A £20,000 monthly Google Ads budget can look healthy in a platform report while quietly eroding margin. Revenue is up, clicks are cheap and conversion volume is growing, yet the blended cost of sale is beyond what the business can afford. That is why the best bidding strategies ecommerce brands use are not chosen because Google recommends them. They are chosen because they support a clear profit target.
For established retailers, bidding is not a switch to flick once and forget. It is the mechanism that decides how aggressively you compete for demand, which products receive budget and how quickly an account can scale without paying for unprofitable revenue. The right choice depends on your conversion data, margins, product mix, average order value and stock position.
Start with the number that actually matters
Before selecting a Smart Bidding strategy, establish your breakeven point. For many ecommerce brands, that means a maximum cost of sale or minimum return on ad spend after product costs, fulfilment, payment fees, returns and any contribution required towards overheads.
A 400% ROAS target is not automatically good. If your gross margin is 25%, it may still be too low once VAT, delivery and returns are included. Equally, a target that is too conservative can restrict volume and leave profitable revenue on the table. The target needs to reflect commercial reality, not an arbitrary benchmark from another account.
This is also where weak account management shows. Optimising solely for platform ROAS can favour repeat customers, branded searches and low-margin products. A serious bidding strategy considers whether the business is acquiring valuable new customers, clearing stock, protecting margin or maximising profitable revenue from proven demand.
The best bidding strategies ecommerce brands should consider
Maximise conversion value without a target
Maximise conversion value tells Google to pursue the greatest possible revenue value within the available budget. It is often the right starting point for mature Shopping and Performance Max campaigns with reliable purchase tracking but insufficient stable data to impose a sensible ROAS target.
The advantage is flexibility. Google can bid harder when it sees a high-value conversion opportunity rather than being constrained by an early target based on incomplete data. The trade-off is control. Without a target, spend can become more aggressive and efficiency may fluctuate, particularly during promotional periods or when product availability changes.
Use it when you need to establish a performance baseline, when a campaign has recently been restructured, or when the account has strong conversion volume but a wide range of order values. Review profitability closely rather than assuming rising conversion value means rising profit.
Maximise conversion value with a target ROAS
Target ROAS is the workhorse strategy for many established ecommerce advertisers. It gives Google a commercial guardrail: pursue conversion value, but aim to generate a defined return for every pound spent.
It works best when the campaign has clean purchase value tracking, consistent conversion volume and a target rooted in margin. For example, a brand that needs at least £5 in revenue for every £1 in media spend can begin around a 500% target ROAS, subject to its actual account history and new customer objectives.
The common mistake is setting the target too high. A 700% target may look attractive on a dashboard, but it can choke delivery, reduce impression share and prevent the campaign from reaching buyers who would still have been profitable. Set targets in stages. If performance is stable, tighten or loosen the target gradually and allow enough time for the system to respond.
Target ROAS is particularly effective when campaigns are segmented intelligently. High-margin bestsellers, low-margin accessories, clearance products and seasonal lines should not necessarily compete under one blanket target. Fluid campaign structures give each product group a more appropriate bidding environment.
Maximise conversions with a target CPA
Target CPA focuses on the number of conversions rather than their revenue value. For a standard ecommerce purchase campaign, it is usually less commercially precise than target ROAS because a £25 order and a £250 order are treated as the same outcome.
There are exceptions. It can make sense for lead generation connected to ecommerce, subscription sign-ups, free sample requests or tightly priced products with a stable basket value. It may also have a role where the primary goal is acquiring a new customer at a known allowable cost.
If you use target CPA for product sales, do so because the numbers support it, not because it is easier to understand. A cost per acquisition target without order-value context can create volume that looks efficient while lowering total profitability.
Manual CPC and enhanced CPC
Manual bidding has a limited but legitimate role. It can offer granular control during short tests, in low-data campaigns or where automation has no meaningful conversion signals to learn from. Enhanced CPC can add a layer of automated adjustment, although it is not a substitute for a properly configured value-based strategy.
For most established ecommerce accounts, manual CPC is not a long-term scaling answer. It is labour-intensive and cannot process auction-time signals at the scale Google’s automated systems can. Brands often cling to it after a poor Smart Bidding experience, when the real issue was broken tracking, blended campaigns, poor feed quality or unrealistic targets.
Automation is not the problem. Feeding it poor data and expecting profitable decisions is.
Match the bid strategy to campaign type
Google Shopping and Performance Max usually benefit from value-based bidding because product prices and basket values vary. Start with maximise conversion value where needed, then move to target ROAS once performance and data quality justify a constraint.
Search campaigns require more judgement. Brand search can often run efficiently with a lower level of intervention, but it should be isolated from non-brand activity so it does not inflate account-level performance. Non-brand search may need a different ROAS target because it is doing more acquisition work and naturally converts at a lower rate.
For Facebook and Instagram campaigns, the principle remains the same even though the bidding options differ. Optimise towards purchase value where the platform has sufficient signal, distinguish prospecting from retargeting, and assess performance against blended revenue and contribution margin. Retargeting is rarely proof that cold acquisition is working.
Your product feed affects what bidding can achieve
Bidding cannot repair a poor Merchant Centre feed. If titles are vague, product types are inconsistent, attributes are missing or disapproved products are left unresolved, Google has less context to match products with relevant shopping queries.
Feed optimisation and bidding strategy work together. Accurate titles, strong imagery, correct GTINs, clean pricing and custom labels make segmentation possible. Custom labels can separate bestsellers, high-margin ranges, seasonal stock, clearance products or products with poor return rates, allowing bids and ROAS targets to reflect actual commercial value.
This is where generic campaign setups fall apart. A single Performance Max campaign containing every SKU may be easy to launch, but it limits control over budget allocation. It can push spend towards products with high revenue but weak margin, or towards familiar products while ignoring a category with stronger growth potential.
Avoid the changes that reset learning unnecessarily
Smart Bidding needs stable signals, but “leave it alone” is not a strategy. The practical approach is controlled change. Avoid rewriting budgets, targets, assets and campaign structures all at once, then blaming the algorithm when results become unstable.
When testing a target ROAS adjustment, change one meaningful variable and judge it against enough data. The right timeframe depends on spend and conversion volume, but daily reactions to normal volatility are nearly always counterproductive. Look beyond the seven-day platform window too. Consider return patterns, lagging conversions, stockouts and promotional effects.
Watch for the warning signs: budget-limited campaigns with an overly strict target, falling spend after a target increase, revenue growth that does not translate into contribution, or Performance Max reporting driven largely by branded demand. These are management problems, not simply bidding problems.
Measure profit, not just attributed revenue
Google Ads reports are useful, but they are not your profit and loss statement. Use them alongside ecommerce platform data, analytics and finance reporting. Track blended MER where appropriate, new versus returning customer performance, product-level margin, refund rates and the difference between reported revenue and net revenue.
At Oxedent, this is the standard expected of serious ecommerce PPC management: bidding decisions should reduce waste and support scalable profit, not make a monthly report look prettier. If conversion tracking is inaccurate or margin data is unknown, fix that first. No bid strategy can compensate for missing commercial inputs.
The most effective next move is usually not a wholesale switch to the latest automated option. Audit the target, product segmentation, feed quality and measurement behind the current strategy. Once those foundations are sound, bidding becomes what it should be: a disciplined way to buy more of the revenue your business can actually afford to scale.