Resources / guide
A practical launch guide covering reader supply, call routing, pricing, payments, safeguards, support and measurement.
01 / Start here
For an operator planning live paid readings with several readers. Use it to specify the service, reader controls and customer protection before buying telephony or building a portal.
The first decision is whether customers pay for an on-demand minute, reserve a fixed session, or choose either. This choice shapes availability, billing, refunds and what the product has to display while a conversation is running.
02 / Core guide
A psychic hotline can be one-to-one calling, scheduled calls, chat, a network of independent readers or a hybrid. The technology should follow the service promise. Decide who your customers are, which countries you can serve, what a session costs, and what happens when a preferred reader is unavailable.
Choose whether readers are employees or independent providers, whether calls are live or scheduled, and whether the customer pays in advance, per minute or per session. Document the reader share and any minimum payout. If you operate in several countries, review local consumer, payment, privacy and telecommunications requirements with qualified advisers before launch.
A caller should be able to discover a reader, see the price and availability, understand billing, consent to the service, connect, see time or balance remaining, and receive a receipt. Specify what happens on a dropped call or when a reader does not answer. A missed connection needs a clear retry or refund path.
Common components include profiles, search or matching, availability, call or chat infrastructure, metered billing, a payment gateway, payouts, account credit, consent records, support tickets and operational reporting. Choose providers for the markets you serve; test call quality, latency, settlement and failure handling with real devices.
Start with a small reader roster and defined service hours. Place test calls across devices and time zones. Track attempted connections, connected calls, paid minutes, payment failures, refunds, support contacts and repeat bookings. Interview both callers and readers before automating more of the workflow.
Launch gate: published pricing and terms · reader contracts · consent and privacy review · payment and payout reconciliation · test calls · refund and dispute owner · support coverage · reporting.
Paid minutes are only one part of the business. Begin with reader capacity and utilisation, then subtract payouts, telephony, payment fees, customer acquisition, support and refunds to estimate contribution. A large theoretical calling capacity is not evidence of demand.
Use the hotline revenue calculator ↗ · Explore hotline systems ↗
03 / Make the choice
The right choice depends on customer expectations and the team’s ability to deliver it consistently. Use the trade-offs as a starting point for a scoped pilot.
| Model | Best fit | Operational consequence |
|---|---|---|
| Instant per-minute call | Customers want a reader now; enough providers can stay available. | Requires live presence, reliable routing, a visible rate and metered billing that stops when the call ends. |
| Scheduled fixed-price session | Specialist readers have limited hours or sessions need preparation. | Easier pricing and planning, but calendar integrity, reminders, no-shows and rescheduling become central. |
| Hybrid network | You can manage both live coverage and booked experts. | Offer the two journeys distinctly; do not assume one balance, cancellation or payout rule fits both. |
04 / Build and test
Give each stage a named owner and a concrete acceptance test. Advance when the working journey and support response are clear.
Write a one-page specification covering who may join, how a customer chooses a reader, session pricing, minimum charge, connection failure, refund grounds and reader share. Decide what support can resolve without an owner. Test these rules with a few real readers and clients before automating them.
Map search or matching, availability, call invitation, acceptance timeout, connection, a live price or balance display, disconnection and receipt. Decide whether the customer and reader numbers are masked. Log call state changes so a dropped call can be reconciled to billing rather than guessed from an audio provider invoice.
Create distinct records for attempted calls, connected calls, billable minutes, payment captures, refunds and reader payouts. Test a declined payment, a call answered after a delay, and a partial refund. A support agent should be able to see why a charge occurred without listening to a private conversation.
Recruit a small roster with confirmed hours. Place test calls across mobile networks and time zones, then open controlled traffic during the hours you can support. Review missed connections, paid utilisation, payment failures and complaints daily; change coverage or routing before adding more readers.
05 / Worked example
Suppose a pilot has eight readers covering two languages. A customer opens a profile, sees a published per-minute rate, prepays a small balance and calls an available reader. The reader has a short acceptance window. If the call never connects, no minutes are billed and the balance remains available. If it drops after seven billed minutes, the record shows seven minutes, the applicable price and a support path. The operator reviews connection and refund rates by shift before extending hours.
For each launch market, record supported phone numbers and call routes, customer and reader countries, billing currency, taxation owner, recording policy, consumer terms, payment availability and payout method. Test both sides of an international call on real networks.
06 / Measurement
Agree on the numerator, denominator and time window before launch. Segment by market or service when different customer journeys would hide a problem in the average.
| Signal | How to define and use it |
|---|---|
| Connection rate | Connected calls ÷ valid call attempts; inspect failed attempts by reader, network and hour. |
| Paid utilisation | Billed minutes ÷ reader available minutes; measure by shift, not only as a site average. |
| Contribution per paid minute | Collected amount less reader share and variable telephony/payment costs; exclude unpaid attempts. |
| Refund and dispute rate | Refunded sessions and disputed charges as a share of completed paid sessions, with reason codes. |
07 / Launch gate
Use this as a team sign-off, then store the owner, evidence and test result beside each item in your project tracker.
08 / Questions
Usually the first question is whether a reliable paid connection and support workflow exists. A responsive web experience can validate the model before a separate app, subject to the telephony and payment choices you make.
Show a reader as available only when they can accept and complete a call. Set a timeout, handle a declined invitation and stop routing to readers with stale presence. Keep scheduled availability separate from instant availability.
Record the connection and billing events, stop the meter, notify both sides and give support a clear recovery path. Your pricing and refund rules should define whether a short failed connection is charged.
It can, but payments, telecom providers, privacy expectations, consumer terms and reader arrangements vary. Choose launch markets deliberately and review the operating rules for each before expanding.
After you know which customer preferences and reader attributes actually affect a successful paid session. Begin with understandable filters and reliable availability; automate matching once the pilot produces useful evidence.
09 / Research notes
These sources informed specific implementation checks above. Product features and regional requirements can change; verify the current documentation before committing to a vendor or launch market.
Editorial review: September 2026. The example and planning metrics are illustrative. Regulatory, tax and contract questions require review for the countries and services you choose.
