Question 1

Your Facebook Ads campaign has an abnormally high CPM and low CTR after 5 days of running. How do you diagnose the problem and optimize?

Accepted Answer

Diagnostic Framework - Top-down analysis funnel: (1) Audience (saturation, overlap check) → (2) Creative (ad fatigue, relevance score) → (3) Offer/Landing Page (message match) → (4) Bidding Strategy (cost cap vs bid cap vs lowest cost) → (5) Timing & Placement. Use Facebook Ads Manager Breakdown to segment data by device, placement, time. Good candidates should mention: Check Frequency (>3 signals ad fatigue); Audience Overlap tool; test new creatives (static vs video vs carousel); check Auction Overlap; question the original objective (awareness vs conversion). For Vietnam market: Mobile-first placement, golden hours 7-10 PM.

Question 2

You have a 100M VND/month budget for Google Ads, needing to allocate between Search and Display. What is your strategy and why?

Accepted Answer

Allocate by purchase funnel: Search (bottom-of-funnel: intent-driven, high conversion) vs Display (top-of-funnel: awareness, retargeting). Principle: 70-80% Search for high search volume industries; 50/50 for markets still being educated. Calculation: Search Volume × estimated CPC → required budget. Display: Lower CPM/CPC but lower conversion rate. Outstanding candidates will: Ask about the industry and objectives first; mention Google's Performance Max (PMax) and how it changes allocation; bring up Smart Bidding (tCPA, tROAS); analyze high-intent keywords (branded, problem-aware, solution-aware).

Question 3

Your data shows Facebook Ads leading in conversions, but Google Analytics shows completely different numbers. How do you explain this discrepancy and which attribution model do you use?

Accepted Answer

Attribution Models: Last-click (GA default) vs View-through (Facebook default) vs Data-driven vs Multi-touch (Linear, Time Decay, Position-based). Discrepancy causes: Facebook counts view-through (impressions); different cookie windows; cross-device tracking; ad blockers; GA4 vs UA differ in session logic. Good candidates should mention: MMM (Marketing Mix Modeling) for a holistic view; Incrementality testing (holdout group) to measure true impact; UTM parameter discipline; presenting to stakeholders using 'directional accuracy' rather than absolutizing one data source.

Question 4

Campaign ROAS is at 2x but the target is 4x. What steps would you take to improve ROAS within 4 weeks without increasing the budget?

Accepted Answer

ROAS = Revenue / Ad Spend. Two main levers: (1) Increase Revenue: AOV (upsell/cross-sell), Conversion Rate (landing page, offer), funnel leaks; (2) Decrease Ad Spend more efficiently: remove poor audience/placements, optimize bidding, negative keywords. 4-week roadmap: Week 1 (audit) → Week 2 (quick wins) → Week 3 (test) → Week 4 (scale winners). Bonus points for mentioning: Cohort analysis to identify high-value audiences; Product feed optimization for e-commerce; focusing retargeting on high-intent users (added to cart but didn't buy); changing offer (discount, bundle, free shipping).

Question 5

Build a complete retargeting strategy for a Vietnamese e-commerce website with an 80% cart abandonment rate. How would you segment audiences and tailor messages?

Accepted Answer

Retargeting Funnel - Audience Segmentation: (1) Homepage visitors (awareness ad) → (2) Product page viewers (product-specific ad) → (3) Cart abandoners (urgency + incentive, 1-3 days) → (4) Checkout abandoners (friction reduction, 24h) → (5) Past buyers (upsell/cross-sell). Channels: Facebook/Instagram + Google Display + Zalo Ads (Vietnam-specific). Exclusion: Recent purchasers. Strong candidates will mention: Dynamic Product Ads (DPA) for e-commerce; Zalo Ads retargeting specific to Vietnam; frequency caps to avoid annoying users; A/B test messaging (discount vs free shipping vs FOMO); parallel email retargeting (if emails available).

Question 6

Your landing page has good traffic but a conversion rate of only 1.2%, 3x below industry benchmark. How would you conduct CRO (Conversion Rate Optimization)?

Accepted Answer

CRO Process: (1) Quantitative Analysis (GA: exit rate, scroll depth, heatmap) → (2) Qualitative Research (session recording: Hotjar/Microsoft Clarity, user survey) → (3) Hypothesis Formation → (4) A/B Testing (variables: headline, CTA, social proof, form length) → (5) Measure & Scale. LIFT Model to evaluate page: Value Proposition, Relevance, Clarity, Anxiety, Distraction, Urgency. Good candidates will mention: Page speed (especially Mobile in Vietnam - 4G dominant); culturally appropriate social proof for Vietnam (real reviews, purchase counts); trust signals (warranty, returns, hotline); how to calculate minimum sample size for A/B tests.

Question 7

You have a 500M VND/month performance marketing budget. The product is a fintech app targeting 25-40 year olds in Vietnam. How would you allocate the budget across channels?

Accepted Answer

Channel Selection Framework based on: Intent Level × Audience Size × Historical CPL × Scalability. Suggested allocation for Vietnamese fintech: Google Search (branded + competitor keywords) ~30%, Facebook/Instagram (lookalike, interest) ~35%, TikTok (younger segment) ~15%, Zalo Ads (mature segment, high trust) ~10%, ASO/App Store Ads ~10%. Always reserve 10-15% budget for testing new channels. Outstanding candidates will: Ask about current install base and organic vs paid ratio; mention Zalo Ads as a Vietnam-specific channel with high trust; differentiate CPI (cost per install), CPL (cost per lead), and CAC; mention MMPs (Mobile Measurement Partners) like Appsflyer/Adjust.

Question 8

After iOS 14.5+, your Facebook Ads data dropped significantly making campaign optimization harder. What have you done and are doing to adapt to this?

Accepted Answer

Privacy-first Measurement Stack: (1) Conversions API (CAPI) - server-side tracking replacing pixel; (2) Aggregated Event Measurement - prioritize 8 most important events; (3) Modeled Conversions - Facebook fills in missing data; (4) First-party data strategy (email list, CRM); (5) Broader audience targeting instead of hyper-targeted. Combine: GA4, server-side GTM. Good candidates will mention: Event Match Quality score in Facebook; Conversion API Gateway (no developer required); increasingly important role of first-party data; lesser impact on Android and implications for Vietnamese market (Android majority).