Fire a performance agency when creative output has stalled, when reporting cannot connect spend to revenue, when the same optimisation explanation repeats for three consecutive months, or when nobody on the account has asked about your margins. Poor results alone are not sufficient evidence. Poor process is.

That distinction saves a lot of money in both directions. Businesses fire competent agencies for results that were never achievable on the timeline they applied, then hire a replacement who inherits the same constraint. Other businesses retain agencies through eighteen months of process failure because the monthly report always looked busy.

The useful question is never "are results good." It is "is this engagement structurally capable of producing results, and if not, whose problem is that." If you've worked through how long before you judge a marketing agency, this is the harder follow-up: what to do once the window has passed.

Why are poor results not sufficient evidence?

Because results are downstream of at least four things, only one of which the agency controls.

The offer. If pricing, product, or positioning is the constraint, no channel execution fixes it. A well-run campaign against a weak offer produces expensive clarity about the offer.

Conversion capacity. If the landing page converts at 1.1 percent and the sales team responds in four hours, media performance is capped regardless of who manages it. The diagnostic is in building a qualified demand engine.

Budget relative to the channel's floor. ₹40,000 of monthly media cannot generate statistically meaningful testing data. The agency is being asked to learn from noise.

Your own inputs. Approval turnaround, asset access, sales data, willingness to appear on camera. If creative approvals average seven days, the account runs at half the cadence you are paying for.

Truth line: Before you change agencies, be certain you are not about to buy the same result at a different price.

The six process failures

These are the grounds. Each is observable within 45 days and none requires waiting for an outcome window.

1. Creative output has stalled

The same three to five assets have run for more than six weeks. Frequency is climbing, click-through rate is declining, cost per result is rising.

This is the most common failure and the most consequential, because creative supply is the binding constraint in roughly seven of ten accounts above ₹5 lakh monthly spend. The volume requirements are in creative volume is the new targeting.

Caveat: check your own approval latency first. If assets sit unapproved for a week, the constraint is yours.

2. Reporting cannot connect spend to revenue

The monthly report leads with impressions, reach, clicks, and engagement. It contains no statement of what was tested, what failed, and what changes next.

A report with no failures in it is a report from an engagement that is not testing anything. Description is not diagnosis.

3. The same explanation repeats for three months

"We're still optimising." "The algorithm is learning." "Seasonality." Each is legitimate once. Three consecutive months of the same explanation is a holding pattern, not a plan.

The test: ask what specifically will be different in the next 60 days. A credible answer names a change in approach — a new creative direction, a channel reallocation, a technical fix shipping. A weak answer describes more of the same with more optimism.

4. Nobody has asked about your margins

An agency optimising to lead or order volume without knowing what a customer is worth is guessing at the target. They cannot set a correct acquisition ceiling, cannot prioritise higher-margin products, and cannot tell you when to stop spending.

This is the single clearest marker of vendor rather than partner behaviour — the distinction is in growth partner versus vendor.

5. Conversion tracking was never verified

Check it yourself. Duplicate events, missing purchase values, test conversions still firing, a page view counted as a conversion. Most accounts we inherit above ₹5 lakh have at least one fault.

If tracking is broken, everything reported since is fiction — and an agency that has not audited it in six months has not looked.

6. They have never disagreed with you

Twelve months of unbroken agreement means nobody is exercising judgement on your behalf. An agency that has never recommended against something you asked for, never proposed reducing spend on a service it delivers, and never told you the problem was downstream is executing rather than advising.

Four or more of these at day 45 means the engagement is structurally failing, and waiting for the outcome window costs you a quarter you will not recover.

What are the failures that are actually yours?

Four honest ones, worth checking before you give notice.

Approval latency. Measure it. If your average turnaround on creative is five days, your ceiling is roughly six assets a month regardless of production capacity.

Withheld numbers. Most clients do not share margins, close rate, or average deal value, then find the recommendations generic. An agency cannot optimise toward economics it has not been given.

Budget below the channel floor. If the retainer cannot fund strategy, creative, and measurement simultaneously, one of the three is being cut — usually strategy. The arithmetic is in why cheap marketing retainers cost more.

Scope that never scaled. An agency contracted at ₹3 lakh monthly spend running a ₹12 lakh account has the same hours it always had. That is a contracting failure presenting as a performance one.

Punishing honesty. If the first piece of difficult feedback was received badly, you trained a vendor within a month. This one is uncomfortable and it is common.

The remediation conversation

Most engagements that end badly could have been fixed at day 45 if anyone had a structured moment to raise it. Before giving notice, have this conversation once.

Name the specific failures, not a general dissatisfaction. "Creative output has been four assets a month for three months" is actionable. "We're not happy with results" is not.

Ask three questions:

• What specifically will be different in the next 60 days?
• What have you learned about our buyer that you did not know at the start?
• What do you need from us that you are not getting?

Agree a 45-day remediation plan with named deliverables and a follow-up date.

Then hold the date. A remediation conversation without a review date is a complaint.

An agency that answers the three questions well has usually earned the 45 days. One that deflects on the second — because there is no accumulated knowledge to describe — has not.

What to do before you give notice

The sequence matters enormously, and businesses that reverse it lose things permanently.

1. Secure administrative access first. Google Ads, Meta Business Manager, Google Analytics, Search Console, Google Business Profile, domain registrar, hosting, CRM. Request owner-level access framed as internal governance, not as a prelude to leaving. A request made calmly is granted far more often than one made in a dispute.

2. Export your data. Campaign history, creative source files, keyword research, audience lists, reports, lead databases.

3. Document current campaign structures. Screenshots of account structure, active audiences, bidding strategies, and what has been tested. This is the institutional knowledge that otherwise walks out.

4. Check your notice period and handover obligations. If the contract has no handover clause — the most common gap in Indian marketing agreements — handover becomes a goodwill exercise performed by a team that just lost your account. The clauses are covered in eleven contract clauses that cost Indian businesses lakhs.

5. Only then give notice.

Ad accounts accumulate conversion history and algorithmic learning that materially affect performance. Losing account ownership typically costs two to four months of rebuilt learning at elevated acquisition cost — a transition cost that belongs in the decision.

How do you avoid repeating this?

Four things to fix in the next engagement, at contract stage rather than in month five.

• Agree the judgement windows in writing — which channel, which signal date, which outcome date, judged on which metric.
• Quantify creative production with monthly concept counts that scale with media spend, not "creative included."
• Name the people and their weekly hours, with consent required before senior substitutions.
• Specify your own obligations — approval turnaround, asset provision, access to sales data — so the clock runs on both sides.

The mechanics are in how to write an SOW that holds and the evaluation questions in how to read a marketing proposal like a CFO.

How Midgrow handles this

We build complete growth systems rather than selling channel management as a line item, and the structural things that cause these failures are the ones we try to remove at contract stage.

• Conversion tracking is audited in week one, because a diagnosis built on faulty data produces the wrong fix.
• Creative production is a quantified scope line that scales with spend, since flat production against rising media manufactures failure 1 on a schedule.
• We ask for margins, close rate, and deal value before proposing anything. If we cannot have them, we say so rather than optimising to volume.
• Reporting names what failed and what changes as a result, monthly.
• Symmetrical thirty-day notice and a defined fifteen-day handover, because an agency that needs a lock-in to retain clients is telling you something.

That spans performance marketing, social media, SEO, and the AEO and GEO layer determining whether AI assistants recommend you at all. Full scope on our digital marketing services page.

The proof is public rather than promised. We generated 10,890 leads at 11.3x ROI for a solar EPC client and delivered 585 percent organic growth with first-position rankings for Autosys Solar — multi-year engagements in energy that lasted because nothing was holding anyone in place.

Book a 45-minute growth diagnostic. Bring your last three reports. We'll tell you whether the problem is the agency, the brief, or the business — including when the honest answer is that you should stay. Start the conversation.

Frequently asked questions

When should I fire my marketing agency?
When process quality is poor rather than when early results are. Six observable failures constitute grounds: stalled creative output, reporting that cannot connect spend to revenue, the same explanation repeating for three months, nobody asking about your margins, unverified conversion tracking, and an agency that has never disagreed with you. Four or more at day 45 means the engagement is structurally failing.

Are poor results enough reason to change agencies?
Usually not on their own. Results depend on your offer, your conversion capacity, your budget level, and your own approval speed and data sharing — only one of which the agency controls. Diagnose which of those is the constraint before switching, or you will buy the same outcome from someone new.

How do I tell if the problem is my agency or my business?
Check four things on your side first: average approval turnaround on creative, whether you have shared margins and close rate, whether your budget can fund strategy, creative and measurement simultaneously, and whether your scope was ever updated as spend grew. If all four are healthy and process failures persist, the problem is the agency.

What should I do before giving notice to an agency?
Secure owner-level access to all ad accounts, analytics, domain and hosting first, then export campaign history, creative source files and lead data, then document current account structures. Only after that give notice. Businesses that give notice first frequently lose account history permanently, along with the algorithmic learning that took months to build.

Does switching agencies cost me performance?
Usually two to four months of transition, mostly from rebuilding algorithmic learning if account ownership changes. If you own the accounts and simply grant new access, the loss is much smaller. This transition cost belongs in the decision — sometimes it justifies a 45-day remediation attempt first.

What is a remediation conversation and how do I run one?
Name the specific process failures rather than general dissatisfaction, ask what will be different in the next 60 days, what they have learned about your buyer, and what they need from you. Agree a 45-day plan with named deliverables and a follow-up date, then hold the date. Without a review date it is a complaint, not a plan.

How do I check whether my conversion tracking is actually working?
Compare platform-reported conversions against your CRM or order records for the same period. Look for duplicate events, conversions without purchase values, test events still firing, and page views counted as conversions. A material mismatch means everything reported since is unreliable, which is itself grounds for concern.

How do I avoid the same problem with the next agency?
Fix it at contract stage. Agree judgement windows per channel with named metrics, quantify creative production with monthly concept counts that scale with spend, name the people working on the account with weekly hours, and specify your own obligations on approvals and data access so the clock runs on both sides.