← All guides
Methods & standards • Data • Data quality

Primary data share: primary vs secondary data in a product footprint

By • • Updated • 9 min read

The primary data share is the percentage of a product carbon footprint that is calculated from primary data: measured activity paired with a company-specific emission factor. Two footprints can show the same kg CO₂e and still tell a buyer very different things. One describes your factory; the other describes an industry average. This guide explains how the share is calculated, why metered electricity can still count as zero, and how to raise yours.

  1. Your activity × your supplier's factorPrimary data
  2. Your activity × a database factorYour own quantities
  3. Typical activity × a database factorSecondary data
  4. Category defaultScreening proxy
BestData qualityWeakest
From your own measured activity and your supplier's own factor, down to a database default.

What is the primary data share?

The primary data share (PDS) is the part of a footprint's emissions that comes from primary data, as a percentage. The PACT Methodology v3.0 requires it in every footprint exchanged under PACT.

PACT calculates it on the footprint excluding biogenic CO₂ uptake. Each component's emissions are weighted by that component's own share, then added up.

If you do not know the share for a component, PACT says to score it 0%. So a footprint built only on database values reports a PDS close to zero, and says so honestly.

The share does not say whether the footprint is high or low. It says how much of the figure describes your own processes and suppliers, rather than an industry average.

0%PDS for any component whose share is unknown
3data quality indicators PACT scores
1 to 5rating scale, 1 best
31 Dec 2027date by which PACT requires the ratings
PACT Methodology v3.0, sections 4.2.1 to 4.2.3.

Primary vs secondary data: what ISO 14067, the GHG Protocol and PACT say

Primary data is measured, or calculated from measurements. Secondary data is everything else: database values, literature, industry averages, proxies. The three main rule sets agree on that, but differ on one important detail.

ISO 14067 defines primary data as values from direct measurement or calculation based on it. Site-specific data is the subset measured inside the product system itself.

The GHG Protocol Product Standard classifies primary data by the activity. It says the source of the emission factor has no bearing on whether the data counts as primary.

PACT is stricter. For an input to count as primary, both the activity data and the emission factor must meet its primary definitions.

So the same footprint can show a higher primary share under the GHG Protocol view than under PACT. If a buyer asks for a PACT file, use PACT's stricter test, and say so in the report.

Both standards set a floor for your own sites. ISO 14067 says site-specific data shall be collected for processes under your financial or operational control.

It adds that site-specific data should also be used for the most important processes outside your control. A note defines these as the processes that together make up at least 80% of the footprint, largest first.

Secondary data, it says, may be used only where primary data cannot practically be collected, or for minor processes. The GHG Protocol Product Standard likewise says companies shall collect primary data for all processes they own or control.

These rules apply whatever your buyer asks. So even before a PACT request arrives, your own energy and material records are the data a footprint is expected to rest on.

Counts as primaryCounts as secondary
ISO 14067Direct measurement, or calculation from itData that do not meet the primary definition
GHG Protocol Product StandardProcess-specific activity or direct emissions; factor source does not matterAverages, proxies, all spend data
PACT v3.0 (for PDS)Company-specific activity AND a company-owned or supplier-specific factorDefault factors, regional averages, EEIO, proxies
Paraphrased from ISO 14067:2018 (3.1.6), the GHG Protocol Product Standard (8.3.5–8.3.6) and PACT Methodology v3.0 (Table 9, 4.2.2).
  • Your own factory: site-specific data is expected, not optional.
  • Your biggest suppliers: aim for their own data on the processes that make up most of the footprint.
  • Small inputs: database values are acceptable, with references in the report.

How PACT calculates the primary data share

PACT's formula is a weighted average. For each component, multiply its absolute emissions by its own primary data share. Add those up, then divide by the total absolute emissions.

"Absolute" means removals are added rather than subtracted, so negative values cannot distort the share. Components are things like each purchased material, each energy input and your direct emissions.

For the upstream part to score above zero, PACT says you must ask suppliers for their footprints. You also need each supplier's own primary data share.

  • Step 1: list every component of the footprint with its kg CO₂e.
  • Step 2: give each a share. Use 100% if both activity and factor are primary, 0% if either is secondary, or the supplier's reported share.
  • Step 3: multiply each component's emissions by its share.
  • Step 4: divide the sum by total emissions to get the primary data share.

Worked example: the PACT calculation in numbers

PACT's own Figure 20 works through four components. The numbers below are taken from it.

Component A is 1,000 kg of material at 0.8 kg CO₂e per kg, so 800 kg CO₂e. Its supplier reported a 36% share. Component B is 2,000 kg at 1.2, so 2,400 kg CO₂e, with a secondary activity figure: 0%.

Component C is 500 kWh of electricity at 0.18, so 90 kg CO₂e, scored 0%. Direct emissions from fuel are 1,000 kg CO₂e, all primary: 100%.

The total is 4,290 kg CO₂e. The primary part is 800 × 36% + 1,000 × 100% = 1,288 kg. So the product's primary data share is 1,288 ÷ 4,290, or 30%.

Now map it to a garment maker. A could be fabric from a mill that shares its footprint. B could be trims bought by weight and costed from a database. C is metered electricity, and the direct line is your boiler fuel.

PACT Methodology v3.0, Figure 20. Total 4,290 kg CO₂e; primary data share 30%.

Why metered electricity can still score 0%

In PACT's example, the electricity has a primary activity figure, read from a meter, but a secondary emission factor. Because both halves must be primary, the line scores 0%.

That surprises many factories. Your electricity bill is real, measured data, and under the GHG Protocol Product Standard it would count as primary. Under PACT's share it does not, unless the factor is also specific to you.

PACT's data hierarchy shows the way up. For purchased electricity, its best case is a supplier-specific factor or one tied to certificates such as guarantees of origin. Either must meet the Scope 2 quality criteria.

Need a footprint for a buyer?ISO 14067 method, PACT v3 file and a PDF report.
Get free estimate →

Data quality ratings: the second number buyers read

Alongside the share, PACT asks for data quality ratings (DQRs). They score how well the data fit your actual technology, geography and time period.

Each of the three indicators is scored from 1, the best, to 5, the worst, and reported separately. PACT v3.0 requires the ratings by the end of 2027; until then they are recommended.

If a supplier's ratings are unknown, PACT says to apply the worst case: 0% share and a rating of 5. ISO 14067 also asks that data quality is described, covering points such as time, geography, technology, precision and completeness.

The share and the ratings answer different questions. The share says how much of the footprint is about you. The ratings say how closely the data match your real process, place and year, whatever their source.

  1. Now
    Primary data share requiredEvery footprint exchanged under PACT v3.0
  2. Now
    Data quality ratings recommendedThree indicators, 1 to 5
  3. 31 Dec 2027
    Data quality ratings requiredReported separately for each indicator
PACT Methodology v3.0, section 4.2 and footnotes 73–75.

How to raise your primary data share

Work down the footprint from the largest line. Your own factory and your main material suppliers usually come first, because those are the inputs you can measure or ask about most easily.

  1. 1
    Measure direct emissionsFuel for boilers and generators, with factors from fuel chemistry: 100% primary.
  2. 2
    Fix electricityUse a supplier-specific factor or qualifying certificates, not only the grid average.
  3. 3
    Ask the biggest suppliersRequest their footprint per kg and its primary data share, ideally as a PACT file.
  4. 4
    Record real weightsUse cut weights and actual losses from your own records.
  5. 5
    Repeat yearlyKeep data recent so the time rating holds, and ask again.
Order of work for a garment, footwear or packaging supplier, following PACT's data hierarchy.

A garment example: where your primary data share comes from

Picture an Indian knitwear factory making its first footprint. The factory's own steps and its fabric supplier carry most of the emissions, so they decide the share.

Its boiler burns fuel it buys and records. With a factor from the fuel's chemistry, that line is fully primary. Its electricity is metered but priced with the grid average, so under PACT it scores 0% until the factor changes.

The fabric comes from a mill that sends a footprint with its own share. Whatever share the mill reports passes through, weighted by the fabric's emissions. Trims, hangtags and cartons come from database values at 0%.

So the factory's primary data share rests mostly on two decisions. Will the mill share its data, and can the electricity factor be made specific? Those two steps usually move the share more than all the small lines together.

The same logic holds for a shoe factory, where soles and uppers dominate, and for a box maker, where board does. Find the one or two suppliers behind most of the emissions, and start there.

What buyers check, and common mistakes

A buyer reading your footprint looks at the share, the ratings and whether they match the story in your report. A low share is acceptable if it is stated plainly.

Most errors come from labelling data by where it was typed in, not where it came from.

Another frequent slip is mixing up the share and the ratings. A database value can be recent and local and still score poorly on technology, because it does not describe your own machines.

Before you send a footprint
  • Primary data share stated as a percentage
  • Unknown supplier shares set to 0%, not guessed
  • Electricity factor type stated: grid average or supplier-specific
  • Data quality ratings for technology, geography and time
  • Reference period of the activity data
  • Sources listed for every secondary value
  • A short plan for what you will measure next
  • "We typed it in, so it is primary." A database value copied by hand is still secondary.
  • "Our meter makes electricity primary." Under PACT's share, only if the factor is specific too.
  • "Supplier data is always primary." Only the part the supplier measured; ask for its own share.
  • "A high share means a low footprint." The share shows where data came from, not how big the result is.
  • "Using the product's weight as activity data." The GHG Protocol says the finished product's mass on its own is not primary data.

What to do next

Start with what you have, state the share honestly, and add primary data year by year. A footprint with a low share that says so is more useful to a buyer than one that hides it.

A low share is normal in a first footprint, and buyers know it. What they need is a clear statement of it and a plan to raise it.

State the percentage beside the footprint, list which components are primary, and give the reference period of your activity data. PACT encourages a short explanation of the share, so add one line on what drives it.

Then say what you will measure next and when. A buyer can plan around a 10% share that will rise next year. They cannot plan around a share that was guessed.

Sources and further reading

The standards and methods this guide draws on, at their publishers:

Questions

What is the primary data share in a PCF?

The percentage of a product carbon footprint's emissions calculated from primary data. PACT v3.0 requires it for every footprint exchanged under PACT.

How is the primary data share calculated?

Each component's absolute emissions are multiplied by its own share, summed, and divided by the total absolute emissions, excluding biogenic CO₂ uptake.

Is my electricity bill primary data?

The kWh are measured, so the GHG Protocol treats them as primary. In PACT's share the line counts only if the emission factor is also specific, not a grid average.

What share should I use if my supplier did not give one?

PACT says to apply 0% to that component, and a data quality rating of 5 where ratings are unknown.

Do data quality ratings use 1 or 5 as best?

In PACT v3.0, 1 is the best score and 5 the worst, for each of three indicators.

Is a footprint with a low primary data share useless?

No. It still shows where emissions sit. It should state its share so the buyer can read it in that light.

Which supplier data should I ask for first?

Footprints for the materials that make up most of your product's emissions, usually fabric, leather or board, from your largest suppliers.

Does ISO 14067 require a primary data share?

No. ISO 14067 asks for site-specific data where you have control and a description of data quality. The primary data share as a percentage comes from PACT.

Footprints to send your buyersISO 14067 method, a PACT v3 file and a PDF report with every source.
Get free estimate →