Feature Request - Suggestive Request Merging

[davidw]

Hi,

Most of the Help Requests (HR) that I deal with involve the communication with several individuals. I use public, private and external request replies for this situation. It is not typical 1-on-1 question-answer format.

Unfortunately, many of the participants will send ‘new’ emails instead of replying to the original request from HelpSpot that includes the Request Reference in the subject line. Yes, I use the merging function to put the requests together, but it seems I am searching and merging more than actually working on the requests. So much so, that I have quit doing it and just go through the ‘Active’ requests and remove, complete, or trash the stragglers, then find the main request and close it if resolved.

If you haven’t already, I would like to propose that you begin the thought process on how an incoming Help Request can be analyzed and a merge be suggested based on the similiar content to other existing and open requests (maybe even closed ones and maybe even Trash as options). Then a dropdown or some selector mechanism where one of the suggestions can be selected and the merge complete.

This would be sophisticated, but very cool and very useful if done right. Enough information must be revealed to the user to make an informed decision on what to merge.

Thanks for listening and developing a great product.

David W.

PS. Of course, any other solution or thoughts on how to handle my situation as it exists is certainly welcome. Currently, I do not use the portal, but only handle email requests.

[davidw]

Hi John,

Thank you for that info. Yes, it does help! That is just about what I sketched out on paper for a plan on how to handle this just a few minutes ago.

How accurate is your request id tagging algorithm? Care to share some of the things you look at to determine a match to an existing request?

Thanks.

[ian]

Hi David,

I think what John has outlined is a good way to go (thanks John!). Building it in would be very tricky and in many ways I think not much more accurate. You’d still have to go through them all and merge them and you’d likely have false positives to deal with, etc.

A more custom solution like what John uses is likely to be much more accurate as you can really tweak it to your situation. That said though, we’re always considering these type of things so thanks for the post!

If there’s any assistance I can offer as you customize things a bit please let me know.