How to Build a 90-Day Media Plan That Accounts for Seasonality, Creative Refresh, and Budget Pacing

Most eCommerce brands do not have a media plan. They have campaigns and a budget.

Decisions happen reactively: scale when ROAS looks good, pull back when it drops, swap creative when performance tanks. There is no forward structure. No anticipation of what is coming. No framework for why the budget is allocated the way it is or what the signal should look like at the end of the period.

That is not media buying. That is firefighting with a credit card.

A properly built 90-day paid media plan for eCommerce is an operating document. It accounts for where the business is in the calendar, what the creative pipeline looks like, how budget should shift across the three phases of the quarter, and what the measurement benchmarks are at each stage. It is built before the quarter starts, not assembled in hindsight to explain what happened.

Image brief: Six-row seasonality table — Planning Period, CPM Trend, Budget Posture, Creative Priority. Q4 Peak row highlighted. alt: "90-day eCommerce media plan seasonality guide by quarter." caption: "CPMs are not constant. The plan that builds seasonality into budget allocation before the quarter starts will outperform the one that reacts after the fact."

Why 90 Days Is the Right Planning Horizon

Thirty-day plans are too short. Creative fatigue compounds before you can address it. Seasonality impacts arrive before you have built the creative or confirmed the budget. Budget decisions are made in isolation rather than in the context of what the next sixty days require.

Annual plans are too long. Channel dynamics shift, platform algorithms change, and creative performance data that seemed predictive in January is irrelevant by October. A plan built twelve months out treats the paid media environment as more stable than it is.

Ninety days is the window where you can see enough of the business calendar to plan intelligently while maintaining enough flexibility to adapt to what the data actually shows. It is also the right accountability unit for any agency-client relationship — it sets shared expectations before spend begins and creates a framework for evaluating whether the quarter worked.

The Three Phases of the 90-Day Plan

Phase 1: Foundation and Learning (Days 1 to 30)

The first thirty days are not primarily about revenue. They are about establishing the data foundation the next sixty days will be built on.

This phase is where you run creative tests, validate audience assumptions, and confirm that the tracking infrastructure is capturing purchase events accurately. For brands with existing accounts, this phase is about auditing what is running and cleaning up the signal before scaling.

See the four-stage funnel audit that should happen at or before Phase 1 of any new plan — the audit output is the Phase 1 starting point. If attribution is broken or traffic quality is misaligned, Phase 2 spend will compound those problems rather than generate clean performance data.

The specific work in Phase 1:

Pixel and Conversions API verification. Confirm that purchase events, add-to-cart events, and initiate-checkout events are firing correctly across desktop and mobile. A misconfigured event at this stage corrupts the algorithm's optimization signal for the entire ninety days.

Creative testing matrix. Five to ten new creative concepts tested with equal budget allocation and a defined promotion threshold. The output is a ranked creative list that forms the basis of Phase 2 scale campaigns. The threshold for promotion to scale needs to be defined before testing starts — cost per purchase relative to account average, minimum hook rate, or CTR threshold. Without a defined threshold, promotion decisions are made subjectively and the test produces no actionable learning.

Baseline MER calculation. Total revenue divided by total ad spend for the period immediately before the plan launches. No attribution model, no platform filters. This is the number you will measure improvement against at the close of the ninety days.

Seasonality mapping. Pull historical data for the same window in prior years. Identify peak weeks, demand dips, and any promotional events that require budget adjustment. Map these against the calendar before any spend decisions are made for Phase 2.

Budget posture in Phase 1: conservative, typically 20 to 25 percent of total quarterly spend, weighted toward learning rather than acquisition.

Phase 2: Scale and Optimization (Days 31 to 60)

Phase 2 is where the plan generates revenue. The creative winners from Phase 1 move into scale campaigns. Budget increases. The algorithm has enough conversion signal to optimize toward purchase events rather than proxy metrics.

The most common mistake in this phase is scaling budget faster than the creative pipeline can support. Budget increases should be paired with creative replenishment. See how creative fatigue signals at scale appear before the CPA increase becomes visible — and why the Phase 2 creative calendar needs to be running in advance of the fatigue window — the fatigue timeline at significant spend levels is typically two to four weeks, which means the creative brief for Phase 2's replacement assets needs to be in motion by day 40, not day 55.

Platform-specific Phase 2 logic:

On Meta, Phase 2 is where broad audience targeting with proven creative can compound efficiently. Advantage Plus Shopping Campaigns or consolidated campaign structures that allow Meta's algorithm maximum latitude work best when the creative quality has been validated in Phase 1 testing. The algorithm needs volume to optimize, and Phase 2 is where that volume threshold is typically crossed.

On Google, Phase 2 benefits from the conversion history built in Phase 1. Smart bidding strategies require a minimum of thirty to fifty conversions in a thirty-day window to optimize reliably — Phase 1 builds that history so Phase 2 can use it.

Budget posture in Phase 2: aggressive, typically 55 to 65 percent of quarterly spend, weighted toward proven campaigns.

Phase 3: Defend and Prepare (Days 61 to 90)

Phase 3 serves two functions that most brands miss.

The first is defending Phase 2 performance gains. The creative that scaled in Phase 2 will begin losing efficiency by week eight or nine. Phase 3 is where the next creative testing cycle needs to be running in parallel with scale campaigns — not scheduled for next quarter.

The second is preparing the infrastructure for the following 90-day plan. Budget recommendations for the next period, creative briefs for the next testing cycle, and any structural account changes the current period's data suggests should be documented before the close. A 90-day plan that ends without a clear input into the next plan is a project. One that feeds the next one is a system.

Budget posture in Phase 3: decreasing acquisition spend, increasing testing reinvestment.

The Seasonality Layer

Most 90-day plans handle seasonality poorly — either ignoring it or reacting only after performance has already shifted. The plan built in advance accounts for seasonal inputs before any budget is allocated.

| Planning Period | CPM Trend | Budget Posture | Creative Priority | |---|---|---|---| | Post-holiday (January) | Declining | Moderate, test-heavy | Fresh hooks, new year framing | | Q1 shoulder (February–March) | Stable | Scale Phase 1 winners | Iteration on proven creative | | Spring peak (April–May) | Rising | Aggressive if category relevant | Category-specific seasonal hooks | | Summer shoulder (June–August) | Variable by category | Conservative or stable | UGC and lifestyle formats | | Pre-Q4 (September–October) | Rising | Build creative library and scale | Q4 hooks, gift-framing creative | | Q4 peak (November–December) | Highest of year | Maximum, pre-approved plan | Promotional, urgency-driven |

Known demand peaks require pre-built creative, confirmed budget above baseline, and a measurement plan for comparing peak-period performance against the 90-day baseline. If the plan does not specify what "success" looks like during a demand spike before the spike occurs, the evaluation happens in hindsight against arbitrary expectations.

Known demand troughs are not failures. They are predictable conditions that should be planned for, not reacted to. The summer shoulder or the post-holiday January dip are the right windows for creative testing investment and audience learning — the organic demand is lower but the CPMs are more favorable and the test results are less contaminated by demand-driven conversion spikes.

Platform cost cycles are predictable. CPMs on Meta and Google increase around major promotional moments as advertiser competition rises. The plan needs to account for rising acquisition costs during these windows and adjust ROAS or CPA targets accordingly — treating elevated CPMs as an account performance problem rather than a predictable seasonal cost input is one of the most consistent misdiagnoses in performance marketing.

Creative Refresh Cadence: The Calendar Most Brands Do Not Have

Creative fatigue is the most predictable performance problem in eCommerce paid media and the least planned for.

At significant spend levels, creative starts showing frequency saturation signals within two to four weeks. Click-through rate declines, cost per purchase rises, and if frequency metrics are not being monitored, the problem is diagnosed as a bidding issue or an audience issue when it is a creative issue.

The 90-day plan needs a creative calendar embedded inside it. Not a separate function that the media team requests when performance drops, but a forward-looking schedule that specifies:

When new creative concepts enter testing — this should happen on a rolling basis throughout all three phases.

Which formats are being produced for each phase — Phase 1 tests multiple concepts, Phase 2 scales winners and introduces iteration variants, Phase 3 runs the next concept generation.

Who delivers each asset and by what date — creative timelines that are not locked before the quarter starts will slip, and creative slippage is the most common reason a media plan underdelivers in Phase 2.

For brands working with a creator network for UGC, the delivery and testing schedule needs to be mapped against the Phase 2 scale ramp. Creator content that will be used as primary Phase 2 scale creative must be delivered and tested in Phase 1. Sequencing this backward from the Phase 2 launch date is how you ensure the assets are ready rather than hoping they arrive in time.

Budget Pacing: Three Buckets, Not a Flat Allocation

Budget pacing across a 90-day plan is not about spending evenly. It is about spending in alignment with what the data supports and what the calendar demands.

Testing budget (15 to 20 percent of quarterly total). Allocated to creative testing, new audience experiments, and channel expansion. This is an investment in learning, not an acquisition line. It runs throughout all three phases but is weighted toward Phase 1.

Scale budget (60 to 70 percent of quarterly total). Allocated to proven campaigns running proven creative against validated audiences. This is the revenue-generating core. It is heavily weighted toward Phase 2 and the first half of Phase 3.

Reserve budget (10 to 15 percent of quarterly total). Held undeployed for two scenarios: opportunistic scaling when creative or channel performance exceeds expectations, and demand peak amplification during the windows identified in the seasonality layer. This budget is not pre-committed. It is deployed based on in-period performance signals.

The reserve prevents both budget management failure modes: spending testing budget at scale before the signal supports it, and failing to capitalize on high-performance windows because all budget is pre-allocated to campaigns that are not outperforming.

KPI Framework: Define Before the Quarter Starts

A plan without defined measurement criteria is a wishlist. The primary KPIs need to be set before the period begins, with a clear methodology for how they will be measured.

MER target. What Marketing Efficiency Ratio does the business need to achieve for the quarter to be operationally healthy? This is set collaboratively with the finance function, not determined by the media team alone. See how contribution margin analysis sets the floor for what MER needs to be at a given spend level — the MER target is derived from margin requirements, not from what the previous quarter produced.

New customer CPA target. What is the acceptable cost to acquire a net-new customer, separate from blended CPA? This should account for LTV. A brand with a strong repeat purchase rate can justify a higher new customer CPA than a single-transaction brand — the lifetime economics support a larger acquisition investment on the first order.

Creative promotion threshold. What performance does a Phase 1 test creative need to reach to be promoted to Phase 2 scale? Define this in advance as a specific number relative to account average — not as a subjective judgment call made when Phase 2 starts.

Channel MER contribution. If running across Meta, Google, and TikTok simultaneously, what is the expected MER contribution from each channel? This gives the media buyer a clear framework for budget reallocation decisions when one channel outperforms or underperforms mid-quarter.

FAQ

Should the 90-day plan change if performance deviates significantly from targets in Phase 1? Yes — but the plan should change in a structured way. Define decision rules in advance: if Phase 1 MER is X% below target by day 20, reduce Phase 2 scale budget by Y% and extend the testing phase. If Phase 1 produces a creative winner that significantly outperforms expectations, accelerate Phase 2 launch by one week. Predefining these decision rules prevents reactive over-correction while maintaining the plan's responsiveness to actual data.

How do we handle a major promotional event that falls mid-quarter, outside the planned peak windows? Treat it as a reserve budget deployment event. The promotional window requires elevated spend, pre-built creative, and a clear measurement plan for separating promotional performance from baseline performance. If the event is large enough to materially shift the quarter's MER, document it as a separate period in the measurement analysis so it does not contaminate the quarterly baseline.

What does the 90-day plan hand-off look like at the end of the period? The close of each 90-day plan produces four outputs: performance analysis against KPIs with root-cause explanations for any variance; creative library documentation cataloging what tested, what scaled, and what fatigued; budget recommendation for the next period based on what the data showed; and the brief framework for Phase 1 of the next plan. These four outputs are what make each plan the foundation for the next one rather than a stand-alone project.

Closing

The 90-day paid media plan is not a document you produce to satisfy a reporting requirement. It is the operating structure that makes every tactical decision more informed.

When you know where you are in the quarter, what creative is coming, how budget is allocated across phases, and what the measurement benchmarks are, you stop reacting and start executing against a forward-looking framework.

Build the plan before the quarter starts. Review it weekly against the KPIs. Adapt it based on what the data shows. Feed what you learn into the next one.

That is how performance marketing becomes a system.

Keep reading

Pieces I've written on related topics that pair well with this one:

• How to Use Spend Pacing as a Diagnostic Tool, Not Just a Budget Control — Spend pacing is more than budget control. Here's how to use it to diagnose creative fatigue, auction pressure, and delivery problems before they hit R…
• The New Customer Rate Metric: Why It Matters More Than ROAS When Scaling Paid Media — ROAS tells you what happened. New customer rate tells you whether paid media is actually growing your business.
• What a Healthy Paid Media Account Looks Like at $10K, $50K, and $200K Monthly Spend — The account structure that works at $10K/month will break at $200K. Here's what healthy paid media looks like at each eCommerce spend stage.
• What Happens When You Turn Off Paid Ads for 30 Days (We Tested It) — We paused paid ads on a $6M brand for 30 days. Revenue dropped 79%.
• Why Your Retargeting Window Is Probably Too Long (And What to Do About It) — A 30-day retargeting window looks comprehensive and costs real money.