Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
DD R2
| Question | Answer |
|---|---|
| What happens when there are 10,000 campaigns | Use pagination or virtualization so only visible rows render. |
| How do you virtualize tables | Render only what’s on screen using a windowing library. |
| How do you debounce filters and why | Delay the filter action until the user stops typing for ~300ms. to prevent performance issues server reloads |
| How do you show loading vs skeletons | Skeletons for layout stability, spinners for full-page loads. |
| How do you prevent flicker | Keep previous data on screen while new data loads. |
| How do you support mobile | Responsive layouts and touch-friendly controls. |
| How do you make this accessible | Keyboard navigation, labels, and ARIA for dynamic content. |
| How do you handle broken data | Fallback UI with clear error messages and retry. |
| How do you let users compare campaigns | Multi-select rows or side-by-side charts. |
| What metrics do you log from this UI | Views, filter usage, clicks, exports, and campaign actions. |
| What is an ad impression | An impression is counted when an ad is shown to a user. |
| What is a click | A click is when a user taps or clicks an ad. |
| What is CTR | Click-through rate = clicks ÷ impressions, measuring how engaging an ad is. |
| What is a conversion | A conversion is a desired outcome after clicking, like a purchase or sign-up. |
| What is ROAS | Return on ad spend = revenue ÷ ad spend, showing ad profitability. |
| What is CPC vs CPA | CPC is cost per click; CPA is cost per conversion. |
| What does “budget pacing” mean | How fast a campaign is spending its budget over time. |
| What does “bid” mean | The amount an advertiser is willing to pay per click or impression. |
| What is an ad auction | A real-time system that chooses which ad to show based on bids and relevance. |
| What is creative rotation | Automatically showing different ad creatives to test performance. |
| How do we avoid double counting | Use unique event IDs and dedupe during aggregation. |
| How do we handle delayed events | Allow late data and backfill metrics when it arrives. |
| What happens if tracking is broken | Detect drops in event volume and flag missing data. |
| How do we validate metrics | Compare raw logs to aggregates and monitor ratios. |
| How do you backfill missing data | Replay historical logs to recompute aggregates. |
| How do you detect fraud | Look for abnormal click rates, IPs, or behavior patterns. |
| What happens if impressions > requests | That signals logging or aggregation errors. |
| How do you handle partial failures | Show partial data with warnings instead of blocking the UI. |
| What does data freshness mean | How up-to-date the metrics are. |
| How do you reconcile logs vs aggregates | Use logs as ground truth and debug mismatches. |
| How would you A/B test a new ad format | Split users into control and test and compare key metrics. |
| What is statistical significance | The likelihood that results aren’t due to random chance. |
| What is a control group | Users who see the existing version. |
| What metrics matter for success | CTR, conversions, revenue, and guardrails like user experience. |
| How do you avoid misleading results | Use enough data and avoid cherry-picking time windows. |
| What if users switch buckets | Use consistent assignment to prevent contamination. |
| How long should tests run | Long enough to reach statistical significance. |
| How do you interpret noisy data | Look for trends and confidence intervals, not just point values. |
| How do you present results to PMs | Show uplift, confidence, and business impact. |
| What would you show advertisers | Clear performance metrics like spend, clicks, and ROAS. |
| Question to ask before solving | Who is the user and what decision are they making? Is the data real-time, near-real-time, or batched daily? How accurate is the data and how often does it change? What scale are we talking about? |
Created by:
Mikay.8