Automation in Google Ads Doesn't Understand Your Balance Sheet

Google's automation is remarkably sophisticated. It can predict conversion probability at the auction level. It can adjust bids in real-time based on hundreds of signals. It can find audiences you never knew existed.

What it cannot do is read your balance sheet.

The Fundamental Disconnect

Here's the core problem: Google's algorithms optimise for advertising metrics. Your business survives on financial ones.

These are not the same thing.

When Google reports a 450% ROAS, it's telling you about the relationship between ad spend and attributed revenue. What it's not telling you:

• Whether that revenue is profitable after COGS
• When you'll actually receive the cash
• Whether it required inventory you couldn't afford to hold
• If the returns rate made the sale worthless
• Whether you needed that revenue this month or could have waited

The algorithm sees a conversion. Your CFO sees something very different.

The Cash Flow Blind Spot

Consider a typical scenario we see in audits:

An e-commerce brand is hitting target ROAS consistently. The account looks healthy. But the business is struggling with cash flow.

Why? Because the algorithm has learned to optimise for:

• Larger basket sizes (which require more inventory capital)
• Products with longer shipping times (delaying cash receipt)
• New customer acquisition (which has higher CAC and lower immediate LTV)
• Weekend purchases (when payment processing is slower)

None of these optimisations are wrong from an advertising perspective. They're all hitting the ROAS target.

But they're slowly bleeding the business of working capital.

Margin Is Invisible

Unless you're feeding margin data into Google Ads (and implementing it correctly, which almost nobody does), the algorithm treats all revenue as equal.

A £50 sale of a product with 60% margin is worth £30 of gross profit. A £50 sale of a product with 15% margin is worth £7.50 of gross profit.

To Google's algorithm, these are identical. Both are £50 conversions.

This creates a systematic bias toward lower-margin sales if those products have higher conversion rates—which they often do, because lower-margin products are frequently lower-priced or more heavily discounted.

The algorithm is doing its job. Your profitability is suffering.

The Inventory Problem

Here's one that rarely gets discussed: automation doesn't understand inventory dynamics.

When your algorithm learns that Product A converts better than Product B, it will push more traffic to Product A. Logical.

But what if:

• Product A has limited stock and long lead times?
• Product A ties up more working capital per unit?
• Product A has higher return rates?
• Product B has excess inventory costing you warehouse fees?

The algorithm doesn't know. It's optimising for conversion, not for inventory efficiency.

We've seen accounts where the "best performing" products from an ads perspective were actually the worst from a business perspective—tying up capital, creating stock-outs, and generating returns.

Payment Terms Don't Exist

Your algorithm doesn't know that:

• You have 60-day payment terms with your manufacturer
• Your payment processor holds funds for 7 days
• Returns can be made up to 30 days post-purchase
• Chargebacks take 90 days to resolve

It sees a conversion and counts the revenue immediately.

This creates a dangerous illusion of cash flow that doesn't match reality. The gap between attributed revenue and available cash can be substantial—and it's completely invisible in your Google Ads dashboard.

Seasonality Without Survival

Algorithms learn seasonal patterns. They know when conversions spike and when they decline.

What they don't learn is your survival threshold.

If you need £50,000 of cash in March to pay suppliers, the algorithm won't help you get there. It will optimise for whatever target you've set, regardless of whether that target aligns with business survival.

We've seen accounts where algorithmically "optimal" performance in Q1 left businesses unable to pay for Q2 inventory.

The Solution Isn't More Data

The common response is: "Just feed more data to Google."

Import margin. Import LTV. Import profit.

This helps, but it doesn't solve the fundamental problem. Because the algorithm still doesn't understand:

• Cash flow timing
• Working capital constraints
• Opportunity cost
• Business phase (survival vs growth)
• Risk tolerance

These are human judgments that need to be translated into bidding strategy and account structure. No amount of data import substitutes for commercial thinking.

What This Means in Practice

When we take on new accounts, we don't just look at advertising metrics. We ask:

• What does this business need, financially, in the next 90 days?
• Where are the working capital pressure points?
• Which products help cash flow and which hurt it?
• What's the real margin by SKU or category?
• How long is the cash conversion cycle?

Only then can we configure an account that actually serves the business—not just the algorithm.

What We Look For

In audits, this is a critical area of investigation: is the account optimising for advertising success or business success?

We examine how targets are set, what signals are being used, and whether the bidding strategy makes sense given financial realities.

Because a "high-performing" Google Ads account attached to a failing business is not high-performing. It just looks like it is.