Online Advertising & Analytics: Metrics, Models & Tools

TAKEAWAYS:

• Online ads use cost-per-click (CPC), cost-per-thousand impressions (CPM), and cost-per-action (CPA) pricing models, each suited to different campaign goals.
• Key metrics like impressions, CTR, and page views are tracked via tools like Google Analytics to gauge ad performance and user engagement.
• Google Analytics and social media analytics (Facebook Insights, Instagram Analytics) help e-commerce businesses refine targeting, content, and ROI.
• A/B testing and personalization (e.g., Daraz’s product recommendations) improve ad relevance and conversion rates.
• Programmatic advertising automates ad buying/selling using real-time bidding (RTB) platforms like Google Display & Video 360.
• Data-driven decisions (e.g., Ncell’s ad campaigns) rely on analytics to allocate budgets and optimize ad spend.

1. Introduction to Online Advertising

Online advertising is the process of delivering promotional content (ads) to users via digital channels like websites, social media, search engines, and mobile apps. Unlike traditional media, digital ads are targeted, measurable, and interactive, allowing businesses to track performance in real time.

Key Players in Online Advertising

flowchart TD
    A["Advertisers"] -->|"Creates ads"| B["Ad Networks"]
    B -->|"Distributes ads"| C["Publishers"]
    C -->|"Shows ads to users"| D["Users"]
    D -->|"Engages with ads"| E["Analytics Tools"]
    E -->|"Reports data"| B
    B -->|"Uses data"| F["Demand-Side Platform (DSP)"]
    F -->|"Bids for ads"| B

• Advertisers: Brands (e.g., Daraz, Ncell) paying to promote products/services.
• Ad Networks: Middlemen (Google AdSense, Facebook Audience Network) that connect advertisers to publishers.
• Publishers: Websites/apps (e.g., Kantipur Online, YouTube) displaying ads.
• Users: End consumers clicking, viewing, or interacting with ads.
• Analytics Tools: Google Analytics, Meta Business Suite track performance.

2. Online Advertising Pricing Models

Advertisers pay publishers based on different metrics. The choice depends on campaign goals (brand awareness vs. sales).

Worked Example: Daraz’s CPM vs. CPC Campaign

• CPM Campaign: Daraz runs a banner ad on Kantipur Online to boost brand visibility. They pay ₹500 per 1,000 impressions. If the ad is shown 50,000 times, the cost is ₹25,000.
• CPC Campaign: For a sale, Daraz targets users searching for "laptops" on Google. They bid ₹20 per click. If 1,000 users click, the cost is ₹20,000—but they may generate 50 sales at ₹15,000 each, netting ₹750,000 in revenue.

3. Key Online Marketing Metrics

Metrics quantify ad performance. Here’s how they’re calculated and interpreted:

Impressions

• Definition: Number of times an ad is displayed (even if not clicked).
• Formula: $$\text{Impressions}=\text{Number of ad views}$$
• Why it matters: High impressions = broad reach, but low CTR may mean poor targeting.
• Example: A Ncell ad on Facebook is shown 10,000 times (10,000 impressions).

Click-Through Rate (CTR)

• Definition: Percentage of users who click an ad after seeing it.
• Formula: $$\text{CTR}=\left(\frac{\text{Number of Clicks}}{\text{Impressions}}\right)\times 100$$
• Good CTR: Typically 2–5% for display ads; 5–10% for search ads.
• Worked Example: If a Daraz ad gets 50 clicks out of 2,000 impressions: $$\text{CTR}=\left(\frac{50}{2000}\right)\times 100=2.5\%$$

Page Views

• Definition: Number of times a user visits a webpage (often linked to ads).
• Why it matters: High page views indicate strong ad engagement.
• Example: A blog post about "best smartphones in Nepal" gets 5,000 page views, many from Google ads.

Conversion Rate

• Definition: Percentage of users who complete a desired action (e.g., purchase, sign-up).
• Formula: $$\text{Conversion Rate}=\left(\frac{\text{Number of Conversions}}{\text{Clicks}}\right)\times 100$$
• Example: If 100 users click a Khalti ad and 10 sign up for a wallet: $$\text{Conversion Rate}=\left(\frac{10}{100}\right)\times 100=10\%$$

Cost per Conversion (CPCV)

• Definition: Average cost to acquire a customer/conversion.
• Formula: $$\text{CPCV}=\frac{\text{Total Ad Spend}}{\text{Number of Conversions}}$$
• Example: If ₹10,000 is spent on ads and 50 users sign up for eSewa: $$\text{CPCV}=\frac{10,000}{50}=₹200\text{ per conversion}$$

4. Tools for Online Advertising and Analytics

Google Analytics

Google Analytics tracks website performance, user behavior, and ad effectiveness. Key features:

• Traffic Sources: Shows where visitors come from (organic search, social media, ads).
• User Behavior: Tracks page views, bounce rate, and time spent.
• Conversions: Measures goals (e.g., purchases, sign-ups) tied to ads.

Example: A Daraz store uses Google Analytics to see that 30% of traffic comes from Facebook ads, with a 5% conversion rate.

Social Media Analytics

Platforms like Facebook, Instagram, and LinkedIn provide built-in analytics:

• Facebook Insights: Shows post reach, engagement, and audience demographics.
• Instagram Analytics: Tracks profile visits, story interactions, and link clicks.
• LinkedIn Analytics: Useful for B2B ads (e.g., NEPSE stock tips).

Example: Ncell runs a LinkedIn ad campaign targeting IT professionals. Analytics show that 40% of clicks come from users aged 25–34, leading to a targeted retargeting strategy.

5. Programmatic Advertising

Programmatic advertising uses automation and AI to buy/sell ads in real time. It eliminates manual negotiations and speeds up the process.

How It Works

1. Demand-Side Platform (DSP): Advertisers (e.g., Daraz) use tools like Google Display & Video 360 to bid on ad space.
2. Supply-Side Platform (SSP): Publishers (e.g., Kantipur Online) list available ad slots.
3. Real-Time Bidding (RTB): Ads are auctioned in real time (milliseconds) when a user loads a page.
4. Ad Server: Delivers the winning ad to the user.

Mermaid Diagram of Programmatic Ad Flow

sequenceDiagram
    participant User
    participant Publisher
    participant SSP
    participant DSP
    participant Ad Server

    User->>Publisher: Loads webpage
    Publisher->>SSP: Sends ad request
    SSP->>DSP: Auction (RTB)
    DSP->>Ad Server: Wins bid (e.g., Daraz ad)
    Ad Server->>User: Displays ad

Advantages:

• Efficiency: Automates ad buying/selling.
• Targeting: Uses user data (demographics, interests) for precision.
• Transparency: Real-time reporting.

Disadvantages:

• Complexity: Requires technical expertise.
• Privacy Concerns: Heavy reliance on user data.

Example: Google’s Display & Video 360 helps Ncell run programmatic ads on YouTube and websites, targeting users interested in mobile plans.

6. A/B Testing and Personalization

A/B Testing

• Definition: Testing two versions (A and B) of an ad to see which performs better.
• Example: Daraz tests two banner designs:
  • Version A: "50% Off on Laptops!"
  • Version B: "Limited-Time Deal: Buy Now!" Results show Version B has a 3% higher CTR.

Personalization

• Definition: Customizing ads based on user behavior (e.g., past purchases, browsing history).
• Example: Khalti sends personalized offers like:
  • "Complete your wallet setup for ₹100 cashback!"
  • "Refer a friend and get ₹500!"

Mermaid Diagram of Personalization Workflow

flowchart TD
    A["User Data"] -->|"Browsing history, purchases, demographics"| B["Data Processing"]
    B -->|"AI/ML algorithms"| C["Targeted Ad"]
    C -->|"Delivered via"| D["Khalti App/Website"]
    D -->|"User interaction"| E["Conversion Tracking"]
    E -->|"Feedback loop"| B

7. Real-World Applications in Nepal

1. Daraz’s Dynamic Product Ads

• Idea Used: Personalization + Programmatic Ads
• How: Daraz uses user browsing history to show dynamic ads. If a user views a smartphone, they see ads for related accessories (chargers, cases) via programmatic bidding.
• Result: 15% higher conversion rate for personalized ads.

2. Ncell’s YouTube Pre-Roll Ads

• Idea Used: CPV (Cost-Per-View) + Targeting
• How: Ncell runs 30-second pre-roll ads before tech tutorials on YouTube, targeting users aged 18–35. They pay ₹10 per view.
• Result: 8% of viewers click the ad, leading to a 2% sign-up rate for new plans.

3. eSewa’s Retargeting Campaigns

• Idea Used: A/B Testing + Retargeting
• How: eSewa tests two retargeting ads:
  • Ad A: "Complete your profile for ₹100 bonus!"
  • Ad B: "Unlock exclusive deals—sign up now!" Analytics show Ad B performs 20% better, so it’s scaled up.

8. In the Real World

1. Google Ads (Global)

   • Idea: CPC & CPA Pricing + Real-Time Analytics
   • How: Google uses AI to match ads with user intent. For example, a search for "best laptop Nepal" triggers ads from Daraz, with CPC bids ranging from ₹10–₹50. Google Analytics tracks which ads lead to purchases.

2. Meta (Facebook/Instagram) Ads (Global)

   • Idea: CPM & CTR Optimization
   • How: Meta’s algorithm shows ads to users based on interests (e.g., "gaming" or "finance"). For Ncell’s 5G launch, Meta ran a CPM campaign targeting gamers, achieving a 4% CTR and ₹300 ROI per ₹100 spent.

3. YouTube (Global)

   • Idea: CPV + Programmatic Bidding
   • How: YouTube’s ad platform uses RTB to sell pre-roll ads. NTC runs a CPV campaign for its internet plans, paying ₹8 per view. Users who watch 30 seconds are 3x more likely to click the ad.

9. Exam Tip

• Focus on definitions and formulas: Exam questions often ask for CTR, conversion rate, or CPCV calculations. Memorize the formulas:
  • $$\text{CTR}=\frac{\text{Clicks}}{\text{Impressions}}\times 100$$
  • $$\text{Conversion Rate}=\frac{\text{Conversions}}{\text{Clicks}}\times 100$$
• Compare pricing models: Know when to use CPM (awareness), CPC (traffic), or CPA (conversions). Example: A bank ad for loans uses CPA; a brand ad for awareness uses CPM.
• Link real-world examples: Always tie concepts to Nepalese companies (Daraz, eSewa, Ncell) or global platforms (Google, Meta). For instance, explain how Daraz uses Google Analytics to track ad performance.
• Diagrams are key: Draw flowcharts for programmatic advertising or A/B testing to visualize processes.
• Avoid vague answers: Instead of "analytics helps track performance," say:

  "Google Analytics tracks user behavior (e.g., bounce rate, time on page) to optimize ad spend. For example, if Daraz sees high bounce rates from mobile users, they adjust ad creatives for better engagement."

Final Note: This unit blends theory (metrics, models) with practical tools (Google Analytics, programmatic ads). Focus on understanding how these tools are used in Nepal’s e-commerce ecosystem (Daraz, Khalti, Ncell) to ace both theory and application-based questions.