324
Audio & Video Production315
Software Development230
Automation & Workflow208
Writing & Content Creation190
Marketing & Growth179
AI Infrastructure & MLOps150
Design & Creative156
Photography & Imaging146
Data & Analytics115
Voice & Speech123
Education & Learning120
Sales & Outreach115
Customer Support112
Research & Analysis86
Salesforce is meeting with customers as often as weekly to decide what AI features to build next, aiming to ship updates faster as needs change.
In short: Salesforce says it is using frequent customer meetings to decide what AI features to build and release next.
Salesforce, the company behind widely used customer management software, told TechCrunch it is “crowdsourcing” its AI roadmap. That means it is letting customers help set priorities for what the company builds, especially for AI features.
Instead of checking in once or twice a year, Salesforce says it meets with some customers as often as once a week. The idea is simple, if one large business customer runs into a problem, other customers are likely to run into the same one.
Salesforce leaders said this approach helped drive products like Agentforce, its software for managing AI agents. AI agents are programs that can take actions for you, like drafting replies, looking up information, or completing steps in a business process (like an assistant that can do tasks, not just answer questions). Salesforce said customer feedback also shapes newer AI work across areas like voice tools and Slack.
Some customers described getting early access to tools before they are released. Engine, a travel management company, said it gave feedback on an AI voice agent that sounded unnatural, and later saw changes after sharing that input. PenFed, a federal credit union, said it built a workflow for handling internal IT requests using Salesforce tools, and Salesforce later made a version available more broadly.
Many companies are still figuring out how to use AI at work, and what is worth paying for. Salesforce’s approach is like building with a focus group that never ends, which could mean faster fixes and features that better match real needs. It also carries a risk, customer requests can be short-term, and early testing does not always turn into long-term use.
Source: TechCrunch AI