Charge for sends, bundle everything an agency resells, and publish the cost of everything you pass through. The meter is the one agencies already use to compare tools (Pricing landscape). The things the leaders meter separately (client workspaces at Smartlead, AI replies at Instantly) become the reasons to switch. Price is not the pitch. Bootstrapped leaders can match any list price (Cloning the leaders). The pitch is what a 10-client agency pays in total, and what it no longer has to argue about.
Every price on this page is a proposal for testing with design partners, not a validated price. Figures labelled "my estimate" come from the hub's cost models (Unit economics, Build costs and team), which are themselves engineering estimates.
What is metered, bundled and sold as add-on
| Treatment | Items | Why |
|---|---|---|
| Metered | Emails sent per month | The unit agencies compare; bounds ingestion and IP load per euro of revenue (Unit economics) |
| Bundled, unlimited | Client workspaces, seats, mailboxes, client-viewer and approver roles, reply classification and AI drafts with approval, objection list, country rule engine, Art. 14 notices, mailbox health, importers, API and webhooks | Each is either a grievance against a leader or core to The wedge |
| Bundled, capped | Verification: one contact per three sends in the plan | Woodpecker already includes it (Built-in email verification); capped because it is the largest variable cost below |
| Add-on at published cost-plus | Inboxes via partner API, domains, pay-on-valid enrichment, seed placement tests | Every supplier price is public; hidden markup is Smartlead's "Powered by Zapmail" weakness (Transparent infrastructure) |
| Optional outcome plan | Per held meeting, on the desk | Hypothesis, not default (The reply desk) |
| Never | Per-client fees, per-mailbox fees, AI reply credits, auto-activated add-ons | The 1-star themes in Customer voice |
The price architecture
EUR prices exclude VAT and are invoiced from the EU entity. USD prices are set separately, not converted daily.
Release A: the desk overlay (from December 2026)
| Plan | EUR | USD | What it covers |
|---|---|---|---|
| Desk | €199/mo | $219/mo | Unlimited client workspaces, seats, AI drafts; Smartlead, Instantly and EmailBison connectors; client portal and reports |
| Desk, founding partner | €149/mo, locked 24 months | $159/mo | First 20 design partners (Go-to-market plan) |
| Desk Outcome (optional) | €79/mo + a fee per held meeting | $89/mo + fee | Fee to be set from interviews: test €5, €10 and €15 |
Release B: the agency platform, sender plus desk (from mid-2027)
| Plan | EUR | USD | Sends/mo | $/1,000 at cap | Extras |
|---|---|---|---|---|---|
| Studio | €79 | $89 | 50,000 | $1.78 | Desk, rule engine, email support |
| Agency | €229 | $249 | 200,000 | $1.25 | + compliance evidence export, shared Slack |
| Scale | €449 | $499 | 500,000 | $1.00 | + named deliverability engineer, priority migration |
| Extra volume | €79 | $89 | per 100,000 | $0.89 | On any plan, prorated |
Annual billing gives two months free, about 17%, the same as Smartlead's discount. Desk customers moving to the sender get their desk fee credited.
No vendor publishes pay-per-meeting software pricing (Business models), and agency per-meeting rates could not be verified: CIENCE prices "from ROI goals" and Belkins hides its prices (Lead-gen agencies). "Held" can be disputed by the agency's client. Keep the outcome plan optional, cap the monthly fee, and drop it if disputes pass 30% of billed meetings (K5 in Kill criteria).
Against the leaders' current prices
List prices as of October 2026, monthly billing.
| Product | Price | Sends | $/1,000 | Client workspaces | Reply AI |
|---|---|---|---|---|---|
| Instantly Hypergrowth | $97 | 125,000 | $0.78 | No client scoping inside a workspace | Not included |
| Instantly Scale bundle | $194 | 100,000 | $1.94 | Same | AI Reply Agent, 5 credits/reply |
| Instantly Light Speed | $358 | 500,000 | $0.72 | Same | Not included |
| Smartlead Unlimited Smart | $174 | 150,000 | $1.16 | $29 each, add-on | SmartAgents (workflows) |
| Smartlead Unlimited Prime | $379 | 500,000 | $0.76 | 3 included, then $29 | Same |
| EmailBison | $599 | 500,000 | $1.20 | Unlimited | None documented |
| PlusVibe Agency | from $497 | 500,000+ | ≤$0.99 | Isolated backend per client | AI reply agent |
| Us, Agency | $249 | 200,000 | $1.25 | Unlimited, scoped | Unlimited drafts, approval |
| Us, Scale | $499 | 500,000 | $1.00 | Unlimited, scoped | Same |
At the cap we are dearer per 1,000 than Instantly and Smartlead. That is deliberate. At 500k sends the sequencer is $0.72 per 1,000 on Instantly Light Speed, out of roughly $10–15 fully loaded, so halving software price saves the customer about 2.5% (Unit economics). Racing to $0.72 buys nothing. The agency's real comparison is the total for its client count:
Smartlead: Prime $379 (Smart caps at 150k) plus 7 extra workspaces at $29 = $582. Instantly: Light Speed $358, with no way to restrict a client to its own campaigns (help) and AI replies extra. EmailBison: $599, no reply AI. Us: $249 with the desk included. At 20 clients Smartlead becomes $872; ours stays $249.
Margin math
My estimates, per plan at full use. Assumptions from Unit economics: ~660 sends per mailbox-month plus 30% spare; replies at Instantly's 3.43%; AI drafts on half of replies. Unit costs: infrastructure at $50–200 per 1,000 mailboxes a month (the Build costs and team estimate); classification ~$0.0009 and a Haiku draft ~$0.0023 at Claude list prices; verification at MillionVerifier's $449 per million volume price. No warmup network and no owned IPs, so neither cost line exists.
| Line (USD/mo) | Studio $89 | Agency $249 | Scale $499 |
|---|---|---|---|
| Mailboxes served | ~100 | ~390 | ~985 |
| Infrastructure | 5–20 | 20–80 | 50–200 |
| Classification + drafts | 3–4 | 14 | 35 |
| Verification (sends ÷ 3) | 8 | 30 | 75 |
| Direct COGS | 16–32 | 64–124 | 160–310 |
| Gross margin | 64–82% | 50–74% | 38–68% |
Three conclusions. Infrastructure decides margin: push-based ingestion (Gmail and Graph notifications, not tight IMAP polling) is the difference between the top and bottom of each range (Sending architecture). Verification is the biggest variable line at volume, which is why it is capped at sends ÷ 3 and bought at volume rates; small-volume verification prices are several times higher (Bouncer at $4–8 per 1,000). Support sits on top: at one person per ~200 customers (Build costs and team) and an assumed €90k loaded cost, that is roughly €37 per customer a month, which makes Studio a funnel, not a profit centre.
Inbox resale is the second P&L line. Partner supply at $2.50–3.00 an inbox resold at a published $3.20 (€3.00) earns 6–22%. That margin is thin but visible, against Smartlead's $4.50 and Instantly's $5 (Infrastructure strategy: build or partner).
Affiliate commission design
| Term | Us | Instantly | Smartlead |
|---|---|---|---|
| Rate | 30% flat from the first referral | 20%, 30% from 51st, 40% from 76th | 15%; 20/30/35% at 20/50/75 a month |
| Duration | Lifetime | Recurring | Lifetime |
| Paid from | Referral's 3rd paid month (months 1–2 accrued, released then) | Weekly, 30-day ageing | ~45 days after trial→paid |
| Base | Subscription only, not inboxes or data | Plan | Plan |
| Cookie | 90 days | Not stated | 90 days |
Sources: Instantly, Smartlead, Distribution.
Paying from month three rewards creators whose referrals stay, as The info economy recommends, and protects against the ~10% early monthly churn of guru traffic. Excluding pass-through infra keeps commission off a 6–22% margin line. The margin check: on an Agency plan, 30% commission leaves $174, minus $64–124 COGS and ~$40 support, so a referred customer contributes roughly $10–70 a month (my estimate). The channel only works at the low end of infrastructure cost, so it opens in H2 2027 after the engine is tuned (K10 in Kill criteria).
What this means for an entrant
- Publish the 10-client total next to the $/1,000 table. The per-client fee is where Smartlead loses. The per-1,000 race is where you lose.
- Put a
pricing.mdand a fair-terms page live with the first invoice. One axis, every limit on one page, one-click cancel (Trust as positioning). - Build push-based ingestion before you scale. It is the swing factor between 38% and 68% gross margin on the Scale plan (Sending architecture).
- Test the outcome plan, do not bet on it. Offer it to design partners; make it the default only if they choose it (The first 90 days).
- Never meter compliance. The rule engine and objection list are why customers stay (EU-native compliant outbound); charging for them invites the "you are the sender" comparison you are trying to escape.