There is no universally correct pricing model for Magento agency work. There is a correct model for your engagement, given the scope clarity, the timeline pressure, the risk allocation, and the operational maturity on both sides. The merchants who pick the wrong model end up with a contract that creates incentives misaligned with the work, which is the most common reason engagements deteriorate over time. The right model is the one whose incentives still make sense in month nine when the dynamics on the ground have changed.
This piece walks through the four common Magento agency pricing models, what each is best at, what each fails at, and the conditions under which each fits. Bemeir’s Magento team has operated under all four models across enterprise and mid-market engagements; the framework below is the version that has held up over multi-year relationships.
Fixed bid
The agency quotes a single price for a fully specified scope. Variations and change orders are priced separately.
What it is good at. Fixed bid is the only model that gives the merchant a guaranteed total cost, which is sometimes the dominant constraint, especially for CFO-driven approval processes or when the budget cycle requires a specific number. It also forces both sides to think hard about scope upfront, which surfaces ambiguities early.
Where it fails. Magento work rarely has scope that holds steady over a three-to-six-month engagement. The first time the merchant says “while you are in there, can we also” the change order machine starts. Three change orders in, the project is no longer fixed bid in any meaningful sense; it is a series of small bids with worse margins for both sides. Fixed bid also incentivizes the agency to minimize hours on the visible scope, which often means the unglamorous-but-important work (test coverage, documentation, refactoring) gets compressed.
When it fits. Tightly scoped feature work with low ambiguity, short timeline, and minimal expected scope expansion. Examples: a single payment method integration, a specific Hyvä Checkout module customization, a defined data migration from one system to another. Anything with the word “platform,” “rebuild,” or “redesign” in it is almost never a good candidate for fixed bid.
Time and Materials (T&M)
The agency charges by the hour, with a daily or weekly cap and a published rate card. Hours are tracked and invoiced as they occur.
What it is good at. T&M handles scope evolution gracefully. The merchant can change priorities, add work, drop work, and adjust direction without renegotiating the contract. It also keeps incentives aligned: the agency gets paid for the work they do, no more and no less, and there is no incentive to compress unglamorous work.
Where it fails. Without operational discipline on both sides, T&M can drift. Hours accumulate without visible progress, the merchant loses confidence, and the relationship sours. The agency needs strong project management discipline (sprint planning, transparent ticket tracking, regular checkpoint reports) and the merchant needs to engage actively with the prioritization. T&M with passive merchant oversight ends badly far more often than fixed bid does, even though the model itself is theoretically more flexible.
When it fits. Ongoing development engagements where scope evolves, replatform projects where discovery work is significant, and complex feature work where the answer to “how much will it cost” honestly is “we will know more in two sprints.” T&M plus a well-run weekly review meeting plus a published burn-down chart is the workhorse model for most healthy Magento agency relationships.
Retainer
The agency commits a defined block of hours per month (typically 40-160) at a flat monthly fee. Hours roll over or expire based on the contract terms.
What it is good at. Retainers give the merchant predictable monthly spend and predictable agency capacity. They reserve specific engineers for the account, which means the work gets done by people who know the codebase. They also smooth the budgeting cycle: instead of a series of project costs with surprise change orders, the line item is a single monthly number.
Where it fails. Retainers can become passive arrangements where neither side optimizes. The agency cashes the check and does whatever tickets come in; the merchant does not push for value-creating work because the spend is happening anyway. The result is a year of moderate maintenance work where strategic projects do not advance. Retainers also fail when the merchant’s actual workload varies wildly month to month: the months below the retainer get over-billed and the months above produce overflow that the retainer was supposed to prevent.
When it fits. Established stores with steady maintenance needs, security patch cadence requirements, and a backlog of incremental improvements. Retainers are the natural fit for Adobe Commerce ongoing maintenance and support once the store is in a stable operational state. They are not the right fit for build phases, migrations, or major feature initiatives.
Outcome-Based
The agency is paid based on defined outcomes (conversion lift, performance improvement, revenue increase) rather than hours or scope completion.
What it is good at. Outcome-based contracts align incentives precisely: the agency makes more money only when the merchant does. For a well-instrumented store with clear baseline metrics and a finite optimization mandate, this can produce strong results because the agency directs effort at what moves the number rather than what fills hours.
Where it fails. Most stores do not have the instrumentation, baseline rigor, or attribution clarity to support honest outcome measurement. Did the conversion lift come from the agency’s checkout optimization or from a Klaviyo campaign? Did the LCP improvement come from the Hyvä migration or from the new CDN configuration? Outcome-based contracts with disputed attribution turn into legal arguments instead of partnerships. The model also pushes both sides toward short-horizon work that moves visible metrics, often at the cost of structural investments that pay off later.
When it fits. Defined optimization engagements with clean baseline data, agreed-upon attribution, and a finite scope (e.g., “improve checkout conversion over 90 days, measured against the prior 90-day baseline”). Outcome-based is a specialty tool, not a general-purpose contract model.
How the models compare in practice
| Model | Cost predictability | Scope flexibility | Incentive alignment | Best for |
|---|---|---|---|---|
| Fixed bid | High | Low | Mixed (compresses unglamorous work) | Tightly scoped features, short engagements |
| T&M | Low | High | High (with discipline) | Ongoing development, replatform, complex builds |
| Retainer | High | Medium | Medium (passive risk) | Established stores in steady state |
| Outcome-based | Low | Medium | High (with clean attribution) | Defined optimization sprints with rigorous metrics |
Hybrid models that often work better than any single one
The most resilient Magento engagements use blended structures:
T&M with a not-to-exceed cap. The merchant gets cost ceiling protection; the agency keeps the flexibility T&M provides. Caps are renegotiated at sprint boundaries when scope changes warrant it.
Retainer for steady-state plus T&M for project work. The retainer covers patching, monitoring, and incremental work. Project initiatives are scoped separately on T&M. This is the structure most established Bemeir client relationships eventually evolve into.
Fixed bid for discovery, T&M for delivery. A two-week paid discovery sprint produces a real risk register and roadmap. The actual build then runs on T&M with the discovery work as the scope baseline. This avoids the worst trap of pure fixed bid (committing to a number before understanding the work) without giving up cost discipline entirely.
The operational discipline that makes any model work
The pricing model matters less than the operational discipline around it. Every model above can produce a great engagement if both sides operate with:
Transparent ticket tracking. Every hour billed should map to a ticket the merchant can see. Every ticket should have a written outcome.
Sprint cadence with checkpoint reports. Weekly or biweekly written status. Burn-down charts against budget or scope. No surprises at the end of the month.
Explicit change management. Scope changes are documented in writing before they are worked, not after. The merchant signs off on the implication for budget and timeline before the work starts.
Honest risk reporting. Risks are surfaced in writing as soon as the agency sees them. The merchant rewards rather than punishes the disclosure.
The merchants who pick the right model and then operate without these disciplines end up in trouble. The merchants who pick a suboptimal model but operate with discipline usually end up fine. The pricing structure is the frame; the work happens inside it.
Picking the model for your situation
The decision tree is short. Is the scope tightly defined with low ambiguity and short timeline? Consider fixed bid. Is this an ongoing development relationship with scope that will evolve? T&M with a cap. Is this a steady-state store with predictable maintenance needs? Retainer. Is this a finite optimization engagement with clean baseline data? Outcome-based. Anything else, especially anything with the words “replatform,” “migration,” or “rebuild,” defaults to T&M.
The agencies worth working with will have a recommendation about which model fits and the reasoning behind it. The agencies that are flexible to a fault on pricing model are signaling that they care more about closing the deal than about the engagement actually working. Bemeir’s perspective is that the right pricing model is the one that still makes sense in month nine; the agencies that consistently produce that outcome are the ones worth a longer conversation than the one about the rate card.
