Google Smart Bidding is one of the most powerful tools in an Indian performance marketer's arsenal — and one of the most frequently misconfigured. The choice between Target ROAS (tROAS) and Target CPA (tCPA) is not merely a technical preference. In India's highly heterogeneous market, picking the wrong strategy costs real money every day the campaign runs.
This guide cuts through the noise with India-specific benchmarks, a clear decision framework and practical calibration guidance to help you get Smart Bidding working correctly in the Indian market context.
What Smart Bidding Actually Does (and Why India is Different)
Smart Bidding uses Google's machine learning to set bids at auction time based on signals: device, location, time of day, search query, audience lists, and dozens more. The algorithm optimises toward your declared objective — revenue (tROAS) or cost per conversion (tCPA).
India introduces several complicating factors that make calibration harder than in Western markets:
- Extreme price variance: Indian consumers search for the same product at wildly different price points. A "kurta" search could represent a ₹299 Meesho buy or a ₹4,500 FabIndia purchase. tROAS handles this better because it anchors on revenue, not a fixed cost target.
- Regional intent variation: Search behaviour in Tier 1 cities (Mumbai, Delhi, Bengaluru) converts at higher rates than Tier 2/3. A single national tCPA target will over-bid metro traffic and under-bid Tier 2 traffic relative to actual close rates.
- Festival seasonality: India has 35+ major festivals causing large conversion rate swings. Smart Bidding needs explicit seasonality adjustments or it undershoots during peaks and overshoots during troughs.
- COD and return rates: Indian e-commerce has 20–35% return rates for some categories (fashion: 25–30%, per Unicommerce 2024 report). If your conversion event is "purchase" without accounting for returns, you're optimising toward a partially false signal.
tROAS: The Right Choice for Indian D2C E-Commerce
How tROAS Works
Target ROAS tells Google to maximise conversion value while achieving your specified return on ad spend. If you set tROAS at 400%, Google aims to generate ₹4 in revenue for every ₹1 in ad spend. The algorithm bids higher for users it predicts will purchase high-value items and lower for users likely to buy low-value items.
When tROAS Outperforms in India
tROAS is the superior choice for Indian D2C brands when:
- Your product catalogue has a price spread of ₹500 or more between cheapest and most expensive SKUs
- You have 50+ conversions in the last 30 days (Google's minimum for reliable Smart Bidding)
- You're running Shopping campaigns where product values vary significantly
- Your average order value varies meaningfully by traffic source or audience segment
| Category | Conservative tROAS | Aggressive tROAS | Note |
|---|---|---|---|
| Fashion / Apparel | 300–350% | 400–500% | Account for 25% return rate |
| Beauty / Skincare | 350–400% | 500–600% | High repeat purchase value |
| Electronics | 400–500% | 600–700% | High AOV, lower return rate |
| Home Décor | 300–380% | 450–550% | Longer consideration cycle |
| Food / Wellness | 250–300% | 350–400% | Subscription uplift important |
Calibrating tROAS for India
Start with a tROAS target that is 20–30% lower than your actual ROAS from the previous 30 days. This gives Google's algorithm enough room to operate without artificially constraining volume. Tighten the target by 10–15% every 2 weeks once the campaign has accumulated 50+ conversions at the initial target.
tCPA: The Right Choice for Indian Lead Generation
When tCPA Works Better
Target CPA is better suited to campaigns where every conversion has roughly equal value — lead generation for real estate, education, financial services, B2B SaaS, and insurance. In these cases, the "revenue" from each lead isn't knowable at bid time, so tROAS can't optimise meaningfully.
tCPA is also appropriate for Indian e-commerce brands with a very narrow product catalogue (e.g., a single product at ₹999) where AOV variance is minimal.
| Industry | Avg tCPA Target | Lead-to-close rate | Customer LTV |
|---|---|---|---|
| Real Estate (₹50L+ property) | ₹1,200–2,500 | 2–5% | ₹2–5L commission |
| Education (online courses) | ₹300–800 | 8–15% | ₹8,000–25,000 |
| Insurance (term/health) | ₹400–1,200 | 5–10% | ₹15,000–60,000 LTV |
| B2B SaaS (SMB) | ₹800–2,000 | 10–20% | ₹30,000–1,20,000/yr |
| Home Renovation / Services | ₹250–600 | 15–25% | ₹25,000–80,000 |
The Regional Close-Rate Problem with tCPA in India
Here's a nuance that catches many Indian advertisers: your sales team likely closes leads from Delhi at 15% but leads from Tier 3 UP towns at 4%. If you run a single national tCPA campaign, Google optimises for lead volume at your target cost — not lead quality. You end up paying ₹500 for a Delhi lead and ₹500 for a Kanpur lead, even though the Delhi lead is 3.75x more valuable to your business.
The solution is campaign segmentation by city tier with differentiated tCPA targets: a Tier 1 (Mumbai, Delhi, Bengaluru, Hyderabad, Chennai, Pune, Kolkata) campaign at a higher tCPA reflecting higher close rates, and a Tier 2/3 campaign at a lower tCPA reflecting lower close rates.
The tROAS vs tCPA Decision Framework
| Business Type | Product/Price Variance | Conversion Type | Recommended Strategy |
|---|---|---|---|
| D2C E-commerce | High (₹500+ spread) | Purchase | tROAS |
| D2C E-commerce | Low (single SKU) | Purchase | tCPA |
| Lead Generation | N/A | Form fill / Call | tCPA (segmented by tier) |
| Marketplace Seller | Medium–High | Purchase / ROAS | tROAS |
| App Downloads | N/A | Install / in-app event | tCPA (tCPI) |
| SaaS / Subscription | Tiered plans | Trial / Demo | tCPA (optimise to qualified trial) |
Seasonality Adjustments: The India-Specific Must-Do
India's festival calendar creates conversion rate spikes that Smart Bidding cannot anticipate without help. Google's algorithm looks back 6–8 weeks to estimate future conversion rates — but Diwali happens once a year. If you don't manually apply a seasonality adjustment, your tROAS or tCPA campaigns will significantly under-bid during the first 2–3 days of the festival (when conversion rates spike) and then over-bid in the following week as the algorithm tries to compensate.
Apply a +30 to +50% seasonality adjustment in Google Ads 24 hours before major Indian festivals (Diwali, Holi, Navratri start, Eid, Christmas/New Year). Remove it 48 hours after the festival peak. According to Google's own documentation, seasonality adjustments work best for short-duration events (1–7 days) — which matches India's festival patterns perfectly.
AdsSarthi's Google Ads automation layer handles seasonality adjustments automatically for all 35 tracked Indian festivals — the system applies and removes adjustments based on the festival calendar without manual intervention. See pricing plans that include Google automation.
Common Smart Bidding Mistakes in India
- Switching strategies mid-learning phase: Smart Bidding needs 2–4 weeks to exit the learning phase. Switching from tCPA to tROAS (or vice versa) resets the learning phase — avoid strategy changes more than once every 6 weeks.
- Setting unachievable targets: If your actual ROAS is 280% and you set tROAS at 600%, Google will drastically reduce volume to hit the target, often destroying campaign performance. Set targets within 20–30% of historical performance.
- Not excluding branded terms from Smart Bidding ROI: Branded search terms convert at 2–5x non-brand rates. Including them in a tCPA/tROAS campaign skews the algorithm toward branded queries at the expense of new customer acquisition.
- Ignoring impression share loss to budget: If you're losing 40%+ of impressions due to budget, Smart Bidding cannot optimise effectively. Fix budget constraints before worrying about bid strategy.
Get a free audit to see exactly which Smart Bidding configuration issues are costing your campaigns money right now.