Five Arrows Logo

This Week in Small Business Tech: Booking & Follow-Up

October 30, 2026

This week, the Vivid Signal platform added new ways to book services in a client portal, capture useful details after calls, test a workflow where it was built, and check current facts from an AI workflow. For a small business, the payoff is fewer dropped handoffs, if you make the changes carefully.

Spend less time reconstructing calls

What changed: Voice Intelligence can summarize recorded calls, suggest updates to contact details and call outcomes, and score a call against your QA checklist. You approve suggested changes before they reach the contact. It is in Labs, requires call recording, works with LeadConnector numbers, and skips voicemails and very short calls.

Why it matters: The next person who speaks with a customer can see what they asked for and what was agreed without replaying the whole call. This is useful call context, not an automatic callback task; automatic follow-up tasks are still listed as a future addition.

Bottom line: If your team records calls, this can make the next follow-up easier to pick up and keep contact details more current.

Let customers book services from the Client Portal

What changed: Customers can browse and book enabled services from the Client Portal, then manage eligible bookings there, including rescheduling or cancellation. Meetings and service appointments appear together under Appointments. You need to enable service booking and the Appointments permission first.

Why it matters: Customers who already use the portal have one place to find services and manage appointments, instead of hunting through old booking links. This works for portal users, not as a public booking page for every visitor.

Bottom line: For a service business already using the Client Portal, make your most common bookable services available there and confirm the access settings.

Test a workflow where you built it

What changed: AI Builder can now run a workflow test from the same chat where you built or edited it. You can watch each step, inspect its inputs and outputs, and read a summary of successes, skips, and errors.

Why it matters: A test can expose missing data or a broken step before you rely on the workflow in production. Important: the test runs real actions. It can send messages, add tags, or write notes, so use your own test contact and check any customer-facing actions before starting.

Bottom line: Test with a dedicated contact first, especially when the workflow includes texts or emails.

Let an AI workflow check current facts

What changed: You can enable web search inside the AI Agent action, or give it specific URLs to consult. It searches when a run needs information it does not already have. This is a premium feature billed per search: the changelog lists 1 cent for OpenAI or Claude models and 1.4 cents for Gemini models.

Why it matters: A service business could have an agent check a trusted rebate page or current product information before it responds. Search only finds information; the agent still needs the right message action and clear instructions to reply, and your team should review the sources it uses.

Bottom line: Start with one trusted source and one narrow question, then check the answers before expanding it.

What This Means for You

These changes help most when they support a real handoff: what happens after a call, where a customer books, or how you know an automation works. Pick one bottleneck, make one change, and test it with your own record before it reaches a customer.

Want a second set of eyes on your setup? Book a 15-minute snapshot at fivearrowsgrowth.com. 15 minutes, no pitch.

Back to Blog