Phone numbers
E.164 is the international standard shape for a phone number: a leading +, then the country code, then the national number, with no spaces, brackets, or dots. CRMs like Salesforce and HubSpot, and every click-to-call and SMS integration on top of them, want numbers in this form so they dial and text correctly. Sift converts your messy column to an E.164-style form in your browser, so the file is never uploaded.
Convert your phone numbers in Sift →
Search for "convert phone numbers to E.164 in Excel" and the incumbent answer is a five-deep nested SUBSTITUTE formula: it strips the brackets, spaces, dots, and dashes, then glues a +1 on the front. It works, for exactly one country. That +1 is hard-coded, so a UK, German, or any non-US number comes out wrong, and there is no branch in the formula that knows the difference. For a mixed-country list, the US-only formula is the wrong tool.
Even when you get the formula right, Excel fights the result. A cell that starts with + reads as the beginning of a formula, so Excel either throws an error or quietly drops the +. And when you save the file as a CSV and reopen it, the guessing starts over and the leading + is stripped again. So the very character that makes a number E.164, the country-code +, is the one Excel keeps deleting. The durable fix is to build the E.164 column outside Excel and import the file directly, so the + is never round-tripped through a spreadsheet.
Sift will not invent a country code. If a number already carries a + it is kept, and an 11-digit North American number starting with 1 becomes +1 followed by the rest. But when a number is a bare national string with no country context, Sift leaves the digits as they are rather than guessing, because a wrong +country prefix is worse than none: it would send click-to-call and SMS to the wrong place, silently. So for a list of unknown-origin national numbers, supply the country yourself, or map to a CRM template that expects the national format. Sift also cannot restore digits that Excel already discarded to scientific notation or a stripped +; for that, go back to the source export.