← Home
35 terms · plain English
Glossary — every term we'll use, defined plainly.
We'll throw words like A2P 10DLC, ASA, lift, NAP, and tool use at you during onboarding. Here's every one of them, stripped of jargon, with links to the relevant system page.
AI
6 terms- AI receptionist
- A voice agent that answers your business line 24/7, qualifies callers, books jobs against your real calendar, and texts the dispatch summary to you. Ours runs on Anthropic Claude + ElevenLabs voice + Twilio. → see systems · ai-receptionist
- Call summary
- A 2–3 sentence Claude Haiku-generated digest of every call: what the caller wanted, what was decided, what's outstanding. Texted to your cell + posted in the portal. Searchable across the call log.
- ElevenLabs
- The text-to-speech engine we use for voice synthesis. Sub-500ms latency. Multilingual v2 model. Voices can be cloned from a 3-minute sample if you want the AI to sound like your brand.
- Emergency triage
- The agent's logic for distinguishing 'pipe just burst' from 'I want a quote next week.' Configured per industry during onboarding. Emergencies transfer to your escalation number; routine calls book or capture as leads.
- Outbound engine
- Targeted cold email + SMS to local prospects (commercial property managers, GCs, realtors) using Claude-generated personalization. Available on Scale plan only. Compliant with CAN-SPAM + 10DLC. → see systems · cold-email
- Tool use
- The Anthropic SDK feature that lets the AI receptionist actually do things — book_appointment, capture_lead, transfer_to_human, end_call — instead of just talking. Each tool runs server-side with type-checked arguments.
Voice / SMS
2 terms- Twilio
- The telecom API powering our voice + SMS layer. We provision a dedicated number per client (or port your existing line), wire the inbound webhook to /api/voice/incoming, and stream conversation turns through Anthropic.
- TwiML
- Twilio Markup Language — the XML response format Twilio expects from your voice webhook. We return TwiML with <Gather> for speech input and <Say> with Polly Joanna-Neural fallback while ElevenLabs synthesizes.
Service ops
12 terms- Answer rate
- Percentage of inbound calls that get picked up before going to voicemail or being abandoned. Industry baseline for service businesses is 65–80%. Our clients land at 97%+ within month one.
- ASA
- Average Speed of Answer — the average seconds it takes for a caller to reach a real agent (human or AI). Human receptionists average 12–25s. Our AI averages 0.4s.
- Average ticket
- The dollar value of a typical job in your industry. HVAC ~$8,500. Plumbing ~$650. Roofing ~$14,500. Used to compute the dollar-value of every missed call. → see industries · hvac
- Baseline
- The pre-install snapshot of your operational metrics: calls/mo, missed-call rate, average speed of answer, average first-reply, review velocity, conversion. We capture this in the audit and use it to compute month-one lift. → see audit
- CSAT
- Customer Satisfaction Score — typically a 1–5 or 1–10 rating after a service interaction. We trigger a CSAT prompt 90 minutes after job completion as the gate to our review-engine routing.
- Dispatch
- The process of routing a job request to the right tech at the right time. Modern dispatch uses ServiceTitan, Housecall Pro, or Jobber. The AI receptionist hands off to your existing dispatch system, never replaces it.
- Housecall Pro
- Field-service platform popular with HVAC, plumbing, electrical. We integrate the AI receptionist with HCP via their public API for calendar, customers, and dispatch.
- Lift
- The delta between baseline and post-install KPIs after the 30-day pilot. The pilot terminates as 'lift achieved → invoice' or 'no lift → uninstall, you owe nothing.'
- Missed-call rate
- Percentage of inbound calls that hit voicemail or get abandoned before a human answers. Industry average: 22%. Translates directly to revenue leak: missed × ticket × 20% close-rate.
- Pay-after-results
- Our pricing model. The 30-day pilot is free. Clients only become paying clients once month-one numbers beat their documented baseline. If they don't, the install uninstalls and the client owes nothing. → see pricing
- Pilot month
- The first 30 days after install. All systems live. We publish a weekly dashboard. Cost: $0. Outcome: invoice or goodbye.
- ServiceTitan
- Field-service operating system used by larger HVAC/plumbing/electrical shops (typically $50M+ revenue). We integrate via their API for calendar, customers, jobs, and dispatch.
Marketing
12 terms- Booking rate
- Percentage of qualified callers who agree to schedule. AI agents typically book 28–42% on the first call vs. 15–22% for an under-trained human receptionist with no calendar tool access.
- Calendly
- The booking tool we use for the 20-minute audit walkthrough call. Embedded into /book and linked from the AI agent's book_appointment tool when relevant. → see book
- Conversion rate
- Percentage of website visitors who take the desired action (call, text, fill the form). Most service-business sites convert at 1.5–3%. We rebuild as 'conversion sites' targeting 6–9%. → see systems · conversion-site
- First-reply time
- The minutes between a lead's form-fill and the first human or AI response. Industry average: 42 minutes. Our Speed-to-Lead engine: under 60 seconds. → see systems · speed-to-lead
- Google Business Profile
- Formerly Google My Business. Your presence in Google's local 3-pack. Profile completeness + review velocity + recency + response rate are the ranking factors we optimize for.
- LCP
- Largest Contentful Paint — the Core Web Vital that measures how long your largest above-the-fold element takes to render. Google ranks pages with LCP under 2.5s higher. Our conversion sites land under 1.6s on mobile 4G.
- Lead score
- 0–100 score we compute on every lead: source weighting (booked > inbound call > calculator > CTA), email/phone presence, business name, industry, payload richness, UTM attribution, deal-value cents. Visible on the lead detail page.
- Local SEO
- Search optimization for queries with local intent ('hvac near me', 'emergency plumber 60601'). Driven by Google Business Profile, NAP consistency, citations, and review velocity. → see guides · google-reviews-playbook
- NAP
- Name, Address, Phone — the three citation fields that need to be byte-identical across Google Business Profile, your site, Yelp, Facebook, and directories. Mismatch tanks local rankings.
- Review engine
- Post-job SMS + email sequence that asks the customer to rate the visit, then routes 4–5 stars to Google and 1–3 stars to your private resolution queue. Lifts review velocity 2–4x in the first 30 days. → see systems · review-engine
- Review velocity
- Reviews earned per month. Google's local algorithm weights recency heavily — a business with 12 reviews in the last 90 days outranks one with 200 reviews from 2 years ago.
- Speed-to-Lead
- How fast the first contact reaches a lead after they raise their hand. Conversion drops 80% if the first touch takes longer than 5 minutes. Our engine fires under 60 seconds, every time. → see systems · speed-to-lead
Compliance
3 terms- A2P 10DLC
- Application-to-Person 10-Digit Long Code. The US carrier registration program for sending automated SMS from a regular phone number. Required for legitimate Speed-to-Lead campaigns. ~$4/mo + a one-time vetting fee. We register you during onboarding. → see legal · sms-consent
- Recording retention
- How long we keep call audio. Default: 30 days. Configurable up to 1 year on Scale plan. Stored in Vercel Blob with at-rest encryption. Deletable on request.
- STIR/SHAKEN
- The US carrier framework for verifying that an outbound call really comes from the claimed number. We provision Twilio numbers with valid SHAKEN/A attestation so your AI doesn't get spam-labelled.
Ready? It takes 30 seconds.
Book a 20-minute audit.
Walk away with the plan — whether you hire us or not.
We'll pull your Google rank, a simulated missed-call flow, your review velocity, and a dollar-value of your current leak. You get the full audit document — yours to keep.