Twelve questions to ask before an avoided-emissions number goes into an LP report

Short answer. Before a figure goes to LPs, check twelve things: the tier, the functional unit, the incumbent, the baseline over time, the levers that carry the number, the equivalence, the volumes, the boundary, the emission factors, double counting, attribution and, for realised figures, the trail back to real sales. If any answer is "we don't know", the first competent question will find it.

Use this list on any number, ours included. It is free to copy.

The twelve questions

1. Which tier is it, and does it fit the company's stage? A good answer names potential, planned or realised, and the stage matches: potential for seed and Series A, planned only once the company has a realistic business plan, realised only for units actually sold. Project Frame Guidelines 2024, Section 3, pp. 26 to 28.

2. What is the functional unit, service life included, and is it the same on both sides? A good answer states what service, how much, how well and for how long, identical for the solution and the incumbent. ISO 14067 6.3.3 and Annex B; ILCD Provision 6.4 IV, p. 68. See the functional unit.

3. What exactly is displaced, in each market? A good answer names the incumbent per segment, in the right country and with the right use pattern, and weights several incumbents by how much each one actually loses to the solution. Project Frame Section 3, pp. 15 to 16; ISO 14067 Annex B.

4. Does the baseline move over the product's life, on the same pathway for both sides? A good answer shows a sourced, bounded pathway, year by year, applied to the solution and the incumbent alike, with a sensitivity on it. Project Frame Section 3, pp. 21 to 22; Frame grid paper (2025), p. 4. See dynamic baselines.

5. Which three inputs carry the number, and what happens when each one moves? A good answer names them, sources them, and varies each over a documented range rather than a default percentage. Where a single input can flip the result, it shows the break-even value. ISO 14067 6.3.7 and 6.6; ILCD Provision 6.10 XII, pp. 147 to 148.

6. Is the swap one for one, and how do you know? A good answer evidences the ratio between solution and incumbent (thickness, dose, efficiency, service per cycle) and does not leave it silently at one. ISO 14067 6.3.3; ILCD Provision 6.10 VIII, p. 147.

7. Do the volumes start from where the company really is, and stay inside their market? A good answer starts from the units already in service, compares year-one additions with what the company can make today, and keeps the obtainable market inside the addressable one. Project Frame Section 3, p. 15; Section 3, pp. 24 to 25.

8. Does the boundary on the page match what was actually counted? A good answer lists the stages included for each side, every exclusion with its reason, and the cut-off, and the inventory really contains the stages it claims. ISO 14067 6.3.4.1 and 6.3.4.3; ILCD Provisions 6.6 IV and V, p. 106. See system boundary and cut-offs.

9. Are the largest emission factors the right substance, place, year and boundary, and do the results pass a physical check? A good answer sources each factor with its version, uses one set of warming potentials for both sides, and confirms that nothing breaks physics: avoided emissions per unit no larger than what the incumbent emits, energy use no lower than the process allows. ISO 14067 6.3.5, 6.4.3 and Annex B.

10. Is anything counted twice? A good answer shows each unit counted once and no figure summed with another view of the same units. ISO 14067 5.12; WBCSD Guidance on Avoided Emissions v2.0. See attribution and double counting.

11. Is it the solution's number, with any attribution shown separately? A good answer reports the solution's avoided emissions first, and any share claimed for the fund as a separate, labelled value. Project Frame Section 3, pp. 29 to 30 and p. 32.

12. For a realised figure: can it be traced to real sales, on the same method as last year? A good answer ties each year's figure to sales the company confirmed, leaves earlier years as they were, uses the same functional unit and method as the year before, and keeps cumulative results apart from forecasts. Project Frame Section 3, p. 27 and Section 5, p. 41; ISO 14067 6.4.7. See potential, planned, realised.

How Carbbin does it

We wrote these from the questions we ask when we review avoided-emissions reports, our own included. A Carbbin case answers most of them on its own pages: the tier, the functional unit and service life, the incumbent per segment, the pathway, the key assumptions rated with their sources, the range, the boundary and the exclusions, every emission factor with its source, and the line that the figure is the solution's, not the investor's.

Common mistakes

• Answering question 1 last. The tier decides how every other answer should be read.
• Treating a precise number as a strong one. A default plus or minus 20% on an unsourced input makes the weakest number look the most certain.
• A design target or a founder's estimate entered as data, with no range.
• Year-one sales that quietly include units already in the field, so the first year looks bigger than any later one.
• A sensitivity that varies one inventory line instead of the inputs that actually carry the number.
• Only checking the arithmetic. The arithmetic usually holds. What breaks is the framing: the functional unit, the incumbent, the boundary, the time basis.

Sources

• Project Frame. Project Frame Methodology: Evaluating Greenhouse Gas Impact for Early-Stage Investments. December 2024. https://projectframe.how/s/2024-Project-Frame-Methodology.pdf
• Project Frame. Quantifying GHG Emissions Intensity of Electricity Grids: Selecting the Right Metric for Net Unit Impact. September 2025. https://projectframe.how/additional-resources-blog/ghg-emissions-grid-intensity
• ISO 14067:2018. Greenhouse gases: Carbon footprint of products. Requirements and guidelines for quantification. Clauses 5.12, 6.3.3, 6.3.4, 6.3.5, 6.3.7, 6.4.3, 6.4.7, 6.6 and Annex B.
• European Commission, JRC. ILCD Handbook: General guide for Life Cycle Assessment, Detailed guidance. First edition, March 2010. Provisions 6.4, 6.6 and 6.10.
• WBCSD. Guidance on Avoided Emissions, version 2.0. 2025.