Why Your Facebook and Instagram Ads Are Not Working: Find the Leak

Your campaign has spent its budget and the report shows reach, clicks and messages. The admissions team still has empty places, or the event team has unpaid registrations. To decide what to change, trace the journey from paid attention to the outcome you actually need.

Do not diagnose failure from the button used to create the advert. Inspect the selected objective, audience, offer, destination and follow-up. A low price per click can coexist with an expensive completed outcome.

1. Confirm what the campaign is trying to achieve

Match the campaign's objective and optimisation settings to the action you want. Meta's campaign-objective guidance starts with the business goal. Reach, engagement, website visits, enquiries and purchases are different results.

If you use a boosted post, inspect its selected goal rather than assuming what it optimises for. Use Ads Manager when the campaign requires settings, testing or reporting beyond the simpler promotion workflow. No setup guarantees that the people who respond will become customers.

2. Check delivery before judging persuasion

Confirm that the campaign is approved, running during the intended period and spending as expected. Review the location, audience restrictions, placement results and destination link. If delivery is limited, investigate that before rewriting the entire offer.

Keep the service area and eligibility realistic. An employer-funded training programme may need a very specific buyer; an event open to a broad local audience may need a different approach. Broad targeting is an option to test, not a rule that overrides your actual market.

3. Make the offer answer the decision

A prospective learner needs course scope, dates, timetable, eligibility and credible recognition. A delegate needs the event's relevance, agenda, fee and booking terms. A vague promise of “excellence” does not resolve those questions.

Use verified information and one clear next step. Make the landing page consistent with the advert. If the price or conditions change between them, the campaign is creating avoidable confusion.

4. Test the destination as a real user

Open the actual advertised link on a phone. Submit a test enquiry with clearly marked test data, confirm that it reaches the right place and check the acknowledgement. Test registration and payment status handling where relevant.

Measure the complete action rather than counting a button tap as a successful submission. Our website enquiry guide explains the checks in more detail.

5. Define a qualified enquiry

Agree what makes an enquiry useful for this campaign. For training, it might mean someone interested in the advertised intake who meets eligibility and can attend the schedule. Keep the criteria relevant and collect no more information than needed.

Log unique enquiries, qualified enquiries and the reasons cases do not progress. Ineligible applicants, unavailable dates and unanswered messages require different fixes. Do not dismiss every lost enquiry as “poor lead quality.”

6. Connect the campaign to staffed follow-up

Assign an owner, a backup and a service window before launch. Give staff current information and an escalation route. Record a next action for cases awaiting a quote, invoice or decision.

A quick automatic greeting is not the same as a useful answer. Review unresolved conversations as well as response time. Use the WhatsApp enquiry routine to make the handoff dependable.

7. Calculate costs at each stage

Illustrative example: KES 7,500 of media spend produces 50 unique enquiries, 20 qualified enquiries and five paid registrations. Media cost per enquiry is KES 150; per qualified enquiry, KES 375; per paid registration, KES 1,500.

If each registration contributes KES 2,000 after direct delivery costs, the five contribute KES 10,000 before campaign costs. Subtracting media spend leaves KES 2,500 before management fees, content, staff time and overheads. These are hypothetical figures, not typical Kenyan ad prices or a client result.

Use a consistent attribution window and reconcile platform-reported results with your own records. A person may respond through several channels; do not count the same paid registration twice. Report uncertainty where attribution is incomplete.

8. Decide what the next test will establish

Change one major variable where feasible: the offer explanation, creative, destination or follow-up process. Set a spending cap, review date and success criteria in advance. Small numbers can vary sharply, so avoid declaring a winner from one or two conversions.

Check what information tracking tools collect before installing them. Do not send sensitive health or learner data into advertising systems. A plan for analytics should include privacy, access and the events you actually need.

Before spending again: identify the largest evidenced loss between delivery, enquiry, qualification and completed outcome. Fix that stage and assess the next cohort over its decision cycle.

Zero Impulse connects paid campaigns, measurement and follow-through. Request a diagnosis with your campaign data and outcome records.