Manage flagging expenses as food data duplicates

All of your purchases will appear in your accounting file, so food emissions are by default calculated using a monetary approach → If you fill the food module, food emissions will be accounted for twice (both monetary and physical approach), it means you must remove the monetary impact to avoid double counting.

Manage quality checks

You are using the advanced module, hence manually processing your client’s data

Your client is encountering errors when uploading data on the end-to-end module, and asks you to check what is wrong with their file

Product descriptions: The accuracy of emission factors matching by the algorithm highly depends on the quality and preciseness of product descriptions. Here is a list of cases that prevent an efficient identification:

Abreviations (for example, using “chckn” instead of “chicken”)

Only mentionning a product brand instead of the product itself (for example, “Mars” instead of “Chocolate bar”)

Vague product descriptions (for example, “Sweet product”, “Drinks”)

This case does not prevent from conducting the study as average emission factors can be used, but it will reduce the precision of the study

Quantities: in the data collection file, we ask for the unitary weights and quantities per product

Negative quantities are frequent: inventory exports often include purchase cancellations, resulting in negative weights

Negative quantities are not managed by the platform’s algorithm

Total quantities per product (positive weight + negative weight) must be kept

If there are not too many lines, you can make the correction manually to avoid additional days of delay → Otherwise, send the file back to the deal owner for the client to make the necessary corrections

Last resort is to create a pivot table to get total weights per client data combination (product description, product category, supplier)

Sometimes clients don’t understand well and fill the unitary weights in both columns → If the total quantity seems too low to you (less than a ton for example), ask the deal owner to make sure with the client that the quantities are okay (disclaimer: usually they are not in these situations)

On the other end, sometimes quantities seem disproportionate (”Quantities” column filled twice) → same process

Units: everything should be in KG, or a weight equivalent

Do not accept other units, if the situation happens → set the status of the task as “Missing data/Info from client” and warn the deal owner

Exception with volumes in L → Sometimes when data collection is complicated, you can agree with the deal owner that 1L = 1KG (not perfect, for example this wouldn’t be true for thick cream)

Non-food products: very often you will find non-food products in the client’s raw data (usually it’s related to packaging, tableware or hygiene products)

Don’t process this data, but extract it and transfer it to the deal owner so they decide if it should be studied in a product purchase module

Missing suppliers: not a blocker, but when the client didn’t fill the “Suppliers” column, make sure with the deal owner that they won’t be able to provide it

Blocker if the client wants to submit SBTI targets

Manage flagging expenses as food data duplicates

Manage quality checks

Go on the module from Admeenly and click on the dropdown under “Change exclusions”, on the right of the screen
​

Once you have selected the most suited option, click on the green checkbox next to the dropdown

Make sure you are done categorizing the data, and click on “OK” on the warning message that appears
​

Once the categorization process is finalized from the module, go to Admeenly > Activity data > go to the food module

Make sure ids have been created for all your custom emission factors

With the search bar at the top right of the table, filter your data to view the needed lines

Select them and click on “Edit EFs”

On the EF manager page that appears, select “Search for an id”
​

In the search bar, paste your EF id

Once you have found your emission factor, click on the arrow on the far right of the line
​

Click on “Save and close” at the bottom of the page

Check that the EF have been well updated when going back to your module

After you validated the client’s raw data, send the file to in addition to the company id, so he can run the algorithm and send you back a categorization file like this

Don’t forget to mention all the client’s specifics to Sigurjón: EF exclusions, supplier EFs… So he can adapt the EF database in his file

If food studies have been done in previous years for your client, warn Sigurjón

Products that were also in the previous years’ inventories will be highlighted in green and automatically categorized with the same EF

Add this categorization file in the client’s Drive folder

Everything happens in the “predictions” tab

Column “name” is the unique list of the client’s products, “selected prediction” contains dropdowns with all suggestions from the algorithm

For each line → If the algorithm’s selection is correct, tick the box in “Auto categorization” (it will give Sigurjón an overview of how well his algo worked for this data)

If the categorization is not correct (no correct categories from the dropdown), use the “manual categorization” dropdowns
​

💡 Tips for large files:

Don’t hesitate to use the filter on column “name” if you see that specific products (ex: “almond” and “apple jam” on the screenshot) appear on multiple lines

Don’t hesitate to add a column with the client’s product categories if they are clear enough for you to use them for bulk categorization (ex: “beef”, “chicken”, “chocolate”…)

The same product might not have the same description through years, meaning that it won’t be highlighted in green on the file
👉 Check recurring product names between the previous years and the current one, and use the file’s filter on the “name” column

Add the client’s raw data → All non-mandatory columns will be useful as analysis tags!

⚠️ Quantity column

The platform won’t accept commas (only a point for decimal numbers) → Select “Automatic” as the number format
​

Copy and paste the content of the columns in grey:

Emission factor id

Category level 1, Category level 2, Category level 3 (4 is not necessary)

Make sure that all the grey columns are well filled (no blanks + you might find "!!! MISSING!!!” sometimes, replace it with the right content)

Modify the content in your “Export” tab and not in the grey columns, or you will break the formulas

When cells are empty for category 2 or 3 (when you use average EFs), copy and paste the upper category from the same line

If Sigurjón was not able to integrate Suppliers or custom EFs directly in the file, replace the initial ids for the products in question directly in the “Export” tab

Agribalyse DB

Chemicals DB

Copy and paste the client’s raw data in a processing file that you will have created in the customer’s Drive folder

Create an “Export” tab whose format must follow this template

All client’s non-mandatory data should be added as additional columns in this file → They will be processed as analysis tags

⚠️ Quantity column

The platform won’t accept commas (only a point for decimal numbers) → Select “Automatic” as the number format
​

Copy and paste the relevant emission factor id per product

Use the “Correspondence table” of the “Agribalyse DB” file, or the “Chemicals DB” file for chemical products

Choose the column of ids that corresponds to the right combination of life cycle steps for your client

For organic products, check if an EF is available in this database

For suppliers EFs or custom EFs, create an id from the Emission factor database and copy/paste in on the file for the relevant products

Fill the categories columns

Category level 1 corresponds to column K or N

Category level 2 corresponds to column L or O

Category level 3 corresponds to column M or P

For chemicals from the “Chemicals DB” file

Category level 1: Culinary aids and various ingredients

Category level 2: Chemicals

Category level 3: enter the type of chemical product (acid, vitamin…)

For custom or suppliers EFs, find the categories for similar products

Don’t hesitate to use the filter on column “name” if you see that specific products (ex: “almond” and “apple jam”) appear on multiple lines for bulk categorization

Don’t hesitate to use the client’s product categories if they are clear enough for bulk categorization (ex: “beef”, “chicken”, “chocolate”…)

Some product descriptions might be the same through years

Create a tab and copy/paste the previous years’ processing files

Use the VLOOKUP formula in your current year’s raw data to match a maximum of lines and reduce the manual categorization work
👉 A product’s description might also vary so you don’t have exact matches → Check recurring product names between the previous years and the current one, and use the file’s filter on the “name” column

Unit = “UNIT”

Parent purchase category = “FoodAndDrinks”

Purchase category = “EMPLOYEE_SURVEY_MEALS”
​

Red meat: ef6af9cd-1f53-44a3-be13-61db8fe1c0dc

White meat: 45c9dee0-162d-41b5-9bcd-5e28e1abb617

Fish: 11df2542-5ba1-458c-851f-7dbf32b84726

Vegetarian: 1e270ef0-33ed-4ec4-b70f-6dd8b76f5251

Vegan: 765196c3-a3c0-4f42-bf31-ff1c2dbb9646

You have to estimate the type of product that correspond to the item and do an hypothesis on the unit weight (align with the client or PM), so you can use the Agribalyse database with “kg” as a unit

Create a tab in your processing file and copy paste your current year’s data that has already been categorized

Create a “Processing” tab with a column containing the descriptions of the products that will be rebaselined

Create a “Matching” column, and with a VLOOKUP formula, identify the products from your reference year that are also registered in your current year’s data

For these identified products → apply the current year’s following data: emission factor id, category level 1, category level 2, category level 3

For the rest of the unidentified data, follow the “Ⓜ️ Manual processing” process above
💡 Some product descriptions do not exactly match with the current year data but are similar → Try to harmonize categorization as much as you can.

Negative quantities will generate errors from OneSchema, make sure you don’t have any in your file

The algorithm does not accept number format with thousands separators (ex: 1, 563, 732.45, or 1 563 732, 45) → To make sure that the format is correct:

Remove any space as thousands separators (otherwise the format will be considered as text)

If you see any “,” as a decimal separator, convert them as points

Select your whole column > “123” in the control panel > Select “Automatic” (you should get 1563732.45 for the same example)
​

Tab named export

Mandatory columns: name of product, category, total weight (in kgs), unit of weight (kg), EF id, category levels associated to the specific EF (3 categories per product most of the time, sometimes 4. They are important for visualisation graphs)

Optional columns but mandatory if your client has filled them in: country of origin, organic, label, any other optional column has filled in

“Processing”: calculations from the algorithm are still ongoing

“Completed”: calculations are finalized → Emissions and visuals on the module are final, and you can set the module as “Done” on the Mission Control
​

“Processing error”: click on “See import logs” to check the source of the issue

It’s often due to an emission factor id that is no longer existing (database updates can happen) → You can check by yourself or with the R& M team what the correct id is

For any source that you can’t manage yourself (other than data format), check with the product team by creating a ticket

Once you have corrected the errors, you can go through the import process again until you the “Completed” status
​

SBTI Knowledge

360L training: SBTI general methodology

SBTI for FLAG companies

As well as for any traditional food study you will process a client’s data, and emissions and analysis will be displayed as usual: Total emissions, emissions per life cycle step, emissions per category, etc…

Client must also report per product emissions related to Land Use Change (LUC), and Land Management (LM), in order to set targets specific to their industry

Manage your food study as usual

Once data is integrated in the client’s module, make an R& M ticket so the owner () will run a query to extract the data under the right format.
⚠️ FLAG also includes non-food products (Paper, cardboard, wood and other materials)

Align with the deal owner if a product purchase study has been done and must be integrated in the export too + who is in charge of this export