Siena raised $17M to give support, shopping and social agents one shared memory of each customer. Your CRM should do the same

Most businesses store every order, but almost none store why the customer bought, returned or complained.

What happened

Siena, an AI customer service startup for consumer brands, raised a $17 million Series A led by York IE. VentureBeat reports this brings total funding to $29.7 million.

Siena calls its direction the Agent of Record. In its announcement, the company argues that a brand's existing systems record transactions (orders, storefront, loyalty, subscriptions, reviews) but none record the relationship: what the customer said and why. Siena Intelligence combines conversations with orders, subscriptions, reviews and loyalty data, and Siena Memory carries that context across support, shopping, social and phone calls. Siena says other tools and agents can draw on the same context through MCP.

Details from VentureBeat:

• Agents can update delivery addresses, create return labels, process refunds and handle subscription changes.
• Siena says it automates up to 80% of customer interactions.
• Pricing starts at $750 a month plus $0.90 per automated ticket, before onboarding and support fees.
• Newer shopping, social, voice and intelligence features are in early access.

Siena also shares one customer result: at Eskiin, its Social Agent took weekly comment engagement from roughly 1,600 to 2,400.

My take

You do not need Siena to steal the core idea. The gap it names exists in almost every CRM I open: deals and orders are logged, but the reason behind them lives in call notes, DMs and support threads nobody reads again.

A practical version for a GoHighLevel or HubSpot setup:

1. After every support or sales conversation, have an AI step extract a few structured fields: reason for contact, objection, product feedback, next best action.
2. Write those to custom fields on the contact, not to a free text note.
3. Make every other automation, from follow ups to win back campaigns, read those fields first.

That gives you shared memory without a new platform. The pricing also shows where the market is heading: per resolved ticket, not per seat. Budget your own builds the same way.