Hey all, founder right here searching for some trustworthy takes 🙏

We constructed an AI-native social advertising and marketing platform earlier this 12 months, primarily centered on DM automation. The flagship use case is the traditional "key phrase on a publish triggers an auto-DM" sample: somebody feedback in your publish, they get a hyperlink, low cost, or invite of their DMs with none guide work.

However we've began noticing one thing in the previous few weeks.

A handful of customers have been utilizing the product for buyer assist as an alternative. Not what we constructed it for. Now we have a characteristic that makes use of AI to establish which incoming feedback and DMs are price replying to (flags actual questions, complaints, shopping for intent, that sort of factor), and these customers are mainly treating it as a unified assist inbox throughout socials.

My thought course of is:

• Most clients in the present day are already on socials.
• They're far more prone to DM a model on Instagram than open a assist ticket on a web site.

Nonetheless, I do know that is additionally a crowded area, with current merchandise like Intercom or ChatBase which might be fairly nice.

The angle I've been kicking round internally:

• A buyer assist software that lives natively the place clients truly are, on socials.
• Cross-platform by default. If somebody DMs you on Instagram in the present day and tweets at you subsequent week, that's the identical individual, and your assist group ought to know.
• Plug in Shopify or your CRM and also you get their order historical past alongside their messages.
• One inbox, all socials, with context feeding an AI agent that both reply on autopilot or suggests replies.

I genuinely don't know if it's price spending time on this path or simply two or three of our customers doing one thing bizarre (sure, I talked to them already).

So asking: does this resonate? Anybody right here working buyer assist for a shopper model and wishing the instruments have been higher fitted to socials?

And if anybody needs to share their expertise, suggestions or chat about it, glad to attach!