Hacker News new | ask | show | jobs
by sandGorgon 3155 days ago
The hardest part is not the client and the server...but the CRM (the Intercom part). Here's something a lot of us would use - we want to bring our own socket API (for example pure websocket, xmpp or something like Pusher.com ) and connect it to a full fledged Intercom-like tool.

This is the "Gitlab for Intercom". The problem that happens is that people usually don't want to rip out their client side messaging SDK, especially if you have a mobile app.

I am personally looking to pay for something like this (we use Pusher.com internally).

Are you focusing on the socket server or the CRM application (like Intercom) ? Because the socket server part is a don't care for everyone. Pusher or Firebase is super cheap. It's the CRM that's tricky. Smooch is one end of the spectrum and it focuses on integration with other tools (helpscout,etc) while Intercom is obviously a CRM. I don't think you'll be able to build both.

If you can build this out with excellent featureset like Intercom and can integrate with something like Pusher (and not lock to your own server), I'm gonna throw money at you!

2 comments

Just want to echo, I’ve recommended intercom.io for a company I was on the board of. It was a huge boon to them. The chat was 5% or less of the value. The analytics… tieing messaging based on actions… the visibility into active/growing, influencers/detractors vs passive/slipping away and guided processes for new features: (show this set of messaging to my feature flag’d customers who have access to beta functionality)… thats the value of intercom.
Wow, this is incredibly valuable feedback, thank you so much! We appreciate your candor (and encouragement!)

You're right in that the socket server is fairly easy to pull in using other tooling. Our focus at the moment is to just build a stable set of pieces that you can use to interact with your website visitor. The operator transports are where things will get interesting, connecting to other tools, etc. This is where connecting things like Pusher will shine, we hope.

Thanks again for the feedback!

The gitlab-for-intercom has genuine value there, so if I were you, I would just use a third party transport and build a killer product there.

Build something I ^H^H people want ;)

IMHO the rest can come later. Just my $0.02 . Excited to see what you come up with.