I expect that with the introduction of Turbot we've started using exclusively Sorting Office for the parsing of all addresses in ETLs. The problem described here about the old ETL becomes then a problem for Sorting Office to solve.
Can you please check if Sorting Office's current algorithm is affected by the same issue?
As long as it's Companies House data only, this is not urgent, as companies telephone numbers are not confidential information and are published by Companies House because of regulations in the first place. But we need to be prepared for when a distracted contributor give us their telephone number by mistake.
I expect that with the introduction of Turbot we've started using exclusively Sorting Office for the parsing of all addresses in ETLs. The problem described here about the old ETL becomes then a problem for Sorting Office to solve.
Can you please check if Sorting Office's current algorithm is affected by the same issue?
As long as it's Companies House data only, this is not urgent, as companies telephone numbers are not confidential information and are published by Companies House because of regulations in the first place. But we need to be prepared for when a distracted contributor give us their telephone number by mistake.