My Agency Isn't Delivering — How Do I Diagnose It and What Do I Do?

A meaningful share of the accounts we review that are reported as underperforming are actually performing and being measured incorrectly. A purchase event that never fires, one that fires without a value, one that fires twice, or a currency configured as dollars on a store selling in dinars — each of these produces numbers that look like failure or false success.

This is a twenty-minute check and it comes first because it is the only explanation that costs nothing to fix and changes everything if true. Place a real order through your own store and confirm the purchase event appears in each platform with the correct value in Kuwaiti dinars.

Then compare reported conversions against the actual orders in your back office for a full month. If the platforms are reporting substantially fewer conversions than you have orders, the campaigns may be working better than anyone thinks and the optimisation is being starved of data.

If reported conversions exceed your actual orders, you have duplicate events, and your performance has been overstated the entire time. This is the more dangerous case because budget decisions have been made on inflated numbers, and the correction will look like a sudden collapse in performance when it is actually the first accurate reading.

Product photography. If your images are inconsistent, poorly lit or missing entirely for half the catalogue, no campaign management fixes that. A great many Kuwait accounts we audit are performing exactly as well as their creative assets allow, and the agency has been asked to compensate for something it cannot compensate for.

Positioning and offer. If nobody can state in a sentence why someone should buy from you rather than the alternative, the agency has been asked to sell something that has not been defined. That is not an execution problem and changing executor will not solve it.

Response time and fulfilment. If campaigns generate enquiries and those enquiries are answered the next morning, or orders arrive late, the marketing worked and the business did not convert it. Look at your own enquiry response times before concluding the leads were poor quality.

And decision speed. An agency waiting three weeks for creative approval or a budget decision is an agency operating at a fraction of its capability, and that delay is yours rather than theirs. Count how long approvals have taken over the last quarter before judging output volume.

Was a number ever agreed? A great many of these situations are not underperformance but a mismatch between an unstated expectation and an unstated commitment. If you expected fifty leads a month and nobody ever wrote fifty down, there is nothing to be underperforming against.

Was the timescale realistic for the channel? Judging SEO at month two, or a paid campaign at week one, produces a verdict of failure regardless of the quality of the work. Different channels pay back on completely different schedules and judging a slow one on a fast one's timeline is a category error.

Was the budget sufficient for the goal? A campaign spending 10 KD a day accumulates conversion data slowly, never fully exits the learning phase, and underperforms for structural reasons rather than through poor management. Sometimes the honest answer is that the goal and the budget were never compatible.

And is the margin the actual constraint? If your gross margin means you need a return of four just to break even, and the category typically delivers two and a half, no agency can close that gap. That is a pricing or product problem being experienced as a marketing one, and it is a genuinely common finding.

Work is not being produced at the agreed volume, and there is no input from you that explains it. Count the deliverables against the contract for the last three months. If the agreement said twelve posts and four have appeared, that is a factual gap rather than a judgement call.

Nothing has been tested or changed. An account that has run the same campaigns, the same audiences and the same creative for four months is not being managed, it is being maintained. Ask what has been tested and what the results were, and a lack of answer is itself an answer.

The reporting avoids commercial numbers, and asking for them produces friction rather than a report. An agency that is performing wants you looking at orders and cost per acquisition, because those numbers make its case.

And nobody senior has looked at your account in months. Ask directly when the account was last reviewed by someone experienced and what changed as a result. In agencies with churn, accounts drift onto junior autopilot, and the drift is invisible until someone asks.

Have one direct conversation with specifics rather than accumulating unspoken dissatisfaction. Bring the numbers, the contract, the deliverable count and the tracking check. A conversation grounded in specifics gets a different response from one that opens with we are not happy, and a good agency will engage with it.

Agree a thirty-day reset with a written scope and two or three measurable outcomes. This gives an underperforming but capable agency a fair chance to correct, and it gives you a clear basis for the decision if they do not. Both parties benefit from the clarity.

Fix the inputs on your side in parallel regardless of the outcome. Product photography, positioning, response times and approval speed follow you to whoever comes next, and a business that switches agencies without fixing these repeats the experience with new people and concludes that all agencies are the same.

And if you do change, secure your accounts before you announce anything. Confirm ownership, retrieve access, export your data and get the handover documentation while the relationship is still functional. The window where this is easy closes the moment you give notice. You can start a free Shopify trial and keep the store, order history and customer list in an account of your own so a supplier change never costs you your data.