Improving accuracy with user habits

Spend-based analysis cannot tell us exactly what a user purchased at a merchant. However, by collecting information about the user — diet, energy tariff, household size, vehicle type — we can adjust the emission factor and significantly improve accuracy.

This also engages users and gives them a clear path to change habits and reduce their footprint.

Example: dietary preference

Question: Which of these best describes your diet?

Option	Emissions per £100 spend on groceries
High meat diet	60 kg CO2e ⬆️
Average diet (default)	47 kg CO2e
Vegetarian	32 kg CO2e ⬇️

Depending on the user's answer, we adjust the emission factor used for groceries (and other food-related categories). As you can see, the differences can be substantial.

User habits are also used to filter tips. For example, if a user is already a vegetarian, we won't surface tips about eating less meat.

Implementation

Example questions

Below are some example questions you can include in your onboarding flow. The number to ask depends on how engaged your users are. For the full list, speak to your Connect Earth contact.

👩‍👩‍👦 Shared purchases

In addition to yourself, how many people do you typically purchase on behalf of? How many people are in your household?

Response: numerical value.

🍽️ Food / diet

Please describe the dietary preferences of the people you typically purchase for:

• Average diet
• Meatless day
• Semi-vegetarian (eat meat 50% of the time)
• Pescatarian
• Vegetarian
• Vegan

🚗 Travel — fuel type

Which type of car do you own/use the most?

• Petrol vehicle
• Diesel vehicle
• Hybrid vehicle
• Electric vehicle

🔌 Household utilities — electricity

Do you pay for renewable electricity?

• Yes
• No

🌡️ Household utilities — heating or cooling

What type of utilities do you pay for?

• Electricity
• Gas
• Heating oil
• District heating or cooling
• Water

Use cases

Single user with a personal account

The user responds to the questionnaire once and their user habits are used for future calculations.

1. The user is presented the questionnaire and completes the questions.
2. The response is sent to the questionnaire endpoint including the user's ID.
3. Connect Earth saves the responses and uses them to adapt calculations for any future data sent in association with that user ID.

Single user with multiple accounts

The user responds to the questionnaire once and their user habits apply to future calculations independent of which account the transactions are created against.

1. The user is presented the questionnaire and completes the questions.
2. The response is sent to the questionnaire endpoint including the user's ID.
3. Connect Earth saves the responses and uses them to adapt calculations.
4. As long as the correct user ID is included in any future requests, the adapted calculations are returned.

Multiple users with a joint account

1. Each user is presented their own unique habits questionnaire and completes the questions.
2. Responses are sent to the questionnaire endpoint with each user's ID.
3. Connect Earth saves the responses against each user ID.
4. When requesting emissions calculations for two different users on the same account, we use the user habits saved against the given user ID.

ℹ️ In some cases this may result in two users seeing slightly different emissions calculations depending on their unique responses to the habits questionnaire.

Example API calls

Submit answers via the questionnaire endpoint. All future calculations will use the updated emission factors.

{
  "userId": "1234",
  "geo": "BE",
  "num_people_purchasing_for": 2
}

You can also provide user habits for individual transactions. These transactions are then recalculated using new/additional habits — for example, a user might want to specify they took a first-class flight or ate a vegetarian meal.

{
  "transactions": [
    {
      "currencyISO": "GBP",
      "user_habits": {
        "diet": { "vegan": 1 }
      },
      "price": 20.22,
      "mcc": "5411",
      "geo": "GB",
      "transactionDate": "2022-3-02",
      "transactionId": "12345"
    }
  ],
  "userId": "1234"
}

Please contact Connect Earth for a full list of possible user habits.