Home / Meta Ads Interview Questions / Scenario and Case Study Questions
Interview questions

Facebook Ads Scenario and Case Study Questions Interview Questions

Facebook ads scenario interview questions: 110 real-world scenarios and data cases with model answers and expert follow-ups, from Basic to Advanced.

  • Model answers
  • Basic to Advanced
  • Search and filter

Real-world scenarios, data cases and expert follow-up questions. Questions are grouped from Basic to Advanced, each with a model answer. Use the filters or search to find what you need, then take the free Meta Ads test to check your score. For official definitions and policies, see the Meta Business Help Center.

What these Facebook ads scenario interview questions cover

These Facebook ads scenario interview questions focus on real-world scenarios and data cases. They cover budget decisions, account audits, performance drops, client conversations and expert follow-up questions. Each question comes with a model answer written by a working Meta ads specialist, so you can compare your own answer and see the points an interviewer or client expects to hear.

Interviewers use questions like these to check how you apply knowledge under pressure, which is what senior roles test most. Start with the Basic questions, then move to Intermediate and Advanced. Click any question to open its answer, use the level filters to focus, or search for a keyword. When you are ready, take the free Meta Ads test to score yourself and find your weak topics.

All questions, from Basic to Advanced

Advanced questions

1AdvancedAdvanced Scenarios (Real-World Style)

You join a company and the Head of Marketing says "Our Meta ads stopped working last month". What are your first 5 actions in the first 48 hours?

  • Get access and review last 90 days trend by campaign: spend, CPA/ROAS, CPM, CTR, CVR, frequency.
  • Audit tracking health (Events Manager, CAPI, dedupe, EMQ, recent website changes).
  • Review change history (Ads Manager "History" tab) – edits, new structures, budget jumps.
  • Check creative age and fatigue, account structure fragmentation, learning status.
  • Talk to sales/ops: offer changes, stock, pricing, lead handling; then present a root-cause hypothesis and action plan.
2AdvancedAdvanced Scenarios (Real-World Style)

Scenario: "If you had AED 10,000 and one week to prove yourself, how would you spend it?"

  • Clarify the goal and KPI; check tracking first (no budget wasted on broken data).
  • Consolidated campaign with proven offers, broad targeting, 6–10 diverse creatives, optimization for the deepest event with volume.
  • Small retargeting budget for warm audiences to capture quick wins.
  • Daily monitoring without over-editing; report learnings + next steps with data.
3AdvancedAdvanced Scenarios (Real-World Style)

Two ad sets in the same campaign: one has CPA AED 40 with 20 conversions, the other CPA AED 60 with 80 conversions. Which is better?

  • Depends: the second delivers 4× volume; marginal CPA and scale matter.
  • If target CPA is 70, both are profitable – the second contributes more profit overall.
  • Don't judge by CPA alone; consider volume, consistency, quality and statistical confidence (20 conversions is small).
4AdvancedAdvanced Scenarios (Real-World Style)

Your manager asks you to cut CPA by 30% next month. How would you approach it?

  • Break down CPA levers: CPM (auction/creative quality), CTR (creative), CVR (landing page/offer), lead quality.
  • Quick wins: pause clear losers, consolidate fragmented ad sets, fix tracking, improve landing page speed.
  • Medium term: new creative concepts, offer improvements, CRM feedback loop.
  • Set realistic expectations and track weekly.
5AdvancedAdvanced Scenarios (Real-World Style)

How would you handle an ad set with excellent ROAS but only AED 50/day spend; when you increase budget, ROAS drops?

  • Small audiences/narrow targeting can't absorb more budget – diminishing returns.
  • Keep it at the efficient level, find horizontal growth (new audiences, creatives, markets) instead.
  • Try cost cap/ROAS goal with higher budget to see if efficient volume exists.
6AdvancedAdvanced Scenarios (Real-World Style)

What would you do if the CEO wants to run ads only to "rich people" with narrow interests even though broad performs better?

  • Show data from an A/B test (narrow vs broad) on the same KPI.
  • Explain that creative and offer can attract affluent buyers while broad gives the algorithm room.
  • Compromise: use interests as suggestions in Advantage+ audience, measure lead quality.
7AdvancedAdvanced Scenarios (Real-World Style)

You discover that 40% of conversions came through Audience Network on a lead campaign and quality is poor. Next steps?

  • Validate with CRM: lead quality by placement (use form IDs/hidden fields or URL params with {{placement}}).
  • Exclude Audience Network for this campaign or switch to Higher-intent form/conversion leads optimization.
  • Monitor whether CPL rises and quality improves.
8AdvancedAdvanced Scenarios (Real-World Style)

You have a winning creative but Meta delivers it 95% to men while your product is for both genders. Should you split by gender?

  • Meta delivers to whoever responds better/cheaper – likely the creative appeals to men more.
  • Rather than splitting audiences, create creatives that appeal to women (different angles/faces/use cases) in the same ad set.
  • Split only if business needs guaranteed spend per gender.
9AdvancedAdvanced Scenarios (Real-World Style)

What would you do if your client's competitor started a huge discount campaign and your CPA doubled?

  • Don't fight purely on price; highlight value, guarantees, service, bundles.
  • Adjust budgets to protect efficiency, focus on retargeting/warm audiences and loyal customers.
  • Monitor Ad Library for competitor activity; prepare counter-offers if margins allow.
10AdvancedAdvanced Scenarios (Real-World Style)

The client wants daily reports and changes daily. How do you manage the client while protecting performance?

  • Educate about learning phase and noise; set a fixed optimization cadence (e.g. twice weekly).
  • Provide daily dashboard snapshot but weekly decision meetings.
  • Use rolling 3- and 7-day metrics to avoid reactive changes.
11AdvancedAdvanced Scenarios (Real-World Style)

You are managing 10 clients. How do you prioritize your daily work?

  • Check alerts first: delivery issues, tracking breaks, rejected ads, spend anomalies.
  • Prioritize by spend size and urgency; use automated rules and dashboards.
  • Weekly deep-dive per client; creative pipelines planned ahead.
12AdvancedAdvanced Scenarios (Real-World Style)

What would you do if Meta's reported leads are 200 but the CRM has 150?

  • Check integration failures, duplicates removed by CRM, test leads, spam filtering.
  • Download leads from Meta Lead Center and reconcile by lead ID/time.
  • Fix integration and set up alerts.
13AdvancedAdvanced Scenarios (Real-World Style)

A campaign in Advantage+ is spending on the 65+ age group which isn't your target. What do you do?

  • Check if those conversions are real and valuable.
  • If not, add a minimum/maximum age control (Advantage+ audience allows minimum age; age maximum as suggestion – for strict control use original audience).
  • Update creatives to speak to the target age.
14AdvancedAdvanced Scenarios (Real-World Style)

Your best ad was accidentally deleted. How do you recover performance?

  • Deleted ads can't be restored, but you can recreate the ad using the same post ID (if the post still exists – via "Use existing post" with the post ID).
  • Duplicate structure; expect some relearning.
15AdvancedAdvanced Scenarios (Real-World Style)

How would you test whether Instant Forms or Website landing page brings better leads?

  • A/B test (or separate ad sets) with same creative/audience/budget: one with instant form, one with website.
  • Compare cost per qualified lead and conversion to sale over 2–4 weeks, not just CPL.
16AdvancedAdvanced Scenarios (Real-World Style)

How would you explain the learning phase to a non-technical business owner?

  • "Like a new employee learning which customers buy; the first week costs are higher while the system learns; if we keep changing things, it starts learning again."
17AdvancedAdvanced Scenarios (Real-World Style)

You notice that Meta reports 30 purchases from a campaign, but 25 were view-through. What do you conclude?

  • Most credit comes from views – the campaign may not be driving clicks/purchases directly.
  • Check click-only results and incrementality; view-through can be valid for video/awareness but can over-credit retargeting.
18AdvancedAdvanced Scenarios (Real-World Style)

What would you do if your lead campaign works only on weekends?

  • Analyze day-of-week data over 4+ weeks; if consistent, use budget scheduling to increase weekend budgets.
  • Check sales team availability and response times on weekdays.
19AdvancedAdvanced Scenarios (Real-World Style)

Your team launched 50 new ads at once in the main campaign and performance tanked. Explain why and what to do.

  • Massive changes reset learning, spread delivery across untested ads, and similar ads compete.
  • Roll back: pause most new ads, add new creatives in smaller batches (5–10 diverse concepts), or test in a separate campaign first.
20AdvancedAdvanced Scenarios (Real-World Style)

How would you use the "History" (change log) tab in Ads Manager?

  • Shows all edits with time and user – useful to link performance drops to changes and to audit team actions.
21AdvancedAdvanced Scenarios (Real-World Style)

What is your approach if your Facebook Page has low engagement but ads perform well?

  • Ads can work regardless, but Page credibility affects trust (people check the Page before buying).
  • Post regularly, showcase reviews, respond to comments; use ad engagement to build social proof.
22AdvancedAdvanced Scenarios (Real-World Style)

Explain how you'd decide whether to use Instagram-only or both platforms.

  • Default: both with Advantage+ placements; the algorithm finds the best mix.
  • Only restrict if creative/brand requires it or data shows one platform consistently converts poorly (with enough spend).
23AdvancedAdvanced Scenarios (Real-World Style)

You're asked: "What's the difference between a media buyer and a performance marketer?"

  • Media buyer focuses on platform execution: campaigns, budgets, bids, targeting.
  • Performance marketer owns the business outcome: funnel, offer, creative strategy, tracking, CRO, multi-channel attribution, profitability.
24AdvancedAdvanced Scenarios (Real-World Style)

You have AED 50k for the next month and 3 products with different margins. How do you allocate?

  • Calculate break-even CPA/ROAS per product; allocate more budget to products with best margin × conversion potential.
  • Test each with enough budget, then shift to winners; consider bundles to increase AOV.
25AdvancedAdvanced Scenarios (Real-World Style)

The client's Page was renamed and ads are disapproved citing "Page mismatch". What do you check?

  • Page name/branding mismatch with ad content or landing page domain; Page transparency changes; new Pages need history.
  • Ensure consistent branding and request review.
26AdvancedAdvanced Scenarios (Real-World Style)

How would you scale a campaign that has a 7-day frequency of 3.2, CPA on target and CTR stable?

  • Frequency is high but performance is stable – scale carefully: add new creatives (to lower creative-level frequency), expand audience/markets with new ad sets, and modest budget increases on the existing ad set.
  • Monitor CTR/CPA daily; if they decline, shift to horizontal scaling.
27AdvancedAdvanced Scenarios (Real-World Style)

What is your opinion: should you use Advantage+ everything?

  • Advantage+ features usually perform well with good data and creative volume, but not always.
  • Test against manual setups; keep control where needed (brand safety, special categories, specific markets, exclusions).
  • Use A/B tests to decide based on your account's data.
28AdvancedAdvanced Scenarios (Real-World Style)

What would you change first in an account with poor results: audience, creative, or bidding?

  • Usually creative and offer have the biggest impact; then tracking/optimization event; then structure; bidding last.
  • But first confirm tracking accuracy.
29AdvancedAdvanced Scenarios (Real-World Style)

How would you prove your value in a 30-day probation period?

  • Week 1: audit + quick fixes; Week 2: new creative tests + structure; Week 3–4: scale winners and show measurable improvement.
  • Document everything with before/after metrics.
30AdvancedAdvanced Scenarios (Real-World Style)

A campaign's CPM is AED 120 in UAE. Is it too high? What do you check?

  • High compared to typical ranges; check targeting narrowness, placements, special ad category, seasonality, ad quality rankings, competitive niche (real estate/finance).
  • Focus on whether CPA is acceptable; improve creative engagement and broaden audience.
31AdvancedAdvanced Scenarios (Real-World Style)

Why might Meta show "Your ad set is not delivering due to audience size too small"?

  • Narrow targeting, excluded audiences overlapping, special ad category restrictions, small custom audience.
  • Broaden or merge audiences.
32AdvancedAdvanced Scenarios (Real-World Style)

What would you do if the client asks for more leads but sales team can't handle current volume?

  • Explain that more leads without follow-up waste budget; improve lead quality and routing instead.
  • Implement automation (WhatsApp bot, CRM) and qualification filters before scaling.
33AdvancedAdvanced Scenarios (Real-World Style)

If you could only look at 3 metrics for an e-commerce account, which would you choose?

  • Contribution margin/profit (or MER), CAC/new customer CPA, and ROAS by campaign – plus CVR as diagnostic.
34AdvancedAdvanced Scenarios (Real-World Style)

How do you detect click fraud or bot traffic from Meta ads?

  • Large gap between clicks and LPVs, extremely low time on site, unusual geo/device patterns, form submissions with junk data.
  • Placement breakdown (Audience Network), exclude suspicious placements, use CAPTCHA/validation on forms.
35AdvancedAdvanced Scenarios (Real-World Style)

Explain "testing velocity" and why top agencies obsess over it.

  • The number of meaningful tests (creatives, offers, landing pages) run per period.
  • More tests = more chances to find winners and adapt to fatigue/algorithm changes.
36AdvancedAdvanced Scenarios (Real-World Style)

What would you include in a client onboarding questionnaire for Meta ads?

  • Business model, margins, offers, target customers, USP, competitors, past performance data, budget, tracking access, CRM details, sales process, brand guidelines, legal constraints, goals and KPIs.
37AdvancedAdvanced Scenarios (Real-World Style)

How do you handle a client that refuses to install CAPI due to "technical difficulty"?

  • Offer easy options: partner integrations (Shopify, WordPress plugins), CAPI Gateway, Zapier/Make for CRM.
  • Explain data loss impact with numbers (e.g. 20–30% underreporting, weaker optimization).
38AdvancedAdvanced Scenarios (Real-World Style)

You find that the same person is receiving ads from 3 of your campaigns at the same time. Is that bad?

  • Indicates overlap; costs can rise and messaging conflicts.
  • Use exclusions or consolidate; ensure sequential messaging makes sense.
39AdvancedAdvanced Scenarios (Real-World Style)

What would you do when Meta introduces a new feature (e.g. new attribution option or Advantage+ change)?

  • Read official docs, understand impact, test it in a controlled A/B test before rolling out account-wide.
40AdvancedAdvanced Scenarios (Real-World Style)

Scenario: "What's the biggest mistake new media buyers make?"

  • Making too many changes too early (impatience), judging on tiny data, over-fragmenting accounts, ignoring tracking quality and creative, focusing on CPL/CPC instead of business outcomes.
41AdvancedAdvanced Scenarios (Real-World Style)

How would you design a dashboard for a real estate client?

  • Spend, leads, CPL, qualified leads %, cost per qualified lead, site visits, bookings, cost per booking, revenue, by campaign/project/market/creative.
  • Data from Meta + CRM; weekly trend lines; lead response time.
42AdvancedAdvanced Scenarios (Real-World Style)

You have a creative with high spend but the worst CPA in the ad set. Why might Meta still spend on it?

  • Meta predicted it would perform; it may be driving view-through or assisting conversions; or the algorithm hasn't updated predictions yet.
  • Check over enough data; if consistently worse, pause it and see whether ad set CPA improves.
43AdvancedAdvanced Scenarios (Real-World Style)

What would you do if your lead form has a 10% completion rate?

  • Too many fields or confusing questions; or low intent clicks.
  • Reduce fields, use prefill, clear intro screen with value proposition, test More volume vs Higher intent.
44AdvancedAdvanced Scenarios (Real-World Style)

Explain your approach to managing Meta ads for a multi-location franchise.

  • Central brand campaigns + local campaigns per location (radius targeting), local creatives and offers.
  • Shared creative library; location-level reporting; consistent tracking per location (custom conversions/UTMs).
45AdvancedAdvanced Scenarios (Real-World Style)

What do you do when your test results are inconclusive?

  • Extend test duration/budget, check if differences are statistically meaningful, test bigger differences (concepts instead of tweaks).
46AdvancedAdvanced Scenarios (Real-World Style)

You inherit an account where every ad set uses bid cap at AED 5 and nothing spends. What happened?

  • Bid caps too low for auctions; previous buyer may have used bid caps for control.
  • Switch to highest volume or cost cap at realistic CPA; gradually tighten if needed.
47AdvancedAdvanced Scenarios (Real-World Style)

What would you do if you're asked to work with a very low budget, e.g. AED 20/day?

  • One campaign, one ad set, one goal (messages/leads), 2–3 creatives; local targeting; strong offer.
  • Supplement with organic tactics; set expectations on volume and learning.
48AdvancedAdvanced Scenarios (Real-World Style)

Scenario: "Tell me about a time a campaign failed and what you learned." How should you structure your answer?

  • Use STAR: Situation, Task, Action, Result – include metrics.
  • Show analysis (why it failed), what you changed, and the improved result/learning.
49AdvancedAdvanced Scenarios (Real-World Style)

How would you use Meta ads to support a Google Ads campaign?

  • Meta creates demand (awareness, consideration), Google captures search intent.
  • Retarget Google visitors (via pixel) on Meta; measure combined MER and brand search lift.
50AdvancedAdvanced Scenarios (Real-World Style)

When would you choose "Engagement → Messages" vs "Leads → Messaging apps" for WhatsApp campaigns?

  • Engagement/messages: optimize for conversations (cheaper volume).
  • Leads/sales with messaging: optimize for lead/purchase events reported from chats (requires event setup) – better quality.
51AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Campaign with 3 ad sets (CBO, AED 900/day). A: spend 600, 12 purchases, freq 3.4, CTR down 35% WoW. B: spend 250, 8 purchases, freq 1.6. C: spend 50, 1 purchase, freq 1.1. Target CPA 60. What do you do?

  • CPA: A = 50, B = 31, C = 50 (but 1 conversion – not reliable).
  • A is near target but fatiguing (freq 3.4, CTR −35%) → add new creative concepts to A's audience or rely on B.
  • B is most efficient with low frequency → room to grow; consider min spend for B or moving B to its own campaign for vertical scaling.
  • C has too little data – keep for a few more days or consolidate into B if audiences overlap.
  • Overall: refresh creatives, rebalance via structure rather than micromanaging CBO.

Expert insight: Also check spend-weighted CPA trend: A's CPA rising daily while B's stable signals CBO hasn't shifted yet – patience 2–3 days or structural move.

52AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Lead gen real estate: Month 1 – spend 60k, 1,200 leads, 120 qualified, 15 site visits, 2 bookings (AED 3M each, 2% commission). Evaluate and plan Month 2.

  • CPL 50; qualified rate 10% (cost per qualified 500); qualified→visit 12.5%; visit→booking 13%.
  • Revenue (commission) = 2 × 60,000 = AED 120,000 → 2× return on ad spend before costs.
  • Biggest leak: lead→qualified (10%). Plan: higher intent forms, budget/timeline questions, investor-focused creatives, CRM feedback (conversion leads), faster follow-up.
  • Target Month 2: qualified rate 20% at CPL ≤ 70 → cost per qualified ≤ 350.

Expert insight: Also compute pipeline value: 15 visits × typical close rate gives expected future bookings; real estate needs 60–90 day cohort view before judging ROI.

53AdvancedData Case Studies (Analyze & Decide)

DATA CASE: E-commerce: Spend 100k, revenue (Meta-reported) 400k, Shopify total revenue 600k (all channels), last month total revenue 520k with Meta spend 60k. Is the extra 40k spend profitable at 45% gross margin?

  • Incremental revenue (total) ≈ 600k − 520k = 80k (assuming nothing else changed).
  • Incremental gross profit = 80k × 45% = 36k vs incremental spend 40k → slightly unprofitable at the margin.
  • Meta-reported ROAS 4 hides this; check seasonality/other channels, consider pulling back or improving creatives/offers before scaling further.

Expert insight: Seasonality check: compare same month last year growth; if market grew 10% naturally, incremental from Meta is even smaller.

54AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Two creatives over 7 days. Creative X: spend 3,000, hook 32%, CTR 1.9%, 25 purchases. Creative Y: spend 3,000, hook 18%, CTR 1.1%, 33 purchases. Which do you scale and how do you iterate?

  • Y wins on CPA (91 vs 120) despite weaker hook/CTR → audience self-selection/higher purchase intent.
  • Scale Y; iterate Y with stronger hooks (keep its message), and test X's hook style on Y's message (hybrid).
55AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Hotel. Bookings via Meta: 80/month, avg booking AED 1,600, spend AED 18,000. OTA commission 17%. Should the hotel increase Meta spend by 50%?

  • Cost per booking = 225; OTA cost per same booking = 1,600 × 17% = 272 → Meta is cheaper than OTA.
  • Booking ROAS = 128,000 ÷ 18,000 ≈ 7.1.
  • Increase spend gradually and monitor marginal cost per booking; scaling makes sense while marginal cost < 272 and occupancy has room.
56AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Business setup firm. CPL AED 35, 600 leads/month, 6% become clients, avg revenue AED 14,000, margin 35%. Can they afford CPL AED 80 if quality rises to 15%?

  • Now: spend 21,000 → 36 clients → cost per client 583; profit per client 4,900.
  • Proposed: to keep same spend 21,000 → 262 leads → 39 clients → cost per client 533. Better, and less sales workload.
  • Yes – higher CPL with better quality is more profitable.
57AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Frequency 2.6, CTR stable 1.6%, CPA 10% under target, audience = UAE broad 25–55. Client wants +100% budget this month. Plan?

  • Performance is stable despite freq 2.6 → room exists but a 100% jump risks CPA.
  • Plan: +20% every 2–3 days (reach ~2× in ~2 weeks), launch 6–10 new concepts, add a cost-cap duplicate at higher budget, expand to GCC/new languages (horizontal).
  • Watch marginal CPA; stop increases when marginal CPA > target.
58AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Retargeting ad set: audience 18,000, budget AED 400/day, frequency 9/week, CPA rising from 40 to 85. Prospecting CPA 95. What's your decision?

  • Retargeting saturated: cut budget to AED 150–200/day, refresh offer/social proof creatives, extend window/sources.
  • Move saved budget to prospecting to feed the funnel; long-term retargeting depends on new traffic.
59AdvancedData Case Studies (Analyze & Decide)

DATA CASE: A/B test: Broad vs Interest stack, AED 10,000 each, 14 days. Broad: 140 purchases, Interest: 118 purchases. Meta reports 89% confidence for broad. Decision?

  • Broad looks better (CPA 71 vs 85) with fairly high but not 95% confidence.
  • Practical decision: move majority of budget to broad (low risk), keep interests as suggestions; optionally retest with larger budget for confirmation.
60AdvancedData Case Studies (Analyze & Decide)

DATA CASE: CTWA campaign: 900 conversations, 270 replied to the first agent message, 90 qualified, 18 sales (AOV 2,500). Spend 9,000. Where to focus?

  • Cost per conversation 10; reply rate after agent message 30% (biggest leak: 70% go silent); qualified 33% of replies; close 20% of qualified.
  • Revenue 45,000 → ROAS 5.
  • Focus: instant automated greeting, faster response, better opening question, follow-ups within 24 h – raising reply rate to 50% could nearly double sales.

Expert insight: Raising reply rate from 30% to 50% (same downstream rates) → 30 sales instead of 18 → cost per sale 300 instead of 500 – operations is the biggest lever here.

61AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Purchase campaign ROAS by placement: FB Feed 3.2 (spend 40%), IG Feed 2.8 (30%), Reels 2.1 (20%), Audience Network 0.9 (10%). Do you cut Reels and AN?

  • AN at 0.9 with 10% spend: investigate quality; if consistently < break-even, exclude for this campaign.
  • Reels 2.1 may still be valuable at the margin (breakdown effect) and assist conversions; test a Reels-excluded version via A/B before cutting.
62AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Monthly e-commerce data: CPM 38→52, CTR 1.4%→1.5%, CVR 2.2%→2.1%, AOV 280→250, ROAS 3.2→2.1. Explain ROAS drop and actions.

  • ROAS ∝ AOV × CVR × CTR ÷ CPM: CPM +37%, AOV −11%, CVR −5%, CTR +7% → combined ≈ −34% (3.2 → ~2.1).
  • Main drivers: CPM increase (seasonality/competition) and AOV drop (discount mix).
  • Actions: bundles/AOV boosters, creative to improve engagement quality, budget discipline during high-CPM period.

Expert insight: Multiply ratios: 0.893 × 0.955 × 1.071 ÷ 1.368 ≈ 0.67 → 3.2 × 0.67 ≈ 2.14 ✓ matches the reported 2.1.

63AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Learning limited campaign: CPA 120, budget 300/day, 2.5 purchases/day, 4 ad sets. What change gives the best chance to exit learning?

  • Weekly purchases ≈ 17 across 4 ad sets (~4 each) – far from 50.
  • Consolidate to 1 ad set (17/week), and/or optimize for AddToCart temporarily, or raise budget to ~850/day for Purchase learning (50 × 120 ÷ 7).
  • Most practical: consolidate + keep Purchase, accept learning limited if CPA is profitable.
64AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Clinic leads: weekday CPL 45 (contact rate 70%), weekend CPL 30 (contact rate 35%). Which is better?

  • Cost per contacted lead: weekday 64, weekend 86 → weekdays better.
  • Fix weekend follow-up (staff/bot) or shift budget to weekdays via budget scheduling.
65AdvancedData Case Studies (Analyze & Decide)

DATA CASE: You have AED 30k/month. Option A: one country (UAE) CPA 150. Option B: add KSA where CPA is 110 but AOV is 20% lower. Which gives better return if UAE AOV = 450?

  • UAE ROAS = 450 ÷ 150 = 3.0; KSA AOV = 360 → ROAS = 360 ÷ 110 = 3.27.
  • KSA slightly better on ROAS; also check margins, shipping costs, return rates, COD refusals for KSA before shifting.
66AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Video ad metrics: 200k impressions, 3s views 50k, ThruPlays 6k, link clicks 1.2k, purchases 12, spend 6k. Diagnose.

  • Hook 25% (good), hold 12% of 3s viewers (weak), CTR 0.6% (low), CVR 1% of clicks, CPA 500.
  • Biggest issue: weak hold and CTR – the body/CTA doesn't convince. Re-edit: show product/benefit earlier, shorter, clearer CTA and offer.
67AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Meta says 200 purchases; backend says 160 orders from Meta UTMs; total orders rose by 100 vs last month after Meta spend increased by 30k. AOV 300. Evaluate.

  • Incremental orders (approx.) = 100 → incremental revenue 30,000 → incremental ROAS ≈ 1.0 on the extra 30k.
  • Meta claims more than incremental; the extra spend broke even on revenue (loss on profit). Reallocate or improve before more scaling; run lift test.

Expert insight: Recommend a structured geo-holdout next month to quantify incrementality before further budget increases.

68AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Lead campaign for education. Form A (5 questions): CPL 60, enrollment rate 8%. Form B (2 questions): CPL 25, enrollment rate 2.5%. Which form?

  • Cost per enrollment: A = 750; B = 1,000 → Form A better.
  • Possibly test a 3–4 question version to find the optimum.
69AdvancedData Case Studies (Analyze & Decide)

DATA CASE: New account, 10 days, AED 2,000 spent, 0 purchases, CTR 1.8%, 900 LPV, 40 ATC, 6 IC. Is the problem ads or website?

  • Ads deliver interested traffic (CTR good, ATC rate 4.4%). Drop happens at checkout (40 ATC → 6 IC → 0 purchases).
  • Website/checkout problem: shipping costs, payment methods, trust, errors – test checkout yourself; also confirm Purchase event works.
70AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Advantage+ sales: spend 50k, ROAS 4.5; audience segments: existing customers 45% of spend with ROAS 8, new 55% with ROAS 1.6. Break-even ROAS 2.5. What do you do?

  • New customer acquisition is below break-even; existing customers inflate total ROAS (and many would buy anyway).
  • Set existing customer cap ~15–20%, improve new-customer creatives/offers, evaluate LTV to see if 1.6 first-order ROAS is acceptable long-term.

Expert insight: Compute blended contribution: existing customers ROAS 8 × 45% spend + new 1.6 × 55% ≈ 4.5 ✓; incremental growth comes only from the new segment.

71AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Cost cap 80 set; campaign spends 35% of budget at CPA 72. Business can accept CPA 95. Next step?

  • There's headroom: raise cap to ~90 in steps; expect more spend at slightly higher CPA still within 95 tolerance.
  • Monitor marginal CPA as spend rises.
72AdvancedData Case Studies (Analyze & Decide)

DATA CASE: 5 creatives in an ad set. Spend share: C1 70% (CPA 55), C2 20% (CPA 48), C3–C5 10% total (CPA 120). What actions?

  • C2 has better CPA but less spend – consider giving it a dedicated test/ad set to see if it scales.
  • Keep C1 (main driver), pause C3–C5 if CPA stays poor with adequate spend, and add new diverse concepts.
73AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Brand awareness campaign: reach 1.2M, freq 4.5 in 2 weeks, ad recall lift 6%, search volume flat. Advice?

  • Frequency high for awareness with low lift – creative may not be memorable/branded; use frequency cap (e.g. 2/week), stronger branding in first seconds, broader reach.
  • Measure brand lift properly; tie to search/direct traffic over longer period.
74AdvancedData Case Studies (Analyze & Decide)

DATA CASE: A clinic wants 200 appointments/month. Current: CPL 40, appointment rate 30%, show-up 70%. Budget needed for 200 shows?

  • Appointments needed = 200 ÷ 0.7 ≈ 286; leads = 286 ÷ 0.3 ≈ 953; budget ≈ 953 × 40 ≈ AED 38,000 (plus buffer for rising CPL at scale).
75AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Purchase CPA rose 40% after you added 15 new ads into the scaling ad set at once. Rollback plan?

  • Pause most new ads (keep 3–5 most diverse), let ad set stabilize 3–4 days; move other new ads into a testing campaign; in future add new ads in batches weekly.

Expert insight: For future, pre-test new concepts in ABO or add in batches of 3–5; it's a common client conversations "lesson learned" story.

76AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Agency report: CTR 3%, CPC AED 0.80, 50,000 clicks, but sales unchanged. Client asks if agency did well. Your assessment?

  • These are traffic metrics; without conversions/sales impact they don't show business value; maybe optimized for clicks/Traffic objective.
  • Ask for conversion data, objective used, tracking; recommend conversion-optimized campaigns with business KPIs.
77AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Lookalike seeded from all leads gives CPL 30 but qualified rate 5%; lookalike from closed customers gives CPL 55 with 18% qualified. Which seed?

  • Cost per qualified: 600 vs 305 → closed customers seed is better; also consider using it as suggestion in Advantage+ audience.
78AdvancedData Case Studies (Analyze & Decide)

DATA CASE: You have 7 days and AED 7,000 for a new restaurant opening: target 1,000 local people to know + 150 reservations. Allocation?

  • ~40% Reach (5 km radius, freq cap) with opening video to build awareness.
  • ~50% messaging/calls for reservations (WhatsApp/IG DM) with opening offer.
  • ~10% retarget video viewers with menu/offer; partnership with 1–2 local creators.
  • Track reservations via WhatsApp tags/promo codes.
79AdvancedData Case Studies (Analyze & Decide)

DATA CASE: Purchases: Monday–Thursday CPA 70, Friday–Sunday CPA 110 consistently for 8 weeks. What do you do?

  • Real pattern: use budget scheduling to lower weekend budgets (or increase weekday), or test weekend-specific creatives/offers.
  • Avoid ad set on/off toggling; keep campaigns running.
80AdvancedExpert Follow-up Drills

You said "consolidate to fix fragmentation". clients: "But consolidation increased our CPA by 25% in week one. Do you roll back?"

  • Expected: merging resets learning; compare against the old structure's trailing 14–28 day CPA, not week one.
  • Give it 7–14 days or ~50 conversions; check if new campaign is exiting learning and CPA is trending down.
  • Roll back only if, after learning, CPA stays worse and conversion volume isn't higher; keep a small share in the old structure during migration to reduce risk.
81AdvancedExpert Follow-up Drills

You said "scale 20% every 2–3 days". clients: "Why 20%? Show me when you'd break that rule."

  • 20% is a heuristic to limit shocks and keep pacing stable; not a Meta rule.
  • Break it when: big seasonal windows (budget scheduling), cost-cap/ROAS-goal campaigns (spend governed by cap, so budget can jump), very high-volume ad sets (hundreds of conversions/week absorb larger changes), or when duplicating at higher budget.
  • Always judge by marginal CPA after the change.
82AdvancedExpert Follow-up Drills

You recommended broad targeting. clients: "Our pixel has only 20 purchases in total. Is broad still right?"

  • With little data, broad depends heavily on creative signals and Meta's general models; it can work but may be slower/noisier.
  • Options: use Advantage+ audience with interest/LAL suggestions, optimize for a higher-volume event (ATC/IC) first, and build data quickly; move to pure broad once signals accumulate.
83AdvancedExpert Follow-up Drills

You said "use CAPI". clients: "We installed CAPI and reported conversions went up 30% but sales didn't change. Did CAPI work?"

  • Yes – CAPI improves measurement (recovering lost signals), not demand directly; reported uplift shows better attribution.
  • The real benefit appears over weeks: better optimization → possibly lower true CPA; check dedupe to ensure the 30% isn't double counting.
84AdvancedExpert Follow-up Drills

You recommended conversion leads optimization. clients: "Our CRM updates stages after 3 weeks. Will it still work?"

  • Slow updates weaken learning; Meta needs timely stage events.
  • Fix: define an earlier qualification stage (e.g. contacted + budget confirmed within 48 hours) and send it in near real time; keep later stages for reporting/value.
85AdvancedExpert Follow-up Drills

You said "check frequency". clients: "Frequency is 1.8 but performance is collapsing on a 10M audience. What now?"

  • Frequency average hides concentration: a subset of high-intent users may see ads 8–10 times while most see once; check frequency distribution where available and CTR by ad age.
  • Also check creative diversity (Andromeda clustering), auction competition, and post-click changes.
86AdvancedExpert Follow-up Drills

You suggested "new creative concepts". clients: "We launched 10 new concepts, none spent. Why?"

  • Retrieval/ranking predicts them weaker than the incumbents, or they're too similar to existing ads (entity clustering), or the ad set is saturated by a dominant ad.
  • Fix: launch them in a separate testing ad set with guaranteed budget, make concepts genuinely different (format, persona, angle), check they're approved and placement-fit.
87AdvancedExpert Follow-up Drills

You said "Advantage+ sales". clients: "How would you prove Advantage+ beats our manual setup without contaminating results?"

  • Use Meta's A/B test with campaign-level split (Advantage+ vs manual), same budget, same creatives, 2–4 weeks; primary KPI cost per purchase/new-customer CAC; check backend orders by UTM/campaign.
88AdvancedExpert Follow-up Drills

You advised "cost cap". clients: "Cost cap campaigns spent everything on day 1 of the month and then nothing. Why?"

  • Possible high-opportunity early period (e.g. paydays, sale) followed by expensive auctions; or budget was lifetime with cost cap pacing; or new learning predictions shifted.
  • Review daily CPA, use daily budgets, set realistic caps and monitor auction conditions; combine with highest-volume ad set as a baseline.
89AdvancedExpert Follow-up Drills

You said "frequency above 2 → horizontal scaling". clients: "Our product is a high-consideration luxury item; frequency 5 converts best. Defend or change your rule."

  • Change it: for high-consideration products, repeated exposure is needed (more touchpoints); frequency is fine if CPA/quality hold.
  • Rule should be performance-led: scale vertically while marginal CPA is good; go horizontal when marginal CPA rises with frequency.
90AdvancedExpert Follow-up Drills

You recommended excluding Audience Network. clients: "Meta says excluding placements increases cost. Who is right?"

  • Both can be right: Meta optimizes for the event – if the event is a click/LPV, AN can deliver cheap but low-quality results; for conversion events with good tracking, AN spend is usually limited automatically.
  • Decide by downstream quality (CRM/backend), not platform CPA, and test exclusion with an A/B split.
91AdvancedExpert Follow-up Drills

You said "use lead form qualifying questions". clients: "People lie in forms – they choose the highest budget. Then what?"

  • Combine stated answers with behavioral validation: OTP, response to call/WhatsApp, follow-up questions; score leads using actual sales outcomes and send those to Meta (not just stated budget).
92AdvancedExpert Follow-up Drills

You said "MER". clients: "MER improved but we cut brand spend. Is that real improvement?"

  • Short-term MER can rise because brand effects lag (adstock); demand may decline later. Monitor brand search, direct traffic, new customer rates over 2–3 months; use MMM/holdouts for brand impact.
93AdvancedExpert Follow-up Drills

You said "run a lift test". clients: "We spend AED 20k/month – is a lift study feasible?"

  • Often underpowered at small budgets; alternatives: geo holdout with on/off weeks, pre/post with control markets, or longer-duration tests; accept directional evidence and combine with MER trends.
94AdvancedExpert Follow-up Drills

You suggested "value optimization". clients: "After switching, purchases dropped 30% and revenue rose 5%. Good or bad?"

  • Depends on profit: fewer orders with higher AOV can be better if margin per order is higher and fulfilment costs drop; compute contribution margin and new customer counts, not just revenue.
95AdvancedExpert Follow-up Drills

You said "event_id dedupe". clients: "Our Purchase dedupe rate is 100% but EMQ is 4. What's the problem?"

  • Dedupe works (same IDs), but server events lack customer information (email/phone/fbc/IP/UA) → poor match quality; enrich server events with hashed customer data.
96AdvancedExpert Follow-up Drills

You suggested "WhatsApp campaigns". clients: "Our agents can only handle 50 chats/day. How do you design delivery?"

  • Cap daily budget to expected chats (budget ≈ 50 × cost per conversation), use automated qualification bot, schedule ads to agent hours, prioritize high-intent creatives, and route overflow to forms.
97AdvancedExpert Follow-up Drills

You recommended "existing customer budget cap". clients: "Our repeat customers are actually our most profitable segment – why limit them?"

  • If retention via email/WhatsApp is cheaper, don't pay Meta to reach them; but if Meta drives incremental repeat purchases (proven via holdout), allow higher share. Decision depends on incrementality, not ROAS.
98AdvancedExpert Follow-up Drills

You said "use partnership ads". clients: "The creator's audience is in a different country than our market. Does it matter?"

  • Partnership ads use your targeting, not the creator's followers – what matters is content resonance with your audience; but creator familiarity in-market boosts trust. Test creators local to the target market.
99AdvancedExpert Follow-up Drills

You said "test concepts not variations". clients: "Give me a concrete definition of 'different concept' for our perfume brand."

  • Different persona (gift buyer vs self-buyer), different emotional angle (confidence vs romance vs status), different format (UGC unboxing vs studio luxury vs comparison), different setting/person; not just color/text changes.
100AdvancedExpert Follow-up Drills

You said "Housing SAC limits targeting". clients: "Can we run the ad without SAC since it's 'just brand awareness' for our developer?"

  • If the ad promotes housing opportunities/listings, SAC applies regardless of objective; mislabeling risks rejection/restriction. Pure corporate brand content (no listings/offers) may differ – check policy and be conservative.
101AdvancedExpert Follow-up Drills

You said "speed to lead under 5 minutes". clients: "Leads come at 2 AM. What's your system?"

  • Instant automated WhatsApp/SMS with info + booking link, chatbot qualification, morning priority queue; consider ad scheduling if night leads are poor quality; measure contact rate by hour.
102AdvancedExpert Follow-up Drills

You recommended "consolidated ad sets by language". clients: "Arabic ad set spends 10% of budget in CBO. Is that a problem?"

  • Only if business needs Arabic speakers specifically; check cost per qualified by language; if Arabic leads are more valuable, use ad set minimum spend or separate campaign/value rules.
103AdvancedExpert Follow-up Drills

You said "look at marginal CPA". clients: "How do you compute marginal CPA when daily data is noisy?"

  • Use multi-day windows before/after budget changes (e.g. 7 days each), regression of conversions vs spend over weeks, or controlled budget tests (split campaigns) to estimate the response curve.
104AdvancedExpert Follow-up Drills

You suggested "catalog retargeting". clients: "Our catalog has 50 products; is dynamic retargeting worth it?"

  • Possibly marginal; with a small catalog, creative-led retargeting (testimonials, offers) may perform similarly. Test incrementality; dynamic ads shine with large catalogs.
105AdvancedExpert Follow-up Drills

You said "use 7-day click attribution". clients: "Our product is impulse (AED 49). Which window and why?"

  • 1-day click (or 7-day click compare) may better represent impulse purchases and reduce over-crediting; but changing optimization window affects delivery – test 1-day vs 7-day click via A/B.
106AdvancedExpert Follow-up Drills

You said "MMM". clients: "How would you validate an MMM result that says Meta's ROI is 3x?"

  • Calibrate with lift/geo experiments; check model fit on holdout periods; sanity-check saturation curves vs budget tests; compare with business intuition (does cutting Meta spend reduce sales as predicted?).
107AdvancedExpert Follow-up Drills

You said "split by country for GCC". clients: "Kuwait has only AED 50/day budget – separate ad set will be learning limited. Now what?"

  • Group small markets (Kuwait, Bahrain, Oman, Qatar) into a cluster ad set; use country breakdown for analysis; value rules or creatives for local pricing; separate only big markets.
108AdvancedExpert Follow-up Drills

You said "Higher intent forms". clients: "Higher intent increased CPL from 40 to 110 and the client may leave. How do you manage?"

  • Show cost per qualified and per deal side-by-side; propose a blended approach (some budget on more-volume form with strong qualification questions); set a 30-day evaluation with agreed success metrics.
109AdvancedExpert Follow-up Drills

You said "Andromeda favors diversity". clients: "Is there any downside of too many creatives in one ad set?"

  • Yes: spend dilution, slower learning per ad, harder analysis, operational load; diminishing returns once concepts are covered. Keep enough diversity, prune non-spenders, and prioritize quality.
110AdvancedExpert Follow-up Drills

Final challenge: "If Meta removed all targeting tomorrow (only country), how would you still win?"

  • Creative strategy (message-market match, personas via creative), offer engineering, landing page CVR, clean CAPI/CRM signals, consolidation and testing velocity, incrementality measurement – the same levers that win today.
Facebook ads scenario interview questions

More Meta ads question topics

Want an expert to run your Meta ads?

Book a free strategy call and get a clear plan for your campaigns.

Book a free call