It answers everything. It guesses at nothing.
One assistant on your phone line and your chat window, working from a knowledge base built out of your own material. It books, it escalates, and when the honest answer is that it does not know, it says so and gets a person on it.
Built out of what you already have.
Your services, your hours, your policies, your prices, and the questions your team answers most. Nothing invented, nothing assumed.
We take your material
Service list, hours, policies, the answers your team gives on the phone every day. What is not in there becomes a handoff rather than a guess.
You approve how it sounds
You hear it before a customer does, and you set what it may and may not say. A refusal you are comfortable with beats an answer you have to correct.
It goes on the line behind forwarding
Busy, unanswered and out of hours route to it. Your number does not change and nothing already printed goes stale.
You read the record
Transcript, summary, reference, and which turns came from your own facts. That last column is also how the month is billed.
Why the bill is shaped like this.
A turn answered straight out of your own knowledge base costs almost nothing to serve. A turn that has to be reasoned out costs real money. You pay the same for both, so the ratio between them is our margin and not your problem.
| Turn | What happens | Billed |
|---|---|---|
| From your knowledge base | Looked up and answered | One interaction |
| Reasoned out | Worked through, then answered | One interaction |
| Handoff | It says it will not guess, logs the question, names who calls back | Not billed |
| Call minutes | Counted in minutes, so a long call is never billed as a short one | Separate from the chat bundle |
Source: fivestarguard metering rules, assistant_metering.py. The handoff row is enforced in the billing code, not in a policy document.
Every number names its source.
Source: fivestarguard rate card, assistant_metering.py:107 and assistant_metering.py:108. Overage assistant_metering.py:109.
Source: seo_billing.py:53, seo_billing.py:64, tiers at seo_billing.py:76-82. A page only earns its fee in a month when your own Search Console shows it reached position ten or better on at least one query and took at least one click.
Source: pricing.js:22 (domain_install), pricing.js:31 (league_season), pricing.js:43 (crm), pricing.js:52 (club_ops).
Answered straight.
It introduces itself as the business's assistant and it never claims to be a specific person. If a caller asks directly, it answers directly. We would rather it be straightforward than clever about it.
The brief forbids it and the knowledge base does not contain one. When a caller asks for a figure the business has not published, it says it does not have that figure, logs the question and names who is calling back. You can hear this behaviour on any of the six demo lines.
It captures an intake with a reference and tells the caller what happens next and when. A refusal without a captured next step counts as a failed call.
Yes, on the same call, with no callback and nobody else on the line. It is billed by the minute at $2.25 and only on the calls that use it. Calls in English stay at the ordinary rate.
500 chat interactions and 75 call minutes, inside the $199 monthly fee. Chat past the bundle is $0.35 each and minutes past it are $2.
Yes. One knowledge base, one set of rules, both channels. A question answered on the website gets the same refusal behaviour as the same question asked on the phone.
Thirty minutes, and we tell you what we would build.
Or ring one of the six live lines first