Notes on live chat and support
Product updates, guides, and ideas for supporting customers on your website and in messaging apps.
Why service businesses lose jobs to missed calls — and how to fix it
Work out what unanswered calls and messages actually cost you each month, then close the three gaps that cause almost all of them.
Read article →Appointment reminders that cut no-shows, with templates you can copy
When to send a reminder, what belongs in it, and copy-ready wording for booking confirmations, day-before notes, and on-the-way messages.
Read article →What to ask before you book a service visit: an intake checklist
The questions that decide whether a visit succeeds: what to ask on every job, what to ask by trade, and which questions to stop asking.
Read article →How to choose field service scheduling software: a 10-point checklist
Ten questions that separate scheduling tools which survive a real week from demos that look good — including when you don't need one at all.
Read article →After-hours and emergency calls: a policy you can actually staff
Define what counts as an emergency, what a caller hears at 23:00, and who is genuinely on call — before a tired phone decides for you.
Read article →Assist mode: approve what the AI proposes, one field at a time
How the review step turns a model's reading of a conversation into changes an operator accepts or rejects individually — with provenance, version checks, and a full audit.
Read article →One operational list: service requests and jobs in a single workspace
See how the Operations workspace joins requests, jobs, scheduling, and resources into one filtered list without becoming a third state machine.
Read article →Business hours, service zones, and emergency policy: the rules a model cannot widen
Configure the deterministic rules Lavenity applies before anything Ven produces reaches a customer — and see what an empty policy quietly allows.
Read article →Measuring an AI Front Office: from conversations answered to jobs completed
Support reporting counts replies. These numbers count what became work — requests qualified, visits booked, jobs completed — plus the correction rate that gates autonomy.
Read article →Build the service catalog your AI qualifies against
Typed fields, required-for-action rules, and the exact question Ven may ask — how a service template, not a prompt, decides what ready means.
Read article →Address, access, and equipment: what turns a request into dispatchable work
Why the property and the asset are separate records, how the address became a readiness requirement, and what the technician brief carries to the door.
Read article →From inbox to completed job: the Lavenity AI Front Office roadmap
See what Lavenity has already delivered across service requests, Ven qualification, scheduling, and jobs — and what comes next before broader automation.
Read article →One AI agent, two specialized roles: how Ven routes every customer turn
Learn how one Ven identity combines Customer Support and Front Office without duplicate replies, unclear ownership, or one oversized permission set.
Read article →How autonomous qualification turns a customer chat into a ready service request
Follow the Qualify flow from service intent and template matching to validated facts, approved questions, audit history, and a request ready for scheduling.
Read article →Safe field-service scheduling: availability, holds, and conflict-free booking
See how Lavenity searches deterministic availability, holds slots temporarily, prevents conflicts, confirms appointments, and delivers reliable notices.
Read article →From confirmed appointment to Job: work orders that preserve customer context
Learn why every confirmed appointment creates one work order, what the technician brief preserves, and how the Job lifecycle stays consistent.
Read article →Shared inbox vs help desk: which support workflow fits your team?
Compare shared inbox and help desk software by channels, ownership, ticket structure, automation, reporting, and the complexity your support team actually needs.
Read article →How to reduce customer support response time without rushing replies
Nine practical ways to reduce first-response time through queue ownership, routing, notifications, saved replies, AI, staffing, and better measurement.
Read article →How to build an AI-ready customer support knowledge base
Structure an AI-ready support knowledge base with clear scope, authoritative answers, useful metadata, privacy controls, test questions, and accountable owners.
Read article →Live chat vs AI chatbot: which one does your website need?
A practical comparison of human live chat and AI chatbots across speed, trust, coverage, cost, and the moments when a hybrid workflow works best.
Read article →How to choose live chat software: a 12-point checklist
Evaluate live chat software with 12 checks for website speed, mobile usability, team workflow, automation, security, reporting, and predictable cost.
Read article →AI customer support: a practical 30-day rollout plan
Launch AI customer support in 30 days with a staged plan for scope, knowledge, testing, human review, controlled automation, handoff, and measurement.
Read article →Proactive chat triggers that convert: a playbook for pricing, checkout, and onboarding pages
Five page types, five different reasons a visitor is hesitating, and five proactive message recipes built around audience, page, and delay instead of one generic greeting for the whole site.
Read article →Should AI send the first message? Why proactive outreach stays human
An AI agent can answer the question a customer asks. Deciding to interrupt someone who hasn't asked anything yet is a different kind of decision — and why proactive outreach in Lavenity stays a human one.
Read article →Ven AI vs Ven AI Agent: assisted drafts or autonomous replies?
Two ways AI can sit in a support inbox: drafting an answer for a teammate to send, or sending the answer itself. How to decide which one belongs in which conversation.
Read article →Grounding AI replies: how to stop confident, wrong answers
An AI agent that cannot cite where an answer came from is guessing in complete sentences. What knowledge grounding actually requires, and how to tell when coverage is thin.
Read article →When AI should hand off to a human: designing the escalation line
An AI agent that never escalates isn't more capable, it's worse at knowing its own limits. How to decide what should hand off to a teammate, and what a good handoff actually carries with it.
Read article →Customer support in Norwegian: Bokmål, Nynorsk and the rules that apply
What a Norwegian-speaking support queue actually requires: two written standards, an AI agent that does not mix them, and GDPR obligations enforced by Datatilsynet.
Read article →Proactive chat messages: start the conversation first
How to plan proactive chat messages that open a real conversation: choosing the moment, the audience, the page, and the person who will reply.
Read article →Announcement banners that customers actually read
A release email reaches an inbox; a banner reaches people already using the product. How to write, target, schedule and retire an in-product announcement.
Read article →NPS and CSAT surveys on your website: ask at the right moment
A practical guide to in-product surveys: choosing between NPS, CSAT and multiple choice, timing the question, asking the right people, and acting on answers.
Read article →How to avoid missing customer messages from your website
A practical guide to combining a shared inbox, clear ownership, Web Push, and email so your support team can catch website conversations sooner.
Read article →Browser push notifications for customer support teams
Learn how Web Push helps support teams notice new customer messages, route conversations correctly, and respond without keeping a dashboard in the foreground.
Read article →Email vs Web Push for customer support notifications
Compare email and browser push notifications for customer support, including speed, context, alert fatigue, and the workflow that makes both channels useful.
Read article →Customer support automation for Ukrainian businesses
Learn how Ukrainian businesses can automate customer support, route website and messenger conversations, and reply faster with Lavenity Automations.
Read article →Customer support automation in Europe: GDPR-aware workflows
Build GDPR-aware customer support automation for European teams with clear purposes, minimal data, controlled webhooks, human handoffs, and workflow history.
Read article →Customer service automation for US small businesses
A practical guide to customer service automation for US small businesses: automate replies, routing, tags, follow-ups, and webhooks without enterprise overhead.
Read article →Turn support reports into concrete actions
A practical approach to Lavenity reporting: choose the right slice, read metrics in context, and finish every review with a specific decision and owner.
Read article →Response and resolution times without metric traps
Understand first response, gaps between messages, time to close, and active handling so the team can move faster without chasing a flattering but incomplete number.
Read article →Measure the contribution of AI and the support team
Follow conversations after Ven replies, review teammate handoffs, latency, and token use, and evaluate AI by customer outcomes instead of automation volume alone.
Read article →One inbox for your website, Telegram, and Instagram
Lavenity now brings live chat, Telegram, and Instagram conversations into one workspace, with a shared queue, clear ownership, and the full customer context.
Read article →Support customers on Telegram without another queue
Connect Telegram to Lavenity, receive new conversations in the shared inbox, and reply as a team without switching between applications.
Read article →Turn Instagram Direct into a team support channel
Instagram Direct conversations can now arrive in Lavenity, where the team sees ownership, customer context, and one queue across every supported channel.
Read article →How to organize one support queue across every channel
A practical workflow for live chat, Telegram, and Instagram: one entry point, clear priorities, an owner for every conversation, and no hidden queues.
Read article →One brand, different rhythm: voice across live chat, Telegram, and Instagram
Facts and promises should stay consistent, while the length, pace, and shape of a reply can adapt to live chat, Telegram, and Instagram.
Read article →How to hand off conversations without losing context
Assignment, history, and a short explanation of the next step help teams transfer requests from any channel without making customers repeat themselves.
Read article →How to prevent duplicate replies in a shared inbox
A simple assignment habit and visible conversation states help teams respond quickly without sending customers two different answers at once.
Read article →Reply to a specific message without losing the thread
Quote any message, keep its author and preview beside your answer, and return to the original context with one click in the inbox or live-chat widget.
Read article →How we made the Lavenity launcher appear in under 100 ms
A 2.9 KB gzip loader, immediate launcher render, idle iframe preload, and a measured local paint time below 10 ms keep chat from slowing down the host page.
Read article →A simpler path from work email to a ready workspace
A focused four-step flow, editable progress, saved account details, and reliable resume behavior make workspace creation easier to understand and finish.
Read article →One guided setup for domains, live chat, teammates, and the first conversation
A shared four-step checklist, contextual prompts, live progress, and one structured settings page make the path to a working inbox much easier.
Read article →Meet Contacts: one workspace for every customer relationship
Lavenity Contacts combines customer identity, activity, tags, internal notes, search, and direct conversation access in one focused workspace.
Read article →Turn a growing contact list into useful segments
Use customer types, activity views, search, and tags to move from one long list to practical support and follow-up segments.
Read article →A faster tagging workflow for customer context
Create color-coded tags, apply them to one or many contacts, filter the directory, and remove outdated labels without leaving the table.
Read article →Keep the conversation going when the team is offline
Lavenity can collect a visitor’s email after they leave a message while every operator is offline, then make it available to the team in the shared inbox.
Read article →Email notifications for unanswered visitor messages
A configurable safety net for new messages: Lavenity waits for your chosen delay and emails an operator only if the conversation is still unread and unanswered.
Read article →Your inbox, All, and Unassigned: clear ownership for every conversation
A practical look at Lavenity’s new inbox views, automatic ownership, and the difference between who owns a conversation and whether it is still open.
Read article →Saved replies, right inside the composer
Type a slash, find a workspace reply by title or text, insert it, and personalize it before sending — no copy-paste library in another tab.
Read article →Images and files that stay with the conversation
Visitors and operators can now exchange screenshots and documents, with previews, upload progress, durable storage, and cleanup when a message is deleted.
Read article →Lavenity 1.0: real-time chat, read receipts, shared inbox
What shipped in 1.0 — WebSocket delivery, messenger-style statuses, typing indicators, and a team inbox that keeps every conversation in one place.
Read article →Install the widget in five minutes
Install the Lavenity chat widget step by step: authorize a domain, copy the site ID, paste the asynchronous script, and test the first conversation.
Read article →Why chat belongs in an iframe
Learn how a cross-origin iframe, Content Security Policy, signed visitor tokens, and domain allowlists isolate chat from the host page’s JavaScript.
Read article →