Fixed-fee, retainer, equity, performance — when to charge what, the published tiers we actually use, and the scripts that move pricing conversations forward instead of into a Calendly black hole.
Pricing note, 5 September 2026. Our published prices changed on this date: Operators moved from $4,950 build / $1,997 per month to $7,500 / $2,950, and Agent in a Day moved from $497 one-time to $2,500 build / $497 per month. The arithmetic below was run against the prices in force when it was written and has been left as it was rather than quietly restated. Current prices are on the pricing page.
I ran a Los Angeles social media marketing agency for three years before TrainYourAgent. I tried every pricing model on the planet — performance fees, equity slices, day rates, hourly, fixed scope, monthly retainer, revenue share. The hard-won lesson: the pricing model decides what kind of work you do, more than the work decides what pricing model you should use.
An earlier version of this paragraph opened with a revenue figure for TrainYourAgent. That figure was invented and it has been deleted, along with the post it linked to. I am not publishing an MRR number for this business until there is one I would show an auditor. What follows is the pricing structure itself, which stands on its own: the four models that survive contact with SMB customers, our published tiers, and the scripts I actually use on calls.
There are exactly four pricing structures that survive contact with real customers:
Everything else — hourly, daily, "we'll figure it out" — fails at SMB scale. Hourly punishes you for being efficient. Daily makes you a contractor. "We'll figure it out" never gets paid.
Fixed monthly fee is the default for ~80% of our customers. The math is clean for both sides: they know their bill, you know your MRR, the conversation stays on outcomes.
The tiering we use, which is the same ladder published on our pricing page:
The build fee front-loads the engineering work and filters tire-kickers. It is refundable two ways: if we do not ship inside 21 days, and under a 30-day money-back guarantee against the spec agreed on the kickoff call.
Why fixed-fee works at SMB scale:
The tradeoff: a customer doing 8x average volume on the cheapest tier is dragging your margin. You need usage caps with overage rates — ours is 5,000 minutes included, $0.40/min after. Without a cap, fixed-fee will eat you alive in the long run.
For custom work — multi-location chains, complex integrations, agency white-label setups — split the deal into:
This is what I use for clients who want something we don't ship as a standard product. The build fee covers actual engineering hours (figure $200–$300/hour loaded). The retainer covers monitoring, model updates, prompt tuning, and ~4 hours of iteration per month.
The script that closes these:
"Our Scale build is $9,950 — 50% on signing, 50% on go-live, and we commit to 21 days. After that we maintain it for $4,997/mo, which includes 25,000 minutes, the on-call rotation and iteration time. That's the floor. If your scope is bigger we'll quote it before signing."
Concrete, bounded, no ambiguity. Closes 60-70% of qualified leads. Tire-kickers self-eject when they hear "50% on signing."
Performance pricing — "$50 per booked appointment, $200 per closed deal" — sounds great on paper. It aligns your incentives with the customer's. In practice it almost never works at SMB scale.
Why:
The ONLY time I'll do performance:
Even then, I price it as a hybrid: $1,500/mo floor + $75 per booked appointment over baseline. The floor protects you. The per-appointment kicker rewards both sides for success.
Equity in a customer is a real option if all three are true:
I've taken equity twice. One went to zero. One is currently worth ~6x what the cash would have been. Expected value is positive in retrospect, but the cash compounds in your business RIGHT NOW. Equity compounds in someone else's company at someone else's pace.
Rev-share is similar but cleaner — you take 5-10% of revenue attributable to the agent for 12-24 months. Easier to track than equity, harder to argue about than performance. I've done two rev-share deals; both worked. The trick is time-box it. Open-ended rev-share is a rope.
Pricing conversations stall in the same three places. Here's what to say.
When they ask "how much?" in the first email:
"It's published, so I'll just tell you: $99/mo if you build it yourself, $1,997/mo plus a $4,950 build fee if we build it, $4,997/mo plus $9,950 for multi-location. If you've got 10 minutes I can tell you which one you actually need — that's the only part that's a conversation."
You're not being cagey. You're framing the range and pivoting to a call. Works ~70% of the time.
When they say "the budget is tight":
"Totally fair. The smallest done-for-you version we ship is $1,997/mo flat plus the build fee, and there's a $99/mo self-serve lane below that if you'd rather wire it up yourself. If the retainer is still high, what we usually do is start with one channel — voice OR chat, not both — and add the second in month 3 once you're seeing returns. Want me to scope that?"
Don't discount. Down-scope. Discounting trains them to negotiate. Down-scoping trains them to expand later.
When they say "let me think about it":
"Of course. What's the open question? If it's pricing, we can scope smaller. If it's technical fit, I can show you the same setup running for [similar customer]. If it's timing, we can start the build in 30 days. Which is it?"
This works because it asks them to be specific. ~40% of the time they tell you the real objection and you solve it in the next two sentences.
Want the full sales playbook + pricing calculator I use on every call? Grab the sales toolkit or book a 30-min call — we'll send the scripts and the pricing sheet.
Three pricing mistakes I made in the SMMA days, in case you're tempted:
AI agencies are scaling faster than ever right now because the marginal cost of an extra customer is near zero. The constraint isn't engineering — it's whether you can hold pricing.
The shops that race to the bottom on price are getting margin-compressed inside 90 days. The shops that hold a four-figure monthly retainer with a real build fee are stacking MRR at a rate that compounds. Same engineering effort, very different outcomes.
Pick a tier. Hold it. Down-scope before discounting. Front-load with build fees. Don't take performance deals without a floor. Don't take equity unless you'd take the cash anyway.
That's the playbook. If you want to talk through your specific pricing, book a call or check the agency partner program.