Marketing Analytics: Answer Channels, Attribution and ROAS with Confidence
A brand can spend lakhs on a creator video, see Google Search conversions jump the next day, and wrongly declare search the hero. That is the central puzzle of marketing analytics: the channel that gets the click is not always the channel that created the customer.
- Marketing channels are routes through which a brand reaches, converts and retains customers - paid, owned, earned, offline and partner-led.
- Attribution assigns conversion credit to marketing touchpoints; it is useful, but it is not the same as causality.
- ROAS = revenue from ads ÷ ad spend; it looks good only if it clears margin, CAC and incrementality tests.
- Last-click attribution over-rewards demand-capture channels like search and marketplaces, and under-rewards demand-creation channels like video, influencers and outdoor.
- Incrementality asks the hard question: “What sales would not have happened without this spend?”
- Best interview answer: map channel role, choose attribution model, test incrementality, calculate unit economics, then recommend budget shifts.
- One-line fix: never say “ROAS is 4x, so scale it” before checking gross margin and incremental lift.
Think of marketing analytics as a decision engine, not a reporting dashboard. The job is to connect where the money went, which customer actions happened, and what business value was truly created.
Core Explanation: Channels, Attribution and Return on Spend
Marketing channels are the routes through which a brand reaches and serves customers. In analytics, do not just name the channel; define its role in the journey.
Attribution answers: “Which touchpoint gets credit for a conversion?” Incrementality answers: “Would this conversion have happened anyway?” These are related, but not the same.
Attribution Models You Must Know
Attribution models are rules for assigning credit. Your answer improves when you explain when each model is useful and where it misleads.
A D2C electronics brand selling on Amazon India may see marketplace search ads deliver high ROAS during a sale. But some of that demand may have been created earlier through YouTube reviews, Instagram creators and price-drop reminders. The strategic so what: marketplace ads often capture high intent, while creators and video may create the intent that later appears as search demand.
Return on Spend: The Metrics That Actually Matter
ROAS is useful, but incomplete. A channel can show high ROAS and still destroy value if margins are thin, discounts are heavy, or conversions are not incremental.
Worked Example: Why 4x ROAS May Still Be Weak
Suppose a brand spends ₹10 lakh on performance ads. The platform reports ₹40 lakh attributed revenue. Reported ROAS = ₹40 lakh ÷ ₹10 lakh = 4x.
Now add two business realities:
- Gross margin is 40%, so gross profit on attributed revenue = ₹40 lakh × 40% = ₹16 lakh.
- A holdout test shows only 60% of reported revenue was incremental. Incremental revenue = ₹40 lakh × 60% = ₹24 lakh.
Incremental gross profit = ₹24 lakh × 40% = ₹9.6 lakh. Against ₹10 lakh spend, the campaign is slightly negative before considering repeat purchases. The interview-grade conclusion is: reported ROAS looks good, but incremental profit is not yet attractive unless repeat purchase LTV makes up the gap.
Definitions You Can Say in One Breath
- Marketing channel: Kotler and Keller define it as “sets of interdependent organizations involved in the process of making a product or service available for use or consumption.”
- Marketing attribution: The rule or model used to assign conversion credit to customer touchpoints before purchase.
- ROAS: Advertising revenue attributed to a campaign divided by advertising spend on that campaign.
- Incrementality: The additional outcome caused by marketing activity compared with what would have happened without it.
- CAC: The cost of acquiring one new customer, including relevant sales and marketing spend.
Case Study: Lenskart and Omnichannel Marketing Measurement
Lenskart shows why modern marketing analytics must connect digital ads, app journeys, store visits and repeat purchases instead of judging each channel in isolation.
Situation: Eyewear is a considered purchase. A customer may discover frames through social media, compare prices online, book an eye test, visit a store, and finally purchase after trying frames physically. If Lenskart judged only the last touchpoint, store sales could look like “offline success” even when digital media created the visit.
The move: The measurement logic for such a business is not just platform ROAS. It needs a connected view of campaign exposure, website or app behavior, store locator usage, eye-test bookings, store visits, purchases and repeat orders. The primary driver is omnichannel journey integration; supporting drivers include first-party customer data, CRM follow-ups, localized store catchments and product assortment visibility across online and offline touchpoints.
Outcome or lesson: The enduring lesson is not “online beats offline” or “stores beat apps.” For omnichannel brands, analytics must measure how channels work together. The strongest answer names the primary driver - connected customer journeys - and the supporting drivers - first-party data, store network, CRM and localized demand generation.
How AI Changes Marketing Analytics: Channels, Attribution and Return on Spend
AI is making marketing analytics faster, but it does not remove the need for judgment. In 2026, three changes matter most:
Use NotebookLM before an interview: upload the company annual report, recent campaign articles and this lesson, then ask, “What are five likely questions on this company's channel mix, attribution risks and ROAS?” Cross-check any numbers manually before using them.
Interview Relevance
“A campaign shows 5x ROAS on Google Search, while YouTube shows only 1.2x ROAS. Would you shift budget from YouTube to Search?”
Use the phrase “channel role before channel ROI”. It signals that you understand full-funnel marketing, not just dashboard reading.
Common Mistake
The costly mistake is treating attributed ROAS as true business impact. It costs candidates because interviewers know platform dashboards can double-count conversions, reward last clicks and ignore margin. One-line fix: always convert ROAS into incremental gross profit before recommending scale.
What to Revise Next
Once channel measurement is clear, move deeper into the customer journey. Revise these next so you can connect marketing spend to user behavior and product value: