@solomon.ives First things first, I’d just make sure to check your units - ECBS reports values in kgCO2e/m2, not /sf.
The longer answer - information about the scope of the LCAs in the database is described in section 3.3 and appendices B and C of the CLF Embodied Carbon Benchmark Study pdf here. The study describes everything much more thoroughly, but I think Appendix C table13 is a particularly helpful summary of the scopes of admitted LCA studies:
Where S = structure, F = foundations, E = envelopes, and I = Interior elements, and the column headers describe life cycle stages included in the LCAs. From a total of ~1,190 LCAs admitted to the benchmark study, 52% (n=619, highlighted) represent LCAs that only consider production stage A for structure and foundations (so, envelope and interior building elements, as well as maintenance and end of life are excluded).
There might be other reasons for why your kgCO2e/sf metric is lower than the ones generated in the ECBS - I would suggest reviewing the tables in section IIIB of Appendix C to identify where your dataset and the one from CLF might be misaligned. For example, the ECBS mostly collected new construction projects - are yours representing renovations? Or, the most common LCA tools used for the studies that make up the ECBS are BATH ICE, Concrete Centre and the EATool - do you use a different tool? To isolate why your intensities don’t match, you could start by reading this appendix, and compiling a list of all the “mismatches” between your dataset and the ECBS one.
That being said, while I know the data is far from perfect, our firmwide database of LCA studies done using Tally (n=61) is mostly in alignment with the ECBS, so far. Let me know what you find out about your dataset - and hope this helps.