the goal is consumer awareness and a change of consumption behavior, the Berlin-team feels it would make sense to add educational information (e.g. if the product is a bottle of coca cola, then a video on how a glass bottle is being produced). Especially in e-commerce, this is easy to integrate.
[b]hmm this is an interesting idea! does the berlin team have a vision for what that looks like? do we compile a bunch of educational videos in our database as well, and link that to the product online (by matching keywords on the webpage and in our database?)
[c]As we think of the product as an extension/widget for your browser (or SDK for a shopping app like amazon) that uses a crawler to detect what product you're currently looking at, you can use the identified keywords to make a standard youtube search "how is a ___ produced" (inserting the identified keywords by the crawler) and then choose the first result of youtube's suggestions. With some more time you can add a simple AI that scans e.g. the top10 results and chooses the most fitting video
[d]ahh very interesting! i would be very grateful if you found a good place to write this up in this proposal!
[e]Any suggestions on how to combine any of these measurements (various indexes) and supply chain information to assign a final sustainability score? How can we deal with the categorical variables and non-categorical ones in this case?
[f]this isn't quite as good for use in a statistical model, but perhaps we can use a scoring system similar to WWF's palm oil scorecard? https://palmoilscorecard.panda.org/2016/methodology
[k]Can you elaborate on how does knowing the value chain help us understand the sustainability of a product?
[l]Not sure why this is here? How does it help to calculate the final sustainability score? Can someone explain a bit?
[o]going to move this above the "various indexes" subsection under the overall "what info do we want to collection section"!
[p]It's just an idea. Maybe it's better to translate sth. that customers are easier to understand and visualize.
[q]Don't forget this is a consumer-facing information/application: Thus, I feel like you although all ideas really are great, that we should not be overwhelming them with information. I think an index (1-10) (aggregating all social and environemntal effects into one index) and one or two more information such as CO2-emission (kg) and company's sustainability (IPE) score.
[r]I totally agree! That's why the those are just data we were thinking about gathering to render the final sustainability score (computed by taking into the consideration of various indexes and the supply chain map) -- which is the most important info that a consumer will see. Then they can look at the details in the supply chain map and the exact values of various indexes if they want.
So you think the most important and relevant indexes are social/environment impact, CO2 emission and IPE?
[s]I think we should design an output that's more understandable for customers and easy to read. For example "you saved xxx amount of trees today!". A research shows that customers will be more motivated with positive signs they are getting, so we don't want to "scare" them with a bunch of numbers.
[t]So then the idea would be to translate the sustainability score into something like # of trees saved?
[w]That's true, what do you think would be a reasonable way to combine them? I know that people are already measuring carbon emission in quantifiable way, but it is difficult to do for the environmental impact. Maybe there are more research in this area that I am not aware of yet, would love to hear what you think!
[x]We could have a "Climate impact" on one side, and "Biodiversity/ecosystem" impact on the other. It's true that climate does impact biodiversity in the long run, but I think the second score would be used to assess the short and medium-term effects a company/product would have on the environment (like chemical dumping, etc)
[ae]Can someone help doing research on the public databases? And what sort of info can we extract from them?
[aj]That would be relevant to the first bullet - though the collaborative filtering idea is very versatile - and can be applied in a variety of ways - the suggestion here is intended to point toward applying it to filter for the more credible, reliable, usable/used, metrics and data.
[ak]The second bullet it quite distinct idea. Having to do with the Human Factors of the User Experience itself - suggesting some different views/perspectives for different usage contexts
[al]i started a new document to outline these use cases! please contribute to that document as well!