Why We Chose Tiered Revenue Share Over Flat Pricing

We could have charged $299 a month and called it done. Every customer pays the same, the billing page writes itself, and finance can forecast revenue in a spreadsheet with one row. That simplicity is exactly why flat pricing is the default for most SaaS companies. It's also why flat-rate models leave money on the table from high-value customers while overcharging price-sensitive ones. We decided the tradeoff wasn't worth it.

Instead, we built Palatai's pricing around a tiered revenue share model. A base subscription gets you on the platform. The variable component — where we earn more only when your agents produce more — scales in tiers that reward momentum. Here's why we made that choice and what it means for how we build product.

Flat Pricing Hides a Structural Problem

Flat pricing treats every customer as identical. The agency whose AI agents close $80K in pipeline last quarter pays the same as the one that logs in twice and forgets the password. Both accounts look identical on a revenue report, but they're fundamentally different businesses with different needs.

This creates two problems simultaneously. The high-output customer is undercharged relative to the value they receive — and eventually, they know it. The underperforming customer is overcharged relative to what they're getting, and they churn. That second group is where most SaaS companies bleed. Businesses using tiered pricing report 44% higher average revenue per user compared to flat-rate models, because they capture value from both ends of the spectrum instead of optimizing for the middle.

Worse, flat pricing gives us no signal. If every account pays $299 regardless of output, we have no built-in metric to distinguish "customer getting massive value" from "customer about to cancel." By the time the churn hits, we've already lost the window to fix it.

Why Revenue Share Alone Wasn't Enough

We already wrote about how revenue share aligns vendor incentives with customer outcomes. That reasoning still holds. When we earn more only when you earn more, we're structurally motivated to make your agents perform.

But a single-rate revenue share — say, 20% of agent-attributed revenue across the board — creates its own friction. A customer whose agents generate $2K a month and a customer generating $50K a month both get the same rate. The smaller customer is still in the early stages, figuring out which agents to activate and how to configure workflows. The larger customer has proven the model and is scaling hard. Treating them identically ignores the difference in where they are and what they need from us.

A flat revenue share also means the high-growth customer's effective cost per dollar of agent revenue stays fixed forever. There's no reward for scale. No reason for us to push their agents into the next bracket. The alignment exists, but it doesn't accelerate.

How Tiers Change the Incentive Structure

With tiered revenue share, the credit rate increases as your agents cross revenue thresholds. At the first tier, your agents generate modest revenue and you receive a baseline credit rate on your next bill. Cross into the second tier, and the rate jumps. The third tier jumps again. Your effective cost per dollar of AI-generated revenue decreases the more your agents produce.

This changes how we build product internally. Our engineering team doesn't just optimize for "agents that work." They optimize for "agents that push customers from one tier into the next." That's a fundamentally different objective. It means we care about activation speed, agent accuracy at scale, workflow depth, and the compounding effects that move an account from $5K in monthly agent revenue to $20K.

For customers, tiers create a visible progression. You can see where you stand and what it takes to reach the next bracket. Companies using outcome-aligned pricing report 23% improved retention rates over traditional subscription models — partly because the pricing itself makes progress tangible.

The Base Subscription Still Matters

Pure outcome-based pricing — pay nothing unless the agents deliver — sounds customer-friendly on paper. In practice, it makes both sides nervous. 64% of SaaS finance executives cite unpredictability as their top concern with pure value-based models. The vendor can't forecast revenue. The buyer can't forecast costs. Nobody plans confidently.

That's why we keep a base subscription. You pay a fixed amount for platform access, infrastructure, and support. The tiered revenue share sits on top. The base gives you cost predictability. The variable component gives us skin in your game. Hybrid models are quickly becoming the default for B2B AI platforms — with companies combining subscriptions and usage or outcome components reporting 38% higher revenue growth compared to single-model approaches.

The base also covers real costs. AI workloads aren't free. Compute, tokens, API calls, and model inference have variable costs that scale with usage. A pure revenue share model without a base would force us to absorb those costs during the ramp-up period when a new customer's agents are still learning. The base subscription makes that investment sustainable.

What This Looks Like in Practice

Say your agents start generating $3K in trackable revenue during your first month. You earn credits at the base tier rate, applied against your next bill. Month two, agents are dialed in — they push $8K. You cross into the second tier. The credit rate increases, and you're now earning more credit per dollar of agent output than you were last month.

By month four, your agents are handling appointment booking, lead qualification, and invoice follow-up across three departments. Agent-attributed revenue hits $18K. You're in the top tier. Your effective platform cost, net of credits, is a fraction of what a flat subscription would have been. And we're earning more than we would have on a flat model too — because your success is our upside.

If a month goes quiet — seasonal dip, you pause a workflow — the credits drop proportionally. We feel that immediately. It creates an automatic incentive for us to check in, suggest optimizations, or adjust agent configurations before a temporary dip becomes permanent disengagement.

Why This Matters for Small Businesses Specifically

SMBs don't have the luxury of paying $500 a month for an AI tool that "might" deliver ROI in six months. 43% of enterprise buyers already consider outcome-based pricing a significant purchase factor. That number is higher among smaller companies where every line item faces scrutiny from someone who also signs the checks.

Tiered revenue share makes the pitch simple: your costs scale with your results. When agents produce, credits increase. When they don't, the variable component drops. You're never paying the same amount for $0 in output as you'd pay for $50K. And the tier progression means growth gets cheaper, not more expensive.

Solo founders running seven departments with AI agents need a pricing model that adapts to their trajectory. A flat $299 doesn't account for the month they activate three new agent workflows and triple their output. Tiers do.

The Tradeoffs We Accept

Tiered revenue share is harder to build, harder to explain, and harder to forecast than flat pricing. We accept those tradeoffs because the alternatives are worse.

• Attribution complexity — We need to track which agent actions led to which revenue events. That requires deep integrations with your CRM, invoicing, and scheduling tools. If we can't show you the chain from agent action to dollar outcome, the model doesn't work. So we build that attribution layer first, not as an afterthought.
• Tier design risk — Set the thresholds wrong and customers feel punished instead of rewarded. We calibrate tiers based on real usage data, not guesses. The boundaries shift as we learn more about how agents perform across different industries and company sizes.
• Revenue volatility — Our earnings fluctuate with customer outcomes. A bad month for your agents is a bad month for our revenue. We manage this with the base subscription providing a floor, but the variable component means our upside and downside are genuinely shared. That's the point.
• Explanation overhead — "Pay $299 a month" fits on a billboard. "Pay a base subscription plus a tiered revenue share with increasing credit rates" needs a pricing page with clear breakdowns. We invest in that clarity because the customers who understand the model become our strongest advocates.

Where Flat Pricing Still Makes Sense

Flat pricing works for tools that don't generate measurable revenue. If your software is a productivity layer — note-taking, project management, file storage — the connection between the tool and a revenue outcome is too indirect to track reliably. Per-seat pricing makes sense there.

It also works in the earliest stages of a product when the data to build tier thresholds doesn't exist yet. Bessemer Venture Partners notes that companies transitioning to outcome-based models typically experience 15-20% short-term revenue variance before stabilization. If you can't absorb that variance, flat pricing buys you time.

But when AI agents are doing the work — closing deals, booking appointments, recovering revenue — the output is measurable. And when the output is measurable, flat pricing stops being simple. It starts being misaligned.

The Bet We're Making

Tiered revenue share is a bet that our agents will perform. Every month, we are financially accountable for whether your AI team produced results. If they didn't, we earn less. If they did, we earn more — and so do you, at an increasing rate.

That bet changes how we hire, how we prioritize our roadmap, and how we staff support. We don't measure success by monthly active users or feature adoption rates. We measure it by agent-attributed revenue across our customer base. That number going up is the only metric that matters.

Want to see how the tiers work for your business? Start a 30-day pilot and watch your agents earn their first credits before you commit to a plan.