Strategic Transparency

We don't just show results.
We show the decisions.

Any agency can show you a ROAS number. What separates strategy from button-pushing is the ability to explain why you chose one path over another - including the path you rejected.

These are real trade-offs from real client accounts. Anonymised, but authentic. Each one shows the options we weighed, the reasoning we applied, and the commercial principle that guided the decision.

See Our Thinking Applied to Your Account Decisions We Won't Automate

The Trade-Offs

Four decisions that define strategic PPC management

Trade-Off 1

We put £0 on the best-selling SKU

A fashion brand spending £35k/month. Their hero product - a £79 dress - generated 42% of all Google Ads revenue. The agency before us had it as the centrepiece of every campaign.

Keep scaling the hero

Rationale

It converts at 4.2%. It has brand recognition. Customers search for it by name. ROAS is 5.8x. By every platform metric, it's the best product in the account.

Projected Outcome

Revenue grows 15%. Contribution margin stays flat. The dress carries 14% margin after 32% returns. At 5.8x ROAS on 14% effective margin, every sale generates £0.41 profit. The £14.7k spent on this product generates £602 profit.

Redirect budget to mid-range accessories

Rationale

Accessories (£25-£45) carry 52% margin with 8% return rates. They convert at 2.1% - lower than the dress - but each conversion generates £12-£18 profit vs £0.41.

Projected Outcome

Revenue drops 18%. Contribution margin increases 34%. The £14.7k reallocated to accessories generates £4,890 profit. The hero dress still sells organically - it just doesn't get paid advertising.

Our Decision

We chose Option B. The hero dress continued selling at 85% of previous volume through organic and direct traffic. It didn't need advertising - customers already knew it existed. The budget was better deployed acquiring new customers through products that actually generate margin.

The Principle

Revenue attribution ≠ revenue dependency. Just because Google Ads claims credit for a sale doesn't mean the sale required Google Ads.

Trade-Off 2

We cut spend 30% while ROAS was above target

A supplements brand at £22k/month, targeting 3.5x ROAS. They'd been at 4.1x for three consecutive months. The founder wanted to scale to £35k/month while ROAS was 'strong.'

Scale to £35k while ROAS is healthy

Rationale

Performance is above target. The algorithm has learned. Momentum is good. Standard agency advice: scale when you're winning.

Projected Outcome

We modelled it. At £35k, projected ROAS drops to 2.8x based on the diminishing returns curve we'd plotted. At 35% margin, break-even is 2.9x. The extra £13k spend would generate approximately −£1,200 in profit. You'd be paying to lose money.

Hold at £22k and optimise for margin, not volume

Rationale

Current spend is near the optimal profit point on the diminishing returns curve. Instead of pushing more money through the same funnel, reallocate within the existing budget toward higher-margin products.

Projected Outcome

Total revenue stays flat. POAS improves from 1.12x to 1.41x. Monthly profit increases from £2,464 to £3,102 - a 26% lift without spending a penny more.

Our Decision

We chose Option B and then went further - we actually reduced spend by 12% (to £19.4k) because the bottom £2.6k of spend was generating marginal ROAS of 1.8x (below break-even). The result: less spend, less revenue, more profit.

The Principle

Scaling isn't always the right move. The optimal spend level is where marginal profit hits zero - not where total ROAS hits target.

Trade-Off 3

We deliberately ran one product at a loss

A pet food brand with a £12.99 starter pack and a £39.99 subscription bag. The starter pack had 3% margin after shipping - essentially a loss-leader. Previous agency had paused it because 'it doesn't meet ROAS targets.'

Keep the starter pack paused

Rationale

At 3% margin, break-even ROAS is 33x. Impossible to achieve profitably. Every sale loses money. Standard CPA-focused thinking says: don't advertise unprofitable products.

Projected Outcome

New customer acquisition drops 40% over 8 weeks. The subscription bags are profitable (42% margin) but customers need the starter pack to discover the brand. Without paid acquisition of the gateway product, the funnel dries up.

Restart the starter pack with LTV-adjusted bidding

Rationale

68% of starter pack buyers convert to subscription within 60 days. Average subscription tenure: 11 months. LTV per acquired customer: £287. Acceptable CAC at 30% of LTV = £86. The starter pack CPA of £14 is well within threshold.

Projected Outcome

Short-term ROAS on the starter pack campaign: 0.9x (a loss). 90-day ROAS including subscription conversions: 4.2x. 12-month customer value of each acquisition: £287 on a £14 CAC.

Our Decision

We chose Option B. We reported two sets of numbers: the campaign-level ROAS (which looked bad) and the customer-level ROI (which was exceptional). The founder needed to see both to understand why we were 'losing money on purpose.'

The Principle

Some products exist to acquire customers, not to generate profit. The question isn't 'is this sale profitable?' - it's 'is this customer profitable?'

Trade-Off 4

We rejected a recommendation that would have improved ROAS by 40%

Google recommended consolidating 6 Shopping campaigns into 1 Smart Shopping campaign for a home & living brand. Their optimisation score was 62%. Accepting would push it to 89%.

Accept the consolidation

Rationale

Google's data: consolidated campaigns perform 40% better on average. More data in one campaign = better algorithm learning. Higher optimisation score = better ad rank.

Projected Outcome

Blended ROAS would likely improve - because Google would shift spend toward the brand terms and high-converting, low-margin products that inflate ROAS. The 6-campaign structure existed specifically to prevent this cross-subsidisation.

Reject and maintain isolation architecture

Rationale

The 6 campaigns each serve a commercial purpose: 2 margin bands × 3 product categories. Each has a different tROAS target calibrated to its break-even point. Consolidation would remove the ability to bid differently on a £200 sofa (18% margin, 5.6x break-even) vs a £15 cushion cover (55% margin, 1.8x break-even).

Projected Outcome

ROAS stays at 3.4x (below the projected 4.8x from consolidation). But POAS stays at 1.3x - because we're not subsidising loss-making sofas with profitable accessories.

Our Decision

We rejected the recommendation and documented exactly why for the client. The 40% ROAS improvement would have come from destroying margin visibility and letting the algorithm optimise for the wrong objective. Google's recommendation was correct for Google's goals - not for the client's.

The Principle

Platform recommendations optimise for platform objectives. Your agency should optimise for yours. These are not always aligned.

The Point

Strategy isn't about knowing the answer. It's about knowing the question.

Every trade-off above has a "standard agency" answer - the one that optimises for platform metrics, follows Google's recommendations, and produces impressive ROAS numbers. Those answers aren't wrong. They're just answering the wrong question.

The right question isn't "how do I improve ROAS?" It's "how do I improve profit?" These sound similar. They lead to opposite decisions more often than you'd expect.

The agency that wins your business should be able to show you their reasoning, not just their results. Results without reasoning is luck. Reasoning without results is theory. You need both.

Frequently Asked Questions

About our decision-making process

Yes. Details are anonymised (company names, exact figures adjusted slightly) but the strategic situations, the options we considered, and the decisions we made are all drawn from actual client engagements. We believe showing our reasoning process - including the options we rejected - is the most honest way to demonstrate how we think.

We test before we commit. Every strategic shift starts with a controlled experiment - typically a 14-day window with clear rollback criteria. If the data contradicts our hypothesis, we reverse. The value isn't in being right 100% of the time - it's in having a systematic process for evaluating options, testing assumptions, and measuring outcomes against commercial metrics, not platform metrics.

We always optimise for contribution margin, not revenue. When two options generate similar profit, we choose the one with lower risk and better cash flow implications. When the data is ambiguous, we choose the option that's easier to reverse. Our decision framework prioritises: (1) profit impact, (2) risk profile, (3) reversibility, (4) strategic alignment with the client's commercial objectives.