Isolation is the feature you are buying.

What does agency size mean in Nerxus One if v1 has no sub-account resale?

It means isolation is the product you can buy, and chrome-resale is not. A tenant holds one organization's leads, keys, senders, diagnostics, DEMOs, sequences, and portals. Clients see a magic-link surface for one business. They do not receive a cloned app with your logo in the sidebar.

The loop does not change at this size. Find, Prove, Win, Build, Run. What changes is the sentence you tell procurement and the sentence you tell a franchise lead. Procurement hears tenant isolation, encrypted secrets, ComplianceGate, four locks. The franchise lead hears we will not give your neighbor's reviews to your manager by accident. They do not hear your own white-label Nerxus.

Nova still orchestrates nine specialists. Scout still never contacts. DEMO is still a Forge working sample. Voice is still inbound. Steward still keeps health as Healthy, Watch, or At-risk. Overnight runner is still not a production claim. Size does not unsay any of those facts. Isolation is the feature you can buy. A lite loop or a resale SKU is not.

Roadmap Not doing for this version includes white-label sub-account resale, autonomous cold voice, cold SMS, review gating, and invented plan prices. Print that. A seller who quietly unsays it is off script.

Read Agencies if you want Monday-Wednesday-Friday in the mix-up voice. This page exists so a size comparison does not invent a SKU. Procurement should leave with the tenant map, the four locks, the no-resale row, and the GoHighLevel compare that opens with their win. They should not leave with a dollar table, a SOC 2 badge, or a promise that overnight already mailed.

How does tenant isolation work for an agency book?

Auth is Clerk. An organization provisions at first sign-in. RLS and secret storage are architecture, not a badge. Places, Resend, and other keys encrypt per tenant. The browser sees masked status. Platform quota is not a shared pot we burn for you.

Business rows dedupe on the listing’s own ID per source inside that tenant. Handoffs persist a uuid. Scout, Radar, Pulse, Forge, Echo, and Steward share it. They do not share it with the tenant next door.

Portals are one business, plain language, no internal notes, no costs, no sibling shops. If a token ever opened two shops, that is a defect.

Audit log language belongs in security and agency extras when commercial policy lands. Do not promise a SOC 2 report from this page. Do not promise HIPAA because a medical book asked.

Why is white-label sub-account resale not a v1 agency feature?

Because resale is a different product: billing pass-through, chrome, support boundaries, and a tenant factory. Shipping isolation is already the hard wall. Shipping a factory on top of a waitlist is how agencies get stuck mid-migration.

Clients who need to "log into the software" in v1 log into a portal that cannot operate Scout. That is intentional. The SMB is not the SaaS user. Making them one is how you inherit their clicks as your incidents.

You can still brand reports and the portal in the ways branding profiles allow. Your mark on the artifact is not a sub-account. Do not stretch that sentence in a proposal.

When resale exists, it will get a page and a changelog entry. Until then, Not doing is the honest row.

Which agents stay shared at agency size, and what must never be shared?

Nova and the roster stay shared inside the tenant. That is fine. Senders, portals, client credentials, and Pulse ownership must stay tight to the book they serve.

Scout remains read-only. A larger shop does not unlock contact-from-search. Relay remains the only mail sender. Voice remains inbound. Echo remains ungated on public buttons.

Steward remains the client record. It is not a white-label CRM you hand to the SMB. GoHighLevel remains stronger if they already live in a mature funnel OS. We remain the loop that starts before the client exists.

Market intelligence, when preview is on, is still provenance-safe and still not a reason to mix two brands' rollups into one slide for a pitch.

What does Monday look like in an agency-sized tenant?

Monday pods run ICPs that match fulfillment capacity. A twenty-person shop that shares one scrap pile is a small team with more Slack. Boards stay signal-named. Saves stay in the tenant that may fulfill them. If you operate two legal agencies, Monday starts by asking which tenant you are in.

Watchtower is how size pays off. Re-scan the book you already own. Timing findings staff a week without a new scrape. Storms, permits, rating drops, license drift. Freshness has a clock.

Unknown sources stay unknown even when a VP wants a green dashboard for a QBR. Agency size is not a license to fake Yelp depth.

If you operate two legal agencies, Monday starts by asking which tenant you are in. Wrong tenant is the mix-up, just earlier.

What does Wednesday look like when several pods attach DEMOs?

Wednesday each pod diagnoses its own short list. Atlas profiles should be hard to grab from the wrong brand folder. A shared "default cover" that still says the other agency's name is an isolation bug in human form.

Forge DEMOs stay on the business uuid. Cloning a site and forgetting the NAP is how size creates one-star reviews you will then try to collect.

Relay domains stay as tight as the contract. ComplianceGate brands every send. Supervised remains the default even if you have an "ops" seat. Volume is Instantly's win, not a reason to lift our gate.

Meetings stay on marketer calendars. Client booking grids stay isolated. Size increases the odds someone pastes the wrong link. The product keeps two grids. Humans have to pick one.

What does Friday look like when Steward is the system of record after the close?

Friday won deals leave hunt boards. Health is a band. Upsells are leftover findings. Invoices, when on, are marketer to client. Portals open one shop each.

A large book will want exports. Exports are how isolation dies in email. If you export, treat the file like a secret. The product's job is to make the export unnecessary for daily work. Offboarding is the exception, not a Friday habit.

Echo and Muse at size need the same locks as at one seat, plus training. A night team in another timezone is still not allowed to gate collect or cold-text a scraped list.

If Friday's QBR needs rank charts, bring BrightLocal. If it needs a full funnel tour, bring GoHighLevel. If it needs the finding, the DEMO, and the leftover, stay here.

Which compliance lock matters most on an agency-size contract?

Isolation, the sending domain, and the collect-page lock, in that order, unless the book is medical. Medical adds the EHR and HIPAA refusal. Say it in the MSA conversation. We are not an EHR. We do not claim HIPAA.

Cold voice and cold SMS stay locked at any headcount. A larger agency asking for a dialer is asking for a different vendor.

Compliance fixers stay drafts. Size does not turn a model into counsel.

Security page language stays conservative. No invented SOC 2. No invented residency. Report vulnerabilities through the channel that page names when it is filled.

How should agency procurement compare Nerxus One to GoHighLevel?

Open with their win. GoHighLevel is the CRM and funnel OS you grow into after a client exists. Snapshots, agency billing theater, and a long marketplace are real advantages for a book that already lives there.

We close the part they leave open: detect a local leak, prove it, attach a working sample, fulfill, then manage the same uuid. We do not start at the signed agreement.

If your evaluation is "replace GHL this quarter," you are probably wrong. If your evaluation is "stop buying a list tool, an auditor, a builder, and a mailer to feed GHL," you are in the right aisle.

White-label in GHL is one reason people stay. White-label here is not in v1. If resale is the only requirement, we will lose that row and we will say so.

What should an agency-size proposal refuse to include this quarter?

Refuse sub-account resale, invented plan prices, HIPAA coverage, SOC 2 badges, live Yelp replies if assisted, overnight playbooks already in production, and cold channels.

Refuse a guarantee that isolation means you can share one tenant across two legal brands safely. The architecture isolates tenants, not magic inside a shared org.

Refuse to put client revenue on health surfaces. Bands only.

Refuse to call us an AI SDR in the RFP response. Corrections belong on /ai-info. Use them.

How do meters and seats likely show up without a dollar table?

The bill's shape is subscription plus meters: scans, diagnostics, builds, voice minutes, emails. Two purses may exist later: data versus action. Only if product agrees. This page will not invent the split as policy.

Agency extras in the pricing shell include seats, isolated client workspaces when we ship them, and an audit log. "When we ship them" is doing work. Workspaces as resale are not v1.

Overage behavior, free tier, and cancel rules wait on real policy. Do not write them into a client MSA as if they were live.

Waitlist and /demo are the doors. Access opens in batches. Volume evaluation is a twenty-minute loop, not a slide.

When should an agency-size buyer stay on specialist stacks?

Stay on Birdeye or Climbo if the contract is enterprise reviews and nothing else. Stay on Synthflow if voice is the firm. Stay on BrightLocal if the contract is rank PDFs. Stay on Semrush if the contract is broad SEO research. Stay on Apollo if the contract is LinkedIn titles.

Stay on GHL if the book is load-bearing there and you only wanted a hunt. Adding us in front is allowed. Ripping out on a promise of isolation is optional and often unwise.

Come here when the size of the book makes the hop between tools into an isolation incident, not only a time tax. CSVs are how large shops mix names.

Lighter list tools still win a bake-off that ends at the spreadsheet. We do not end there.

How should an agency-size shop treat secrets and Places keys across a large book?

Per-tenant encrypted keys. Masked status in the browser. The platform does not burn a shared quota to make a trial look generous. A client's key, if the contract says it is theirs, does not live in the agency tenant as a convenience.

Rotating a key does not destroy the business uuid. Handoffs persist first. Radar and Steward keep the row. A shop that rebuilds the book because someone rotated Google is doing the second import this product exists to avoid.

Contractors do not get the raw secret in Slack. Seats see status. If you need a hunter in another country, they hunt inside the tenant, not inside a pasted .env.

Two legal brands mean two tenants and two key rings. Saving a seat by sharing a Places key across brands is how a QBR slide leaks a competitor-client's metro.

What does an audit log owe procurement that a Slack export does not?

An audit log, when it ships as an agency extra, owes who approved Relay, who deployed, who opened a portal token, and who changed the matrix. A Slack export owes whoever was in the channel and a lawyer who has to read jokes.

Until that log is a real object on /pricing extras, do not write it into an MSA as if it were live. The pricing shell already says when we ship them. That clause is doing work.

Activity in Mission Control is the honest preview: approvals and runs you can see. It is not a compliance certificate. Do not staple it to a SOC 2 sentence we have not claimed.

If procurement's only pass/fail is a badge wall, we may lose. Say so. Isolation, the gate, and the four locks are the architecture you can defend today.

How do multi-location clients work without a sub-account factory?

Each location is a business row with its own uuid, portal, and collect URL. A group can still be one legal client in Steward if that matches the contract. Collapsing listings to save clicks is how downtown sees the suburb's reviews.

Senders should be as tight as the locations' brands. One domain for twelve rooftops is convenient until a complaint lands. Then every rooftop pauses.

Watchtower diffs per row. A two-star week at one bay is not a campaign you blast to the other eleven as a success story.

If the group demands a chrome-resale console for their managers, that is v1 Not doing. You can still show portals. You cannot show a fake product factory. Put that in the proposal so procurement does not hear a different story than delivery.

How should an agency-size QBR use reports without turning Atlas into a rank vendor?

Atlas composes branded memos with evidence and assumptions. Twenty-six report types live in the studio. Public share is a token. Credibility gate refuses bare dollars. That is a QBR door. It is not BrightLocal.

If the buyer bought rank tracking, bring the tracker. If they bought the finding and the leftover, bring the DEMO history and Steward's band. Mixing those artifacts into one slide that implies we measured position nightly is a green check from nowhere.

Health stays Healthy, Watch, or At-risk. Client revenue stays off the surface. A VP who wants a cash number gets FEE invoiced, not DEV modeled.

Schedules can refresh a report. They cannot invent a live Yelp reply or a HIPAA posture. Assisted sources stay labeled. Medical books still get the EHR refusal in the same type size as the chart.

How should an agency-size shop talk about credits without inventing a dollar table?

The bill's shape is subscription plus meters: scans, diagnostics, builds, voice minutes, emails. Two purses may exist later. This page will not invent the split as policy and will not print a floor.

Pods that share a tenant share a meter. That is a reason not to park a second legal brand in the same org to save a seat. Isolation and money will argue the same way.

Overage behavior waits on real policy. Do not write pause-versus-surprise into a client MSA as if /pricing had already locked it. Waitlist and /demo are the doors. Access opens in batches.

If a seller quotes a number they invented to win procurement, they are off script. Your FEE to the SMB is still yours. DEV stays modeled and off the invoice.

How should an agency-size shop retire a client without leaving a sender live?

Close the Pulse board. Pause Relay. Take the domain off that book if the contract ends. Portal tokens die. Watchtower can stop on that uuid. A retired shop that still receives a sequence from last quarter is how you inherit a complaint after the invoice stops.

Secrets that were theirs leave with the contract. Do not keep a client's Places key in the agency tenant as a souvenir. The uuid can remain in Steward as a closed record. The key should not.

Exports at offboarding are how isolation dies in email. If you must export, treat the file like a secret and delete the shared link. The product's job was to make daily exports unnecessary. Offboarding is the exception, not a Friday habit.

If the client moves to another agency, they take their site and their reviews. They do not take your other book. A portal token that still opens after you said goodbye is a defect. Treat it like one.

Steward can keep the closed record so you do not re-hunt them next quarter. Exclusions in the ICP should name them. A retired client who reappears on a No-Booking board is a junior without a filter. Isolation includes time, not only space. A leftover finding on a closed uuid is an upsell to a current client, not a reason to reopen a dead sender. Keep the wall.

How does an agency-size team join the waitlist without buying a resale fantasy?

Say you are an agency, approximate book count, whether you need /demo, and that you understand v1 has no sub-account resale. Batches open. No dollar.

Read Agencies, ten-clients, guardrails, and security while you wait. Write the tenant map: one org per legal brand.

When access opens, provision Clerk orgs the way you want walls, connect keys per contract, set named approvers, run one pod through the loop before you roll ten.

If you only needed roles, you wanted Small team. If you only needed the mix-up story, you wanted Agencies. If you needed the sentence isolation is the feature and resale is not, you are on the right page.

What should an agency-size RFP packet include and exclude?

An agency-size RFP packet should include tenant isolation, Clerk organization as the unit, encrypted per-tenant secrets, ComplianceGate, the four locks, Supervised as default, a read-only portal scoped to one business, and the honest Not doing row for white-label sub-account resale. It should exclude SOC 2 badges, HIPAA coverage, invented plan dollars, live Yelp replies if assisted, and overnight playbooks already running in production.

Procurement hears walls. A franchise lead hears we will not give your neighbor's reviews to your manager by accident. They do not hear your own white-label Nerxus. A seller who quietly unsays that last sentence is off script. Print the roadmap Not doing list: resale, autonomous cold voice, cold SMS, review gating, invented prices.

Compare GoHighLevel in the same packet. Open with their win. Mature CRM, funnels, snapshots, agency billing theater. We close the part they leave open: detect a local leak, prove it, attach a working DEMO, fulfill, then manage the same uuid. If the evaluation is replace GHL this quarter, you are probably wrong. If it is stop buying a list tool, an auditor, a builder, and a mailer to feed GHL, you are in the right aisle.

If procurement's only pass/fail is a badge wall, we may lose. Say so. Isolation, the gate, and the four locks are the architecture you can defend today. Audit log language belongs in extras when commercial policy lands. Do not staple Mission Control activity to a SOC 2 sentence we have not claimed.

Voice stays inbound. Scout never contacts. DEMO is a Forge working sample. Health stays Healthy, Watch, or At-risk. Client revenue stays off every surface in the packet. DEV stays modeled. FEE is what you invoice. Mixing those in an RFP response is how a VP staffs the wrong week.

How should two legal brands map to Clerk organizations?

Two legal brands should map to two Clerk organizations because a tenant holds one organization's leads, keys, senders, diagnostics, DEMOs, sequences, and portals. Saving a seat by parking a second brand in the same org is how the mix-up arrives early. Isolation isolates tenants, not magic inside a shared org.

Write the tenant map before the first hunt. One org per legal brand. Named approvers for Relay and DEPLOY in each. Places keys as the contract allows. A client's key, if the contract says it is theirs, does not live in the agency tenant as a convenience. Rotating a key does not destroy the business uuid. Handoffs persist first.

Pods that share a tenant share a meter. That is a second reason not to merge brands to look cheaper on a waitlist form. Isolation and money will argue the same way. Contractors hunt inside the tenant, not inside a pasted .env. Seats see masked status. They do not see the raw secret in Slack.

If you operate two brands and still want one Nova, you want two Novas that look the same. The roster does not fork. The wall does. Wrong tenant on Monday is the mix-up, just earlier. Ask which org you are in before you run an ICP.

What does a franchise evaluation hear that a pod does not?

A franchise evaluation hears one business per portal, one uuid per rooftop, and no chrome-resale console in v1. A pod hears which ICP they may hunt and which sender they may use. Both sentences are true. They are not the same meeting.

Each location is a business row with its own portal and collect URL. A group can still be one legal client in Steward if that matches the contract. Collapsing listings to save clicks is how downtown sees the suburb's reviews. Watchtower diffs per row. A two-star week at one bay is not a campaign you blast to the other eleven as a success story.

Senders should be as tight as the locations' brands. One domain for twelve rooftops is convenient until a complaint lands. Then every rooftop pauses. If the group demands a factory for their managers, that is v1 Not doing. You can still show portals. You cannot show a fake product. Put that in the proposal so procurement does not hear a different story than delivery.

Nova stays shared inside the tenant that owns those rooftops. The SMB still does not sit in Nova. A regional manager can hold several magic links. They cannot hold Scout. If they want the dentist to run competitors, that person would be a second tenant. We do not sell that on this page.

How do seats and named approvers work before a dollar table exists?

Seats and named approvers work before a dollar table exists as humans who share one Nova and must not share one sender by accident. The pricing shell already says seats, isolated client workspaces when we ship them, and an audit log. When we ship them is doing work. Workspaces as resale are not v1. Dollars wait on commercial policy.

A named human owns Relay. A named human owns DEPLOY. Scout can be looser. Overnight dispatch, when the runner is real, is a principal decision. Until the runner is the open half, do not tell the packet playbooks already ran in production. Governance is live. That sentence is allowed.

Two to ten people who only needed roles wanted Small team. Agency size starts when the book needs a harder wall than a role sheet. Headcount eight with one book is still a small team. Headcount three with two legal entities is already this page. Count books, not only humans.

Overage behavior, free tier, and cancel rules wait on real policy. Do not write them into a client MSA as if they were live. Waitlist and /demo are the doors. Access opens in batches. Volume evaluation is a twenty-minute loop, not a slide about agents. If a seller quotes a number they invented to win procurement, they are off script.

What isolation tests should procurement run in a demo?

Procurement should run isolation tests in a demo by asking for two businesses, two portals, and a proof that a token from shop A cannot open shop B. Ask whether a Scout save in one org appears in another. Ask whether Relay can pick a sibling agency's verified domain. Ask whether Steward can invoice a client it cannot see. Those answers should be no.

Ask to see the four locks rendered disabled, not hidden. Cold voice, cold SMS, review gating, unverified email. Ask what happens if they type text the list into Nova. The refuse should name consent. Ask whether overnight already mailed. The answer should name dry-run and the open runner.

Ask for a SOC 2 report or a HIPAA BAA if that is their only pass. We will lose that row and we will say so. Ask for white-label sub-accounts. Not in v1. Ask for a shared platform Places key so the trial looks generous. Per-tenant keys. We will not burn a shared pot.

Bring your own Places key conversation. Unknown sources stay unknown even when a VP wants a green dashboard. Agency size is not a license to fake Yelp depth. If the demo uses a seed tenant, say seeded. Seeded run history is not proof that a night already mailed for their book.

How does an agency-size shop brief a new AE on walls they cannot see?

An agency-size shop briefs a new AE on walls they cannot see by giving them an ICP, a Pulse board, a sender they are allowed to use, and an Atlas profile that already says the right legal name. A shared default cover that still carries last year's brand is an isolation bug in human form. Nova can be the shared door. The cover cannot be the shared brain.

Run one hunt through DEMO before they get Relay. A pod that sends first and brands later will mail a finding with the other client's voice. Supervised is how a principal catches that before the domain pays. Steward templates should not include internal notes from the last book. Portals stay one business.

Teach the tenant map. If they operate two legal agencies, Monday starts by asking which org they are in. Teach export hygiene. Exports are how isolation dies in email. The product's job is to make the export unnecessary for daily work. If they export, treat the file like a secret.

Teach retirement. Close the Pulse board. Pause Relay. Take the domain off that book if the contract ends. Portal tokens die. A retired shop that still receives a sequence from last quarter is how you inherit a complaint after the invoice stops. Secrets that were theirs leave with the contract. The uuid can remain as a closed record. The key should not.