For most eCommerce in 2026, automated (Smart Bidding) wins: it adjusts bids per auction using signals no human can act on fast enough. Manual CPC is now a niche tool. The real choice is which automated strategy: use Maximise conversion value + a target ROAS once you have enough conversions and clean revenue values, and Maximise conversions (or no target) while data is still thin. The catch: Smart Bidding is only as good as the conversion data you feed it.
“Should I bid manually or let Google do it?” used to be a real debate. In 2026 it mostly is not, because Smart Bidding sees things you cannot: the user's device, time, location, browser, query context and much more, and prices each auction accordingly, in real time. But automated is not one thing, and picking the wrong strategy (or the right one too early) can still hurt. Here is how to choose.
Manual vs automated, in one paragraph
Manual CPC means you set the maximum you will pay per click and adjust it yourself. It gives total control and zero intelligence, you cannot react to the hundreds of signals that decide whether a given click is worth 20 cents or two euros. Automated (Smart Bidding) hands that per-auction decision to Google, which optimises toward a goal you set (conversions, or conversion value). For a store with real conversion data, that trade, giving up manual control for machine speed and signal, is almost always worth it.
The Smart Bidding menu
| Strategy | Optimises for | Best for |
|---|---|---|
| Maximise conversions | The most conversions for the budget | Early on, or when every order is worth roughly the same. |
| Maximise conversions + target CPA | Conversions at a set cost each | Lead-gen, or stores wanting a fixed cost per sale. |
| Maximise conversion value | The most total revenue | Most eCommerce, where order values vary. |
| Maximise conversion value + target ROAS | Revenue at a set return (e.g. 4x) | eCommerce with enough data and clean revenue values. |
For a typical online store, the destination is usually Maximise conversion value with a target ROAS, because it optimises for money, not just order count. The question is when to get there.
Which to use, and when
Match the strategy to how much data you have and what you are optimising for:
- New campaign, thin data: start on Maximise conversions or Maximise conversion value with no target. Let it gather conversions and learn before you constrain it.
- Enough conversions, values flowing: switch to Maximise conversion value and introduce a target ROAS, set just above your break-even ROAS.
- Orders all similar in value: a target CPA can be simpler and just as good.
- Tiny or very unusual account: the rare case where manual control still earns its keep, but for most stores this is not you.
The prerequisite that makes or breaks it
Smart Bidding predicts the value of each auction from your conversion data. If that data is wrong, the bidding optimises confidently toward the wrong thing, and you will not easily see it happening.
The mistakes that undo good bidding
- Target set too aggressively. Too high a ROAS (or too low a CPA) target starves the campaign, it stops bidding on all but the safest conversions and volume collapses.
- Changing the target too often. Big or frequent target changes reset the learning period. Adjust in small steps and give it time to settle.
- Too few conversions per campaign. Fragmenting the account leaves each campaign without enough data to bid well, a core reason to consolidate rather than fragment.
Not sure whether your bidding strategy fits your data, or whether an aggressive target is quietly starving your account? A free audit checks your bidding, your conversion setup and your targets, and tells you what to change.
