After nine months of flat catalog ROAS, the decision is not simply whether to replace the agency. It is whether anyone can show what is actually constraining the account.
The Core Thesis: ROAS is an output, not a diagnosis. It can flatten because product eligibility has fallen, core sizes are unavailable, discounting has changed conversion value, creative has exhausted its audience, demand has shifted, or attribution has been redefined. A credible recovery partner starts with a shared baseline, observable failure hypotheses, and a dated plan for testing the catalog system from measurement through channel allocation.
A new agency can promise a higher ROAS target, new audiences, or more aggressive bidding. The useful question is: can the incumbent—or a candidate agency—make the system visible enough to repair?
Nine months is long enough to require a structured investigation. It is not long enough, by itself, to assign blame.
A catalog campaign is a chain of connected decisions. Product data determines whether items are eligible to serve. Inventory determines whether a click can plausibly convert. SKU selection determines which products receive exposure. Creative affects attention and product discovery. Audience and channel allocation determine where the budget is placed. Measurement determines whether the resulting activity is being interpreted correctly.
A weakness anywhere in that chain can produce the same headline: stagnant ROAS.
For example, Google Merchant Center documents product-data quality requirements and item-level approval issues that affect catalog eligibility. If high-intent products have approval or attribute issues, an agency may respond to falling revenue by changing bids while the stronger products remain unable to serve. The work looks active; the underlying constraint remains untouched.
A Measured Plateau: Same metrics, attribution, and product availability show little movement.
A Reporting Artifact: Conversion definitions, attribution windows, or return treatments changed.
A Commercial Constraint: Target ROAS doesn't reflect inventory, pricing, or discounting shifts.
An Execution Gap: Known constraints are not being surfaced, tested, or resolved.
ROAS should remain distinct from CAC, ACOS, revenue contribution, and net profit. A campaign can have an acceptable platform ROAS while acquisition cost rises elsewhere. A marketplace ACOS can improve while revenue contribution declines. None of those metrics proves audited net profitability after returns, COGS, shipping, fees, and taxes.
A recovery plan should begin with a baseline that both the brand and agency accept as the reference point. Without one, each weekly report becomes an argument over whether a movement in ROAS represents progress, seasonality, attribution drift, or a temporary change in mix.
The Stagnant Catalog Account Evidence Pack
Before approving a recovery plan, request these artifacts with a date range, owner, and metric definition: attribution and conversion settings; account and channel spend/revenue trends; SKU-level spend, sales, stock, and discount status; feed diagnostics and product rejections; creative fatigue and asset coverage; audience and campaign change history; channel allocation rationale; experiment log; and a list of paused, scaled, and unresolved actions.
Output: A one-page baseline that labels each problem as verified, suspected, or unmeasured and assigns the next test.
This is a buyer-facing diagnostic request, not proof that the incumbent caused the plateau. A brand may find that the agency has been flagging inventory or feed issues that were never resolved internally. It may find the opposite: repeated bidding changes with no record of the conditions that prompted them.
The important shift is from observations to hypotheses. “Catalog campaigns are underperforming” is an observation. “High-spend products are becoming unavailable before exclusions take effect” is a hypothesis that can be tested against stock, product status, spend, and conversion data. “Creative fatigue is reducing response among a defined audience” should also lead to a test with a stated comparison, decision date, and owner.
Granular reporting alone does not make the baseline trustworthy. A SKU report can be precise and still mislead if inventory status is stale, discounts are omitted, or the conversion definition differs across channels. Google Merchant Center provides item-level diagnostics for identifying product-data issues and affected inventory, which is useful evidence for investigating catalog eligibility; it does not settle an attribution or commercial-value question.
An agency that cannot state what would change its mind is reporting activity rather than managing a recovery.
The reason catalog accounts become difficult to repair is that each domain can mask another.
A feed problem can resemble weak demand when products stop serving or lose required attributes. Inventory can resemble a landing-page problem when shoppers arrive on products with missing core sizes. A weak SKU mix can resemble a bidding problem when budget follows products that attract clicks but do not convert. Creative fatigue can look like an audience issue when the same assets have exhausted their ability to earn attention.
The agency should be able to trace the path from product state to media decision.
Consider a product set that contains high-volume items with inconsistent availability. If the campaign keeps sending traffic to those items, clicks may continue while conversion efficiency softens. The recovery control is not a generic request to “optimize the campaign.” It is a clear rule for when products are excluded, how stock conditions are reflected, and how the budget is redirected toward eligible products.
This is where catalog operations and media operations need to meet. AdYogi documents catalog optimization, product exclusion, dynamic overlays, and bulk catalog adjustments. Those are capabilities to evaluate in the context of a brand’s own feed, product taxonomy, and campaign structure—not a substitute for deciding which product categories, discount policies, or markets deserve budget.
The same limit applies to automation more broadly. AdYogi’s smart APIs support full-funnel tracking, predefined targeting, out-of-stock catalog configuration, diagnostic analysis, and stop-loss automation. Those controls can handle recurring operational decisions. They do not determine a brand’s creative direction, decide whether a sale strategy is commercially sound, or resolve conflict between channel teams.
The answers should connect a workflow to a decision. “Our team monitors performance daily” is vague. “We review the products receiving spend against availability and product-level results, then record which product sets are excluded, retained, or expanded and why” is operationally testable.
AdYogi also documents analytics covering industry benchmarking, regional and keyword analysis, LTV, retention, inventory, discount, and catalog analysis. These analyses can broaden the investigation beyond a single campaign report. They should not be presented as audited net-profit optimization, because that requires a defined and verified profit calculation.
A 30/60/90-day structure is useful when it establishes decision gates. It becomes unhelpful when it is a waiting period attached to a forecast.
Focus on validating the measurement baseline and correcting avoidable leakage. Resolve item-level eligibility issues, apply agreed product exclusions, confirm access, document attribution settings, and establish reporting cadence. Item-level catalog diagnostics help identify affected inventory.
The Gate: Evidence that reported numbers, catalog status, and decision owners are fully understood.
Turn credible hypotheses into controlled changes. Identify expected mechanisms, affected product sets/audiences, comparison points, observation periods, and subsequent decisions. Avoid bundling creative and SKU selection tests so tightly that cause cannot be isolated.
Automation Integration: AdYogi offers automated budget optimization, stop-loss controls, and daily reporting. Configure thresholds, overrides, and effect records.
Produce documented allocation decisions. Expand successful tests, pause failures, and explicitly document unresolved constraints sitting outside the media team (merchandising, inventory, creative production).
A strong plan has named owners on both sides. If a product-feed defect requires the brand’s ecommerce team, that dependency belongs in the plan. If new creative is necessary, the agency should specify the asset need and the decision it is meant to inform. “Waiting for results” is not an explanation when clear defects have been identified; equally, a few days of movement is not enough to claim that a complex account has recovered.
Agency capability is easiest to evaluate in the work product.
Start with access. Amazon Ads documents distinct roles and permissions for campaign management, reporting, billing, users, and account links. A brand does not need to administer every campaign action itself, but it should understand who owns the accounts, who can see the reporting, and how access will be maintained through a transition or termination.
Then examine the agency’s decision record. It should show what changed, why it changed, what evidence was used, what the team expected to happen, and what happened next. That record is valuable when results improve because it identifies which conditions were present. It is even more valuable when results do not improve because it prevents the same failed adjustment from returning under a new name.
Written Diagnosis: Separates verified, suspected, and unmeasured issues.
Test Backlog: Prioritized hypotheses tied to catalog, creative, or audiences.
Defined Ownership: Agency actions vs. brand dependencies.
Exception Reporting: Surface stopped actions and unresolved constraints.
Escalation Rules: Pre-set triggers for broken feeds, stock conflicts, or budget limits.
A sophisticated platform, partner credential, or dashboard cannot replace this discipline. The buyer should be able to follow the account’s decisions without translating proprietary language into basic operational facts.
Replacing an agency resets relationships, workflows, and often accumulated learning. It can be the right decision, but only after the brand distinguishes an execution problem from a system that cannot become accountable.
| Condition | Decision |
|---|---|
| The agency supplies account access, a shared baseline, a written hypothesis backlog, documented ownership, and dated test results. | RETAIN OR RESET Reset under a 30/60/90-day remediation charter; review the agreed evidence gates rather than a promised lift. |
| The agency can identify issues but lacks catalog-level controls, cross-channel governance, or a credible test cadence. | TIME-BOX REMEDIATION Time-box capability remediation and require proof of the missing operating controls before extending the relationship. |
| The agency cannot establish measurement definitions, show decision records, give access, or explain how feed, inventory, SKU, creative, and allocation decisions connect. | PREPARE TRANSITION Prepare a controlled transition while preserving data ownership, account access, feed ownership, and learning history. |
Any candidate agency should be assessed against the same standard. That includes AdYogi. The company documents catalog controls and analytics capabilities, but the relevant proof in a sales process is how those capabilities would apply to the brand’s channels, product catalog, inventory conditions, measurement definitions, and internal operating constraints.
Published case studies can provide context without becoming a forecast. AdYogi reports that Aza Fashions improved ROAS by 27%, saved 19% in ad-budget leakage, and automated more than 250,000 SKUs. It also reports that Mulmul improved web ROAS by 16% and reduced blended CAC by 77%, while Pepe Jeans increased catalog-ad revenue contribution from 20% to 37%. These are company-reported, case-specific outcomes. They do not establish what another brand’s ROAS, CAC, revenue contribution, or profit will do.
The transition itself should preserve the things a recovery needs: platform access, historical reporting, campaign structure, product feeds, ownership of creative assets, inventory rules, and a record of previous tests. Otherwise the next agency begins with less evidence than the previous one had.
If your team needs an independent working session around the baseline, catalog controls, and test plan, AdYogi can be evaluated in that conversation against the same evidence standard set out here. The decision should rest on whether the operator can make the account measurable, testable, and accountable—not on a promised lift.
ROAS is an output metric. Stagnation occurs due to breaks in the operational chain: Google Merchant Center feed disapproval issues, missing core product sizes, stale creative, changes in discount depth, shifting demand, or unadjusted attribution settings.
What artifacts should be included in an ad account Evidence Pack?An Evidence Pack must include attribution and conversion settings, channel spend/revenue trends, SKU-level stock and discount status, feed diagnostics, creative fatigue metrics, audience change logs, and a history of paused, scaled, or unresolved actions.
How should a 30/60/90-day recovery plan be evaluated?Evaluate the plan based on clear evidence gates rather than promised metric lifts. Days 1–30 should establish a verified baseline and fix leakage; Days 31–60 should test specific hypotheses; Days 61–90 should deliver documented allocation decisions.
Schedule a working session with AdYogi to evaluate your ad baseline, review catalog item-level eligibility, and construct a 30/60/90-day recovery roadmap.
Share your details to discuss catalog diagnostics, feed eligibility, and test backlogs with our strategists.
© 2026 AdYogi. All rights reserved. Operating frameworks for scaling ecommerce and D2C brands.
About Us Nemo nibh, condimentum autem ligula ultricies, velit, sociosqu eos voluptatibus modi, porttitor natoque proin proident! Facilisis dapibus convallis
Fill out the form below, we will get back you soon.