Your cost side is fixed and small: DM Champ white-label starts at a $229 one-time AppSumo purchase (Tier 3, 10 sub-accounts) and Tier 6 is $999 one-time for unlimited sub-accounts with credit reselling. That one-time cost structure is what makes every model below work: your platform cost does not scale with your client count, so almost every incremental client dollar is margin.
All client-side prices below are illustrative examples, not benchmarks or promises. High-ticket niches support the top of each range; local services the bottom.
Model 1. The First-Value Offer (get paid to start)
The mistake most agencies make is the free pilot. Free pilots attract clients who cost you a month of setup and then ghost. The first-value offer flips it: a paid, small, fast deliverable that proves value inside two weeks.
The offer: a two-week "AI setter installed" sprint at a flat $500 to $1,500 (illustrative). You connect one channel (start with WhatsApp or Instagram, where the actual conversations happen), load the client's FAQ and offer into the agent, and run it on live inbound leads. The deliverable is not a report; it is the client watching the AI qualify their own leads in their own inbox.
Why it works: speed to value builds trust, the client has paid so they show up to the handover call, and the sprint is designed so the obvious next step is Model 3 (the retainer). Track one metric during the sprint: conversations handled without a human. It sells the retainer for you.
The same AI delivering this read is one click away. Ask it anything. Push back. See if you can break it.
Model 2. The Performance Core (base plus results)
For clients with real lead volume, anchor the deal to outcomes: a modest base fee plus a per-result component. Per booked appointment, per qualified lead handed to sales, or per show, whatever the client's CRM can verify.
The offer: $300 to $500 base (illustrative) plus $10 to $50 per qualified appointment, with a cap so the client can budget. The base covers your fixed attention; the performance component means the invoice grows exactly when the client is winning.
The discipline this model demands: define "qualified" in writing before launch, and make sure attribution is checked in the client's own system so there is never an argument about whose number is real. Agencies skip this step once and never skip it again.
Model 3. Continuity (the retainer that compounds)
This is the core business. Monthly management of the AI agent: prompt tuning, new campaign flows, follow-up sequences, monthly reporting. Typical agency retainers for this run $200 to $500 per client per month, with high-ticket niches (clinics, coaches, real estate) often at $500+, as covered in the $10k MRR playbook.
On the cost side, AI usage is either platform credits you rebill through Stripe at prices you set (Tier 6 includes credit reselling) or the client's own Anthropic key via BYOK. Either way, usage is a pass-through, not a margin leak.
The compounding part: every month of conversation history makes the agent measurably better tuned to the client's objections and offers, which raises the switching cost in your favour. Churn on a working AI setter is low because turning it off means the client's DMs go dark.
Model 4. Ascension (white-label your way to a portfolio)
Once five or more clients run on your retainer, the ascension move is to stop selling a service and start selling your platform. With full white-label (your domain, your logo, your app name), clients log into your branded product. At that point you can:
- Raise the packaged price, because a branded platform reads as a product, not a reseller arrangement.
- Sell annual prepay on the platform fee.
- Sell adjacent modules: campaigns, comment-to-DM funnels, appointment booking.
This is where the one-time platform cost gets absurd in your favour: a Tier 6 purchase at $999 one-time supports an unlimited-client white-label portfolio with no recurring platform fee. Ten clients at an illustrative $400/month is $4,000 MRR on a fixed cost most agencies earn back inside the first client's first month. The ceiling on this model is real: FueGenix, the clinic DM Champ was built in, went from $5M to $10M in annual revenue with this exact AI handling its WhatsApp sales, with $5.0M in AI-attributed closed deals counted deal by deal in its CRM.
Launch tactics that carry across all four models
- Sell the sprint (Model 1) publicly; upsell the retainer (Model 3) privately at the sprint debrief, with the client's own numbers on screen.
- Never quote "AI" as the product. Quote the outcome: answered DMs at 2am, qualified leads, booked calls. The benchmark page gives you the third-party-checkable numbers to put in the deck.
- Pick one niche and stack receipts in it. Two case studies in one vertical outsell ten features every time.
- Put your money model in writing before you need it. The agencies that stall are the ones re-negotiating structure on every deal.