AI Receptionist Quick-Start Checklist: From Sign-Up to Live in 15 Minutes

Create Your Account

Go to talkingwidget.ai/pricing and choose the plan that fits your call volume. Your account is active immediately — no waiting, no manual review.

• Visit talkingwidget.ai/pricing and select your plan
• Complete account registration with your business email
• Confirm your email address via the verification link
• Log in to your dashboard — your AI receptionist environment is ready
• Note your provisioned phone number (in the Phone Numbers tab)

Choose Your AI Persona

Your AI receptionist needs a name, a voice, and a personality that matches how you want your business to sound. This is the first impression every caller receives — take two minutes to get it right.

• Go to Settings → AI Persona in your dashboard
• Choose a name — common choices are the business name ("Parker from Parker Electrical") or a friendly given name
• Select your voice style — warm professional Australian is the default and the most versatile starting point
• Set the tone slider: formal, balanced, or conversational
• Preview the voice with a sample phrase and confirm it sounds natural

Configure Your Business Details

This step feeds your AI receptionist the knowledge it needs to handle real calls. The more complete you make it now, the fewer follow-up calls your team will need to take.

• Enter your business name and trading name (if different)
• Add your business hours — including public holidays if relevant
• List your core services — keep each description to one to two plain-English sentences
• Set your pricing policy — "quote on request", "fixed pricing on website", or provide actual prices for common services
• Write your opening greeting — e.g. "Thanks for calling [Business Name], I'm [Persona Name]. How can I help you today?"
• Enter the top 10 questions callers commonly ask, with your preferred answers
• Add emergency escalation contact details (the mobile number a caller should get if there is an urgent issue)

Connect Your CRM

Every call your AI handles should automatically create or update a lead record in your CRM. This takes one click via OAuth — no API keys, no developer required.

• Go to Integrations → CRM in your dashboard
• Click your CRM and authorise via OAuth — you will be redirected back automatically
• Map the call data fields: caller name, phone number, enquiry type, booking status
• Set the pipeline stage new leads should enter in your CRM
• Run the built-in connection test — a test record should appear in your CRM within 30 seconds

Connect Your Calendar

Live calendar access is what transforms your AI receptionist from a message-taker into a booking engine. The AI checks your real-time availability during each call and books confirmed appointments directly — no double-booking, no manual entry.

• Go to Integrations → Calendar
• Authorise your calendar via OAuth
• Select which calendar(s) the AI should read availability from
• Set your booking buffer — minimum time between appointments (e.g. 15 minutes for travel)
• Enable booking confirmation emails — caller and team both receive a confirmation automatically
• Test with a dummy booking from the dashboard — verify it appears correctly in your calendar

Set Up Call Routing

Call routing determines which calls your AI handles and which calls reach a human. Getting this right before going live prevents your AI from attempting to handle situations it is not yet configured for.

• Go to Settings → Call Routing
• Set business hours mode: AI answers all calls, or AI answers when team is unavailable
• Set after-hours mode: AI takes messages and books callbacks, or AI takes messages only
• Configure emergency escalation — callers mentioning urgent keywords (gas leak, flood, emergency) are transferred immediately
• Set the human escalation number — where calls go when a caller explicitly asks to speak to someone
• Enable missed-call SMS follow-up — callers who hang up before resolution receive an automatic follow-up text

Test Your AI Receptionist

Never send real callers to an untested AI. The built-in test call feature lets you run through every scenario in minutes — from the dashboard, or by calling your test number from a mobile phone.

• Open Dashboard → Test Call and run the built-in scenario simulator
• Call your provisioned test number from a real phone and simulate a standard booking request
• Test a pricing enquiry — confirm the AI responds with your configured pricing policy
• Test an after-hours call — confirm the AI does not attempt to book during closed hours
• Test an unclear or rambling caller — confirm the AI stays patient and guides them to a resolution
• Test a caller requesting a human — confirm the escalation transfer fires correctly
• Review the call transcripts for each test in your dashboard
• Verify that test bookings appear in your calendar and test leads appear in your CRM

Go Live

You have tested. Everything looks right. Time to route real callers to your AI receptionist. This step activates the live number and adds the voice widget to your website.

• Go to Settings → Phone Numbers and toggle your provisioned number from Test to Live
• Set up call forwarding from your existing business number to the AI number via your carrier (takes 2 minutes via your carrier's app or settings page)
• Copy your embed snippet from Dashboard → Website Widget
• Add the snippet to your website's </body> tag — the voice widget appears automatically on every page
• Update your Google Business Profile with the new AI phone number if replacing an existing number
• Announce to your team — let staff know calls may now be answered by AI, and when to expect escalations

Monitor Your First Week

Your AI receptionist improves quickly with real call data. During the first week, review your analytics dashboard daily and act on what you find. Most issues that surface in week one are small fixes that take under five minutes each.

• Check the Analytics Dashboard daily — total calls, resolution rate, escalations, missed calls
• Review 10% of call transcripts (or all escalation transcripts) each day
• Look for knowledge gaps — questions the AI could not answer — and add those answers to your FAQ list
• Check that CRM records are being created accurately after each call
• Confirm calendar bookings are appearing correctly and confirmations are being sent
• Verify missed-call SMS follow-ups are being delivered to callers who did not get a resolution

Optimise

After your first week of live calls, you have real data to work with. The optimisations in this step typically push your resolution rate from the 70s into the 85-plus range and significantly reduce unnecessary escalations.

• Add every unanswered question from week-one transcripts to your FAQ list
• Adjust your greeting message if transcripts reveal callers are confused about what the AI can help with
• Tighten your booking confirmation language — if callers are unclear about what happens next after booking, rewrite the confirmation message
• Review your after-hours messaging — if late-night callers are attempting to book immediately instead of requesting a callback, adjust the after-hours greeting
• Expand your service descriptions if calls reveal callers are asking basic "what do you do?" questions before getting to their actual need
• Review escalation patterns — if the same type of call is escalating repeatedly, that is a signal to improve your AI's handling of that call type rather than always sending to a human

5 Common Mistakes to Avoid

These are the five setup mistakes that generate the most support tickets in week one. All five are easily avoided if you know what to watch for.

All 10 Steps at a Glance

Print this or bookmark it. Every step, every time estimate, in one view.