Target CPA vs Target ROAS: The One Question That Decides

One question about your conversions, then one check on your data, decides the bid strategy.

Most tCPA vs tROAS debates skip the only question that matters.

Does a ₹500 order mean the same to you as a ₹5,000 order? If yes, you are buying conversions, and Target CPA is the tool. If no, you are buying value, and Target ROAS is the tool, provided Google can see that value.

This post covers that one decision and how to set the target. For every other strategy (manual CPC, Maximize clicks, Target impression share), our guide to Google Ads bidding strategies has the full map, and if you landed here looking for CPA as an affiliate pricing model, that is a different subject covered in our CPA marketing explainer.

What changed in Google Ads in 2026?

The names got shorter. Google's own Help pages say that starting June 2026, "Maximize conversions with a Target CPA" is being relabelled "Target CPA" and "Maximize conversion value with a Target ROAS" is being relabelled "Target ROAS" (Google Ads Help, About Target ROAS bidding).

So if an older screenshot or a colleague says "Maximize conversions with a target CPA", they mean Target CPA. The mechanics did not change with the label. Both remain Smart Bidding strategies that set a bid for each auction.

Both also work as a standard strategy on one campaign or as a portfolio strategy shared across several (Google Ads Help, About Target CPA bidding). More on when a portfolio helps further down.

How does Google treat your "target"?

As an average, not a ceiling. This is the part people misread most often.

Google says plainly that with Target CPA "some conversions may cost more than your target and some may cost less", and that the goal is to keep the overall cost per conversion at the target (About Target CPA bidding). So a ₹600 target that produces one ₹1,400 lead on a Tuesday is working as designed.

Target ROAS behaves the same way on value. Google describes the reported average target ROAS as "the traffic-weighted average ROAS that your bid strategy optimized for" (About Target ROAS bidding).

Aggressive targets carry a cost, and Google warns about it on both pages. Set tCPA too low and "you may forgo clicks that could result in conversions, resulting in fewer total conversions." Set tROAS too high and it "may limit the amount of traffic your ads may get." The bidder protects your ratio by sitting out auctions. Your budget stays unspent and the ratio looks lovely on shrinking volume.

When Target CPA is the right call

Pick tCPA when you would pay the same for any conversion in the campaign. A clinic booking form, a coaching institute's counselling request, a B2B demo request where your sales team closes leads at similar rates: those fit.

Google's value-based bidding page frames it the same way: Target CPA suits advertisers who "care more about driving conversion volume" (About Smart Bidding using value-based bidding).

The trap is counting the wrong thing. If "conversion" includes page views, PDF downloads and form submits in one bucket, tCPA will happily buy you the cheapest of them. Fix the primary conversion action first; our conversion tracking guide covers primary versus secondary actions.

Google also says you can start Target CPA "with no conversion history" and recommends judging it on the last 30 days with at least 30 conversions (About Target CPA bidding). Starting is easy. Reading the result on 8 conversions a month is where people fool themselves.

When Target ROAS is the right call

Pick tROAS when conversions differ in value and you can report that value. Ecommerce is the obvious case: a ₹900 order and a ₹6,000 order should not be bid for at the same price.

Lead gen can qualify too. Google explicitly allows "real economic values, like revenue or proxy values, like a lead score" and requires two or more unique values reported to the conversion goal (value-based bidding for Search and Shopping). A real-estate developer that values a 3BHK site-visit lead above a 1BHK one fits here, once those values actually flow back into Google Ads.

The volume bar is published. For Search and Shopping, Target ROAS needs at least 15 conversions in the past 30 days; for Display, 15 conversions with valid values; app campaigns need 10 a day or 300 in 30 days (About Target ROAS bidding). Conversions with a value of zero do not count toward that.

Our view: if you are uploading values from a CRM, get the values stable before you switch. Google's own best-practice page asks for at least 4 weeks or 3 conversion cycles of consistent uploads first, and recommends daily uploads (Value-based Bidding Best Practices).

Sources: Google Ads Help pages for Target CPA, Target ROAS and value-based bidding, checked September 2026.

Worked example: a Pune D2C brand and its breakeven ROAS

Everything in this section is hypothetical, built to show the arithmetic. Swap in your own numbers.

A Pune D2C brand sells activewear. Average order value is ₹1,800 and gross margin is 40%, so each order leaves ₹720 before ad cost.

Breakeven ROAS is revenue ÷ margin: ₹1,800 ÷ ₹720 = 2.5, or 250%. Below 250%, every order loses money on ad cost alone. Our ROAS calculator does this sum for you, and our guide on what counts as a good ROAS explains why 250% is a floor and not a goal (shipping, returns and COD return-to-origin eat into that ₹720 too).

Last 30 days on Maximize conversion value: ₹1,00,000 spend, 180 orders, ₹3,24,000 revenue. That is a ROAS of 324%. Gross margin is 180 × ₹720 = ₹1,29,600, so the campaign leaves ₹29,600 after ad cost.

The founder reads a blog that says "aim for 5x" and sets Target ROAS at 500%.

Here is one way that can play out. The bidder drops out of every auction where it predicts less than ₹5 back per ₹1. Suppose spend falls to ₹20,000 and orders to 56. Revenue is ₹1,00,800, ROAS reads 504%, and the dashboard looks like a win.

Now the margin: 56 × ₹720 = ₹40,320, minus ₹20,000 = ₹20,320. Profit fell by ₹9,280 while ROAS "improved" by 180 points. The exact drop in your account will differ; the direction is the point Google itself warns about.

The sensible version: start the target near the 324% the campaign already earns, say 320%, and move up in small steps while watching profit in rupees. A tROAS between breakeven (250%) and your current actual is where the useful decisions live. Above your current actual, you are asking for volume to shrink.

The same brand also runs a bulk-order enquiry form for gyms and corporate gifting. Each enquiry is worth about the same to the sales team, so that campaign runs on Target CPA. One business, two strategies, same question deciding both.

Learning periods, and what Google says about changing targets

Google's learning-period page says a new or changed bid strategy "can take up to around 50 conversion events or 3 conversion cycles" to calibrate (Duration of the learning period). Learning status is triggered by a new strategy, a changed setting, or adding and removing campaigns, ad groups or keywords from the strategy.

Changing the target is different. Google's page on how its bidding algorithms learn says "changing a target won't trigger a 'learning' status" and that you can change targets "by as large a magnitude as you would like" (How our bidding algorithms learn). The Target ROAS page adds that the bidder reacts immediately but needs 1–2 conversion cycles to hit a new target.

We still move targets in steps, and that is opinion, not a Google rule. The reason is diagnosis. If you jump from 320% to 500% and volume falls, you cannot tell how much fall came from the target versus seasonality or a competitor. Small steps give you a readable cause and effect.

Standard or portfolio strategy?

A standard strategy sits on one campaign. A portfolio strategy groups several campaigns, ad groups or keywords under one shared target (Portfolio bid strategy definition).

In our experience, portfolios help when individual campaigns are too thin to read on their own, for example five city campaigns doing 6–8 conversions each a month, and the business target is the same across them. Keep campaigns on separate strategies when their economics differ; a portfolio averages a high-margin category with a low-margin one and hides the loss-maker.

Performance Max has its own quirks here, which we cover in the Performance Max guide.

How to switch from Maximize conversions to a target safely

This is the sequence we use on managed accounts. It applies to Maximize conversion value to Target ROAS in the same way.

1. Take a 30-day baseline. Write down conversions, CPA or ROAS, spend and conversion lag for the last 30 days, excluding the most recent few days if conversions arrive late. This is the number you will judge against, not the number you wish you had.
2. Check the floor. Work out your breakeven CPA (margin per order or value per lead) or breakeven ROAS. If the current actual is already below breakeven, a target will not rescue it; fix the offer, landing page or tracking first.
3. Set the first target at or near actual. Google's recommended tCPA is based on the last 30 days' average CPA adjusted for conversion delay. Start there, or within about 10% of your baseline (our rule of thumb). Do not start at your dream number.
4. Leave it alone for the learning window. Give it up to around 50 conversions or 3 conversion cycles. Avoid adding keywords, changing conversion actions or moving campaigns in and out of the portfolio during this time.
5. Tighten in steps. Move the target by a modest amount (we use 10–15%), then wait at least 1–2 conversion cycles before the next move. One change at a time.
6. Watch four numbers weekly. Conversions or conversion value against baseline, actual CPA or ROAS against target, spend against budget (steady underspend usually means the target is too tight), and profit in rupees. If volume drops faster than efficiency improves, step back.

Want a second pair of eyes before you flip the switch? Our Google Ads management team runs this exact baseline-and-steps process, including the breakeven maths and the conversion-value plumbing that tROAS depends on. Management fees for performance work typically run ₹25,000–₹1,00,000 a month in the Indian market, and ad spend is always separate.

Mistakes we see most often

• Running Target ROAS with a static conversion value, so every order reports ₹1. That is tCPA with extra steps and a confusing report.
• Setting tCPA from a competitor's number or a blog benchmark instead of your own 30-day actual.
• Judging a strategy change after four days when the product takes a week from click to purchase.
• Treating a rising ROAS as good news without checking whether revenue and profit fell alongside it.
• Counting a lead as the finish line when sales can tell you which leads close. If you know cost per customer, our CAC calculator will show whether your CPA target holds up.