Master a Full Campaign Metrics Teardown for Interviews
A Diwali ad that sold chocolate also gave local tailors, jewellers and electronics shops a celebrity ad of their own. That is what made Cadbury Celebrations' “Not Just A Cadbury Ad” more than a festive campaign - it forced marketers to ask a sharper question: did the campaign create business impact, or only beautiful noise?
- A campaign metrics teardown connects objective - audience - channel - funnel - cost - incrementality - learning.
- Never judge a campaign by one number. Reach without relevance, CTR without conversion, and ROAS without incrementality can all mislead.
- Use the funnel: awareness metrics show exposure, engagement metrics show interest, conversion metrics show action, and brand metrics show memory.
- The strongest metric is usually incremental business impact: what happened because of the campaign, compared with what would have happened anyway.
- For paid campaigns, always calculate break-even ROAS = 1 ÷ gross margin. If gross margin is 40%, break-even ROAS is 2.5x.
- A good teardown ends with decisions: scale, pause, change creative, change audience, change channel mix, or run a better experiment.
Big Picture: A Campaign Teardown Is a Business Logic Chain
A campaign is not “good” because it has high views, awards, or engagement. It is good when the metrics prove that the right audience moved closer to the business objective at an acceptable cost.
Core Explanation: How to Tear Down Any Campaign
Start with the campaign objective. A brand launch, app install push, festive sale, store footfall drive and reactivation campaign need different metrics. The mistake is to force every campaign into the same “reach, likes, sales” template.
Then map the campaign to the customer journey. A useful campaign dashboard should answer four questions:
The Metrics Dashboard: What to Track and What “Good” Means
Use the ranges below as interview guardrails, not universal laws. Category, creative, audience quality, seasonality and channel mix can change the benchmark dramatically.
A Small Worked Example: ROAS Is Not Enough
Assume a D2C brand spends ₹5,00,000 on a performance campaign. It generates ₹18,00,000 in attributed revenue and has a gross margin of 40%.
This is why good candidates say: “The dashboard says it worked; the experiment tells us how much it really worked.”
Definitions You Should Be Able to Say Cleanly
- Marketing - Philip Kotler: “Marketing is the science and art of exploring, creating, and delivering value to satisfy the needs of a target market at a profit.”
- Campaign metric: A quantified signal used to judge whether a campaign moved the target audience toward the intended business outcome.
- ROAS: Revenue attributed to advertising divided by advertising spend.
- CAC: Total acquisition cost divided by the number of new customers acquired.
- Incrementality: The additional outcome caused by the campaign compared with a credible no-campaign baseline.
Case Study: Cadbury Celebrations and “Not Just A Cadbury Ad”
Mondelez India turned a Diwali chocolate campaign into a hyperlocal advertising platform for small businesses, making campaign impact broader than chocolate sales alone.
Situation. Diwali is a crowded gifting season in India. Chocolate gift packs compete not only with other confectionery brands, but also with sweets, dry fruits, apparel, electronics and local retail promotions. During the pandemic recovery period, many small businesses also needed visibility.
The move. Cadbury Celebrations and Ogilvy India built “Not Just A Cadbury Ad,” a campaign that used technology to create localized versions of a celebrity-led festive ad. Instead of only saying “buy Cadbury,” the campaign allowed local stores to be featured in digital ad versions, turning a national brand campaign into a platform for neighbourhood businesses.
The result and lesson. The campaign earned major creative recognition, including Cannes Lions recognition, but its marketing lesson is bigger than awards. A full teardown would not stop at views or sentiment. It would track brand salience for Cadbury Celebrations, local merchant participation, engagement with personalized ads, festive purchase intent, earned media, and eventual sales or distribution lift.
So what: Cadbury did not win because of one factor. The primary driver was a brand idea with social usefulness. Supporting drivers were celebrity recognition, AI-led personalization, festive cultural relevance, digital shareability and the distribution strength of a large FMCG player.
How AI Changes Campaign Metrics Teardown
AI changes campaign analysis in three practical ways. First, it helps create more variants - copy, visuals, landing pages and audience segments. Second, it improves prediction - who is likely to convert, churn, repeat or respond to an offer. Third, it makes measurement faster by summarizing messy signals such as comments, reviews, call transcripts and competitor ads.
But AI does not remove the need for business judgment. In 2026, a good teardown should measure whether AI improved campaign economics, not merely whether AI produced more content.
Practical student workflow: Put the campaign film transcript, brand website, Meta Ad Library screenshots and public articles into NotebookLM. Ask it to produce: “funnel metrics, possible proxy metrics, missing data, likely control group design, and three interviewer questions.” Then use your judgment to separate evidence from assumption.
Interview Relevance
“Pick any real marketing campaign you admire. How would you evaluate whether it actually worked?”
Use this sentence in interviews: “I would separate optimization metrics from decision metrics - CTR helps improve the campaign, but incrementality tells me whether the campaign deserves more money.”
Common Mistake
The most common error is calling a campaign successful because it got high views, likes or awards. That costs candidates because it shows weak commercial thinking. Fix: always link the visible metric to a business outcome and ask, “What would have happened without the campaign?”
What to Revise Next
This is a natural capstone topic because it forces you to combine STP, consumer behavior, media planning, brand equity, digital metrics, unit economics and experimentation. Since this is the final lesson, revise by taking three campaigns - one FMCG, one app-led business and one B2B or financial-services campaign - and teardown each using the same structure: objective, audience, funnel, cost, incrementality and decision.