How do you do batch editing / creation?

Question

How do you do batch editing / creation? (Desktop web)

Batch editing is when you have multiple records to be edited at the same time but each has different values, even for the same field. For example you get an email with an attached xls of students who changed their names over the summer break. You're updating the names of 100 people, each one to update for their new surname.

The two approaches I've seen are templates and no templates. Templates: download xls template > copy-paste your information to the template and save > upload file > preview rows to update highlighting diffs > save to DB. No template: upload the file with your data > preview data > select columns for mapping and validation > save to DB.

I'm not crazy about either one and maybe I'm missing stuff. The template option seems to be less development work but all that copy paste opens the door wide for human error. The no template option, seems like i would be building a massive system for a relatively minor function.

How do you handle batch editing?

Answer 1

I've seen both of those options as well; clearly there are differences between "create" or uploading data for the first time, and "edit", which in this case I assume means uploading new data that is going to be merged into or overwrite the original data.

Depending on the complexity of the data being uploaded, I've also used a text field into which a user can paste either a simple list (delimited in some way), or a matrix of values, all of which get parsed and uploaded into the proper data structure.

As far as the template option, yes it is less work for the development team, but it violates one of my core principles as a UX designer: Our systems should be designed to take the work OFF of the user and put it ON to our technical systems wherever possible and appropriate. The system should do the work, not the user.

To that end, if your goal is actually to create the best user experience, I recommend doing the development to make it as easy as possible for people to upload their data, then give them a robust diff tool to understand what changes will be made by that upload, along with clear and easy controls to accept, reject, or change those results.