You hired an email agency. You paid them $1,500 a month. They sent emails. The emails looked good. The reports showed solid open rates and decent click-throughs.

Your revenue from email last quarter: $4,200.

Their invoice last quarter: $4,500.

That is not a marketing problem. That is a structural problem. And the structure — not the agency, not the copy, not the design — is the reason it happens over and over again across thousands of Shopify stores.

This article is about the economics, the psychology, and the science behind why the retainer model produces this outcome reliably — and why performance-only pricing is not just a better commercial arrangement, it is the only arrangement that is mathematically aligned with your interests as a store owner.

The Principal-Agent Problem: Why Retainer Agencies Are Structurally Broken

In economics, the principal-agent problem describes what happens when one party (the agent) is hired to act in the interests of another (the principal), but the two parties have different incentives. The agent does not automatically do what is best for the principal — they do what is best for themselves, within the constraints of the contract.

It is one of the most studied problems in economic theory, documented formally by economists Michael Jensen and William Meckling in their 1976 paper on agency costs. Every major governance failure in corporate history — from Enron to the 2008 financial crisis — has a principal-agent problem at its core.

Email marketing agencies operating on retainer have a textbook principal-agent problem.

How the Misalignment Manifests in Practice

Consider three decisions a retainer agency makes every month:

Decision 1: How many flows to build. Building a complex, well-segmented win-back flow takes 15–20 hours of skilled work. For a retainer agency, that is a significant cost against a fixed monthly fee. The rational economic behaviour — from the agency’s perspective — is to build the minimum viable flow that satisfies the client enough to renew. Not the flow that maximises your revenue.

Decision 2: What to report on. Revenue attribution requires deep Klaviyo analysis, honest interpretation of what is working, and the courage to report underperformance accurately. Open rates are easy to calculate, always look acceptable, and require no difficult conversations. The rational behaviour under a retainer is to report what looks best, not what is most useful.

Decision 3: How aggressively to test and iterate. Aggressive A/B testing and flow restructuring require time and risk. If a rebuilt flow underperforms during the testing period, the agency has a difficult client conversation. Under a retainer, the rational behaviour is to leave working-ish things alone rather than risk a temporary dip. This is how accounts stagnate.

None of these decisions require the agency to be dishonest or incompetent. They are the natural output of a correctly-functioning incentive structure pointed in the wrong direction. The retainer model produces mediocrity not through malice but through mathematics.

The Mathematics of Retainer vs. Performance: What You Actually Pay

Let’s make this concrete. Most Shopify stores paying a retainer email agency are in one of three situations:

The performance model row is not a projection. It is the benchmark for what a properly structured email operation generates for a store doing $50K/month with an engaged list, correct flows, and competent execution. The industry standard for email revenue contribution in a well-run Shopify store is 25–40% of total revenue.

The gap between 7% (common retainer outcome) and 30% (performance benchmark) on a $50K/month store is $11,500/month. Annualised: $138,000. That is the cost of the misaligned incentive structure — not the retainer fee, not the design budget. The lost revenue.

Why Email Is the Right Channel to Apply Performance Pricing To

Performance pricing exists in many marketing contexts — affiliate marketing, paid search, influencer deals. But email is uniquely suited to it for four structural reasons that do not apply to other channels.

1. Revenue Attribution Is Exact

In paid advertising, attribution is contested. A Facebook ad might have contributed to a purchase that was closed by a Google search. Multi-touch attribution models exist precisely because no single paid channel can claim clean ownership of a conversion.

Email attribution in Klaviyo is different. Klaviyo tracks email opens and clicks, then attributes purchases made within a defined window (typically 5 days for clicks, 1 day for opens) directly to the email that preceded them. The revenue number is specific, verifiable, and not subject to interpretation.

This makes email the cleanest channel for performance pricing. The number is not estimated — it is reported.

2. Email Costs Scale With List Size, Not With Revenue

When a paid advertising agency generates more revenue for you through ads, their costs scale accordingly — more ad spend means more budget to manage, more creative to produce, more bidding to optimise. Their cost base genuinely grows.

For email, the cost structure is fundamentally different. The marginal cost of sending one more email to an already-built flow is effectively zero. Once flows are built and segments are structured, the infrastructure generates revenue without proportional additional cost. This means an agency can offer performance pricing without their costs becoming unmanageable as your revenue grows.

3. The Asset Belongs to You

Your email list is owned infrastructure. Unlike paid traffic — which disappears the moment you stop spending — your list persists. Your flows persist. The automation you build this month generates revenue next month, next quarter, next year, independent of ongoing ad spend.

This permanence means the value of well-built email infrastructure compounds over time. A performance model aligns the agency’s incentive with building infrastructure that generates sustainably — not just infrastructure that looks impressive in month one.

4. The Science of Email Engagement Is Well-Established

Email marketing has over two decades of controlled research behind it. We know, with statistical confidence, what works. The optimal send windows. The segmentation logic that outperforms broadcast sends. The flow timing that maximises recovery rates. The subject line structures that drive opens for specific audience types.

This knowledge base means performance is not a gamble — it is an engineering problem. A performance-only agency is not betting on luck. They are applying known mechanisms to a measurable outcome. The risk in performance pricing is not “will email work?” — it is “will this agency execute it competently?” That is a much smaller risk.

Incentive Alignment: The Science of Why It Changes Everything

The concept of incentive alignment has been studied across economics, organisational psychology, and behavioural science for decades. The consistent finding: when two parties share the same outcome metric, they make better decisions together than when their metrics diverge.

The seminal work by Jensen and Meckling (1976) on agency costs quantified the economic loss created when agent incentives diverge from principal objectives. In their model, agency costs include:

• →Monitoring costs — what you spend trying to verify the agency is doing what they said (your time reviewing reports, asking questions, managing the relationship)
• →Bonding costs — what the agency spends on reports, calls, and updates to demonstrate they are delivering (their overhead that you ultimately pay for)
• →Residual loss — the value lost because the agent’s decisions are not perfectly aligned with the principal’s interests, even after all the monitoring and bonding

In a retainer model, all three categories of agency cost are high. In a performance model, monitoring costs collapse (the revenue number is the report), bonding costs collapse (the agency has no incentive to produce elaborate reports defending mediocre results), and residual loss shrinks dramatically because the agent’s financial incentive is now identical to the principal’s.

What Alignment Does to Decision-Making Quality

Consider the same decision — whether to rebuild an underperforming abandoned cart flow — under each model:

The same people, the same knowledge, the same tools — different incentive structure, completely different decision. This is what incentive alignment does. It does not make agencies smarter. It makes their self-interest point in the same direction as yours.

What a Performance-Only Email Operation Actually Looks Like

Theory is useful. Specifics are more useful. Here is what a performance-only email engagement looks like in practice — the exact scope, the accountability structure, and the economics at each stage.

Stage 1: The Audit (Before Any Work Begins)

A legitimate performance-only operation starts with a full Klaviyo account audit before agreeing to terms. This is not a sales pitch — it is a diagnosis. The audit identifies:

The audit output is a specific dollar figure — the estimated monthly revenue gap. That number becomes the basis of the engagement. Both parties know what the opportunity is before any work begins.

Stage 2: The Build (Infrastructure First, Revenue Second)

The first 30 days of a performance engagement are infrastructure-heavy. Flows are built in revenue-impact order. Segmentation is structured. Deliverability is assessed and corrected if needed. Campaigns are planned against a calendar tied to purchase moments, not arbitrary send schedules.

None of this generates an invoice. The agency absorbs this cost entirely. Their calculation: if the infrastructure is built correctly, the revenue will follow. If it does not, they have lost the investment. That risk is not abstract — it is the mechanism that ensures quality of execution.

Stage 3: The Revenue and the Invoice

Once flows are live and generating, Klaviyo’s revenue attribution tracks what email is producing. The agency earns 10% of that number. The accountability is built into the payment structure — every month, the invoice is a direct function of performance. No revenue means no invoice. Underperformance means lower income for the agency, not just a difficult conversation.

The Three Objections Every Founder Raises — and the Direct Answers

Performance-only pricing sounds straightforward in theory. In practice, founders have specific concerns. These are the three we hear most often, and the direct answers to each.

Objection 1: “What if you just send a lot of emails to inflate revenue attribution?”

This is the most common concern and it is a legitimate one. If an agency earns 10% of attributed revenue, what stops them from sending aggressive, high-volume campaigns that extract short-term revenue at the cost of long-term list health?

The answer is structural: list decay is self-defeating under a performance model. An agency that burns your list through over-sending destroys the asset that generates their income. Deliverability drops. Open rates collapse. Unsubscribes increase. Revenue declines. Their 10% of a declining number becomes 10% of very little, very fast.

Under a retainer, an agency that burns your list loses one client. Under performance pricing, an agency that burns your list loses their income stream immediately and permanently. The incentive to protect list health is more powerful under performance pricing than under any other model.

Objection 2: “What if my list is too small for this to work?”

List size is less important than list quality and infrastructure. We have seen stores with 3,000 engaged subscribers outperform stores with 15,000 cold ones. The variables that determine email revenue are not primarily list size — they are segmentation quality, flow completeness, and send relevance.

A store with 2,500 highly engaged subscribers, a properly built abandoned cart flow, and a segmented welcome series will consistently generate more email revenue than a store with 20,000 subscribers, no flows, and broadcast campaigns.

Objection 3: “How do I know the revenue attribution is accurate?”

Klaviyo’s revenue attribution is the industry standard for Shopify email tracking. It attributes purchases to specific emails within a defined conversion window (the default is a 5-day click window and a 1-day open window). Both the store owner and the agency see the same Klaviyo dashboard — there is no separate tracking system controlled by the agency.

The attribution window can be reviewed, adjusted, and agreed upon at the start of the engagement. Both parties look at the same data. The invoice is generated from the same number the store owner sees in their own Klaviyo account. There is no black box.

The Five Questions to Ask Any Retainer Agency Before Your Next Invoice

If you are currently paying a retainer for email marketing, these five questions will tell you everything you need to know about whether that arrangement is working in your interest.

What to Expect in the First 90 Days of a Performance Engagement

The performance model does not produce revenue on day one. Infrastructure takes time to build. Here is what a realistic 90-day trajectory looks like and why.

By day 90, a store that entered with zero flows and an untouched list should be generating 20–35% of total revenue from email. The exact number depends on list size, list quality, product category, and average order value. But the direction is consistent: infrastructure built correctly compounds. Every week a flow runs, it generates data. Every data point informs the next optimisation. Revenue grows not because campaigns get louder but because targeting gets sharper.