I'm a software engineer at Academia.edu. We have a feature which allows users to see which papers mention which users.

That feature is only available to premium users. During the first month, you can email us to cancel and we'll refund you. So users can test-drive the mentions feature, and if they decide they don't like it, they can cancel. We think it's useful, and we don't want users to be coerced into using it.

Also, we do make mistakes. Sometimes we mistake a user for another academic with the same name! But Google finding only one paper with a name on academia.edu does not mean that there is only one paper that mentions that name on the site.

A cold call email usually a scam.

but as the software engineer of Academia.edu said:

I always knew it has to be their way of fishing us to subscribe but knowing Academia.edu is quite known website, I was wondering anyone has tried this and find their result accurate.

Btw, I didn't aware that they will refund us if we cancel the subscription. Thanks for that info.

Have you done a check by searching your name in other database/search engine like google scholar, researchgate, etc?

Yes, I have but none came out. That's why I'm sceptical when they say I have 7 mention in Academia.edu

In this case, I would ignore it if it requires paid subscription.

Yup, that's what I did. I was just wondering have anyone tried it and whats the outcome.

Honestly speaking, this is the first time I heard about this site, it's basically just a database similar to researchgate, researcherid, etc. So I am guessing, they compile all your works and then offer you to join. You have at least 7 published works, right?