A bidding strategy can make a profitable product catalogue look unscalable, or turn a healthy budget into expensive revenue with no margin left. The best Google Ads bidding strategies are not the ones Google recommends by default. They are the ones that reflect your real contribution margin, stock position, conversion data and appetite for growth.
For established eCommerce brands, bidding should never be a set-and-forget platform setting. It is a commercial control. If your account is optimising towards inflated revenue, duplicated conversions or last-click signals that ignore returning customers, automated bidding will scale precisely the wrong outcome.
Start with the number that matters: profitable revenue
Google Ads can bid towards clicks, conversions, conversion value, impression share or a target return. Only one of those is usually aligned with a scaling eCommerce business: profitable conversion value.
That does not mean every campaign should run on Target ROAS from day one. It means every bidding decision should begin with a clear answer to one question: what can you afford to pay to acquire a sale while still making money?
Your break-even ROAS is the foundation. A brand with a 60% gross margin, 10% fulfilment and payment costs, and 15% operating costs has very different bidding headroom from a brand with 25% gross margin and frequent discounting. Applying a blanket 500% ROAS target across every product, campaign and season is lazy management, not strategy.
Build targets around contribution margin where possible. Factor in product cost, shipping subsidies, transaction fees, returns, discount depth and whether the order comes from a new or returning customer. Revenue is useful, but it is not profit.
The best Google Ads bidding strategies by account maturity
Maximise conversion value: the default for controlled growth
For most eCommerce accounts with accurate purchase values and enough conversion volume, Maximise Conversion Value is the strongest starting point. It gives Google room to find higher-value orders without immediately restricting delivery with an ambitious ROAS target.
This is particularly useful when a campaign is under-spending, when you have recently rebuilt a Performance Max or Shopping structure, or when seasonality has changed demand. The algorithm can explore more auctions and identify valuable pockets of traffic that a tight target may suppress.
There is a trade-off. Without a target, spend can rise faster than efficiency. Use this approach where you have a defined testing budget and the margin to tolerate short-term fluctuation. Review performance against blended business metrics, not a single day of platform ROAS.
Target ROAS: best for mature, stable campaigns
Target ROAS is usually the right choice once a campaign has reliable conversion tracking, sufficient purchase volume and a consistent pattern of demand. It tells Google the return you want while allowing the system to bid more aggressively when it predicts a valuable sale.
The common mistake is setting the target too high. A 900% target might look attractive in a report, but it can restrict reach, reduce revenue and hand market share to competitors. If your true break-even point is 350%, insisting on 900% may protect a vanity number while capping profitable growth.
Set the target based on what the business needs, then adjust gradually. Major target changes can destabilise delivery. A sensible move is often 5% to 15% at a time, followed by enough time for the campaign to gather fresh data. The exact window depends on volume, but reacting to two weak days is not optimisation.
Target ROAS works best when campaign groups have similar economics. Combining high-margin accessories, low-margin hero products and clearance stock in one asset group or Shopping campaign gives the algorithm conflicting instructions. Fluid campaign structures and clean product segmentation make bidding far more predictable.
Maximise conversions with Target CPA: useful, but selective
For direct-to-consumer retail, Target CPA is rarely the primary bidding strategy for purchase campaigns. It treats every conversion as broadly equal, whether the order is worth £20 or £400. That can push spend towards easy, low-value sales rather than commercially valuable customers.
It has a place when the conversion itself has a fixed value or a tightly defined commercial role. Think newsletter sign-ups with a proven downstream value, sample requests, finance applications or a campaign designed solely to generate first orders below a fixed acquisition cost.
If you use Target CPA for new-customer acquisition, separate that objective from standard revenue campaigns. Otherwise, Google may chase cheap customers at the expense of total revenue and returning-customer profitability.
Manual CPC and Maximise Clicks: rarely a scaling answer
Manual CPC still has limited uses: short diagnostic tests, brand protection, low-volume launch activity or situations where conversion tracking is temporarily unreliable. It gives more direct bid control, but it cannot process auction-time signals at the scale Google’s automated systems can.
Maximise Clicks is even more limited for eCommerce. Cheap traffic is not the same as profitable traffic. Unless the goal is a very specific data-gathering exercise, bidding for clicks invites low-intent visits, weak engagement and wasted budget.
If your account only performs when you manually control every bid, the issue is often not automation. It is usually poor feed data, fragmented conversion tracking, mixed product economics or insufficient volume for the structure in place.
Tracking decides whether smart bidding is actually smart
Google’s bidding system only knows what you send it. If purchase values are wrong, tax is inconsistently included, refunds are ignored, or several tags record the same transaction, a sophisticated bid strategy becomes a fast way to scale bad data.
For retail brands, purchase conversion tracking should be the primary optimisation signal. Check that transaction IDs deduplicate sales, values reflect the amount you actually intend to optimise towards, and consent configuration does not create unexplained reporting gaps. Enhanced conversions can improve signal quality, but it does not repair a broken tracking implementation.
Also look beyond platform attribution. Google Ads may report a strong ROAS while total paid revenue stalls, branded search absorbs demand created elsewhere, or returning customers account for the apparent efficiency. Compare Google’s reported results with your store data, blended MER, new-customer rate and actual profit after advertising.
Feed quality changes bidding outcomes
Bidding cannot rescue a weak Shopping feed. Google needs clear product titles, accurate GTINs, relevant categories, competitive pricing and dependable availability data to match products to high-intent searches.
Feed optimisation is not a cosmetic task. It determines the queries your products can enter, the relevance Google assigns and the conversion data your bidding model receives. A generic title such as “Blue Trainers” gives the platform far less to work with than a specific, commercially useful title containing brand, model, colour, material and key product attribute where appropriate.
When margins differ materially, use custom labels to reflect them. You may segment by margin band, best seller status, seasonal priority, price point or clearance stock. This lets you apply targets that match commercial reality rather than forcing every SKU into the same ROAS requirement.
How to set a target without strangling growth
A target is not a wish. It is a constraint placed on the auction system. Before increasing a Target ROAS, ask whether the campaign is already limited by budget, whether it is missing profitable impressions and whether the marginal return still clears your profitability threshold.
If a campaign is spending its full budget at 450% ROAS and your profitable floor is 350%, raising the target to 600% may reduce volume without improving the business. A better decision may be to increase budget, maintain the target, and monitor whether marginal performance remains acceptable.
The opposite is also true. If a campaign is spending heavily at 280% ROAS against a 400% break-even requirement, lowering the target to chase scale would be commercially reckless. Cut waste first. Review search terms where available, product performance, feed quality, exclusions, pricing competitiveness and landing-page friction.
Treat Performance Max as a profit channel, not a black box
Performance Max can be highly effective for eCommerce, particularly when supported by a strong feed and clean conversion data. But it needs commercial guardrails. Separate products with distinct profitability profiles, exclude irrelevant URLs where needed, provide meaningful creative assets and keep brand performance under scrutiny.
Do not judge it only by the number shown in the Google Ads interface. Assess incremental revenue, product mix, new-customer contribution and whether branded demand is inflating the result. The right bid target is the one that produces profitable growth after those factors are considered.
For serious retailers, the winning approach is disciplined rather than dramatic: accurate data, campaign structures built around economics, targets adjusted with intent, and a willingness to remove spend that cannot justify itself. At Oxedent, that is the standard worth holding every bid strategy to. If a bidding setting cannot explain how it protects margin while creating room to scale, it has not earned more budget.