Well-meaning people using Excel, are without realizing it, too often making analysts and coders cry for inefficiency reasons. This is very costly, both financially and emotionally, and must be improved:

• I suggest to have everybody invest a few hours in "Data Organization in Spreadsheets ...", from the awesome carpentries.org, for which I am a certified instructor. "We teach foundational coding and data science skills to researchers worldwide."
• Advertise that course, this page, or your own, widely in contexts, including in the default 'new Excel sheet'.
• Try a pre-structured spreadsheet, like what I made a rough sketch for in the Excel file in this repo.
• When the 'damage' is already done, and you have to figure out the (implicit and hidden?) structure of an already existing Excel spreadsheet, look at https://mobile.twitter.com/steltenpower/status/1330290665322934278
• Questions will arise and Digital Competence Centers and the like should be great at answering them
• It is not uncommon that finding and cleaning of data takes way more time than everything else in a project. Put that in your planning and make sure managers read it, to make sure they think about investing in data quality.
• (Mental note: Excel's comma-delimited .csv is actually semi-colon (;) delimited. See if there are carpentries.org courses to explain this and maybe mention how to roundtrip from Python/R with Excel on CSV)
• Why electronic notebooks make you work together faster and fewer errors
• If datacarpentry and softwarecarpentry sound scary, try Library Carpentry

If you actually want or need further structuring of data:

• oversimplified ultra-summary of FAIR data principles, from there go to GO_FAIR for the real deal.
• Data-centric Manifesto Data 1st!, applications should be 'mere visitors'.
• an example of a research Data Management Plan template
• Some of my related brainstorm-ideas on getting&using data FAIR&early in happy cooperation workflows

Excel is not all bad if you know how to use it, plus there are nice extensions:

• E2D3 (Excel to D3, where D3 stands for Data Driven Documents; aka the most used library for defining your own visualisations)
• Excel querying Linked Open Data cloud
• Perfect XL helps some