Your Marketing Dashboard Is Lying to You (And You're Making Decisions Based on the Lies)
A business owner logs into their Google Ads dashboard. They see 4x ROAS. Four dollars back for every dollar spent. That looks healthy. They increase the budget.
Six months later, the accountant calls. Profit is flat. Cash flow is tighter than it was before the ad spend went up. The dashboard still shows 4x ROAS. Nothing has changed in the numbers. Everything has changed in the bank account.
This isn't a hypothetical. It happens every week across Australian SMEs running paid advertising. The dashboard shows a version of reality that feels complete but isn't. And the scariest part isn't that the data is wrong. It's that your brain is wired to trust it anyway.
The metric everyone watches is the metric that matters least
ROAS (Return on Ad Spend) is the default metric in every Google Ads and Meta Ads dashboard. It divides revenue by ad spend. Simple. Clean. Dangerously incomplete.
Avinash Kaushik, former Chief Strategy Officer at Croud and one of the most cited voices in marketing analytics, lays out a progression that most SMEs have never seen. He calls it the accountability ladder, and it exposes why ROAS gives you a false sense of control.| Metric | Formula | What it tells you | What it hides |
|---|---|---|---|
| ROAS | Revenue / Ad Spend | "We made $4 for every $1 spent" | Cost of goods, margins, overhead, the cost of your team's time |
| ROI | (Revenue - Campaign Cost) / Campaign Cost | "After subtracting campaign costs, we're still ahead" | Product costs, fulfilment, returns |
| POAS | Gross Profit / Ad Spend | "After subtracting what we sold and what it cost to sell, we're ahead" | Overhead, salaries, rent, software |
| POI | Net Profit / Total Investment | "After everything, did we actually make money?" | Nothing. This is the truth. |
But the service delivery costs $6,000. Admin, tools, and overhead attributable to those jobs come to $2,500. The actual profit from $3,000 in ad spend is $500. That's a 0.17x POI.
The dashboard said 4x. The bank account said 0.17x. Both numbers are correct. Only one of them tells you whether you should keep spending.
Kaushik's point is stark: "ROAS gives full credit without subtracting ad spend. ROI subtracts campaign cost. POAS subtracts COGS. And POI reveals true profit after all costs." Most businesses never get past the first rung. The ones that do often discover their "best performing" campaigns are actually their least profitable.
Why your brain falls for it
The dashboard problem isn't just about incomplete data. It's about the way human brains process numbers under uncertainty. And this is where Daniel Kahneman's Nobel Prize-winning research on cognitive bias becomes uncomfortably relevant to your Tuesday morning ad review.
Anchoring bias. The first number you see becomes your reference point for everything that follows. When you open your dashboard and ROAS is front and centre, it anchors your entire assessment of campaign performance. Kahneman and Tversky demonstrated this in their famous UN experiment: people shown the number 65 guessed an average of 45 for an unrelated question. People shown 10 guessed 25. The anchor had nothing to do with the answer, but it moved estimates by 80%. Your ROAS number is doing the same thing. It sets the frame before you've had a chance to think critically. Confirmation bias. Once you believe a campaign is working (because ROAS says so), you unconsciously filter for data that confirms it. Click-through rate is up? Confirms the story. Cost per lead dropped? Even better. You stop looking for contradicting signals because your brain already has its narrative. As Kahneman writes in Thinking, Fast and Slow: "A reliable way to make people believe in falsehoods is frequent repetition, because familiarity is not easily distinguished from truth." Your dashboard repeats the same metrics every single day. The illusion of validity. Kahneman identified a specific cognitive trap where having more information makes you more confident in your judgment, even when the additional information doesn't improve accuracy. A dashboard with 47 metrics feels more informative than one with 6. In reality, research shows dashboards exceeding 12 KPIs have 40% lower engagement rates because users experience cognitive overload and default to the handful of metrics they already understand. More data doesn't mean better decisions. It means more confident bad decisions.This is why only 35% of small businesses can accurately tie revenue back to their marketing spend. It's not that business owners are careless. It's that the tools they use are designed to show them what's easy to measure, not what's important to know. And their brains are wired to accept the easy answer.
Goodhart's Law is eating your ad budget
There's a principle in economics that explains why optimising for dashboard metrics makes them progressively less useful. It's called Goodhart's Law: "When a measure becomes a target, it ceases to be a good measure."
Sam Tomlinson, whose newsletter on digital marketing strategy reaches thousands of practitioners, applies this directly to advertising. His observation: most marketing dashboards are "80% data pukes and 20% useful insights." The data looks impressive. The decisions it drives are often wrong.Here's how Goodhart's Law plays out in a real Google Ads account. You set a target ROAS of 4x. Your campaigns hit it. So you scale. But to maintain 4x ROAS at higher spend, the algorithm chases the easiest revenue. It shows your ads to people who were already searching for your brand name. It captures demand you would have won anyway. Your ROAS stays at 4x. Your incremental revenue from advertising approaches zero.
Tomlinson's solution is what he calls "metric trios": never show a KPI alone. Always pair it with a forecast (what you expected) and a benchmark (what's normal). A 4x ROAS means nothing in isolation. A 4x ROAS against a forecast of 6x and an industry benchmark of 5x tells you the campaign is underperforming. Same number. Completely different decision.
This is why Kaushik argues that an effective dashboard should be 80% action recommendations and 20% data, not the reverse. The question isn't "what are the numbers?" It's "what should we do differently because of the numbers?"
The cockpit problem: why more instruments make worse pilots
A fighter jet cockpit has hundreds of instruments. A pilot doesn't watch all of them equally. They scan a critical few, in a trained sequence, looking for deviations from normal. The rest is noise until something goes wrong.
Most marketing dashboards are the opposite. They show everything, weight nothing, and leave the interpretation entirely to someone who isn't trained to read them. That's not a dashboard. It's a wall of numbers with a logo on top.
Kaushik's rule is simple: maximum six KPIs on any dashboard. His framework for choosing them is even simpler. Every metric should answer one of three questions:
- Why are we doing this? (The strategic outcome)
- What is happening? (The operational indicator)
- How much did it cost? (The efficiency measure)
For an Australian SME running Google Ads and Facebook Ads, a truthful dashboard might look like this:
| Metric | Why it matters | Review frequency |
|---|---|---|
| Cost per qualified lead (not just cost per lead) | Connects ad spend to leads your sales team actually wants | Weekly |
| Lead-to-sale conversion rate | Shows whether the post-click experience is working | Monthly |
| POAS or POI (not ROAS) | Tells you whether advertising is actually profitable | Monthly |
| New customer acquisition cost | The real price of growth, including all costs | Monthly |
| Pipeline value | Forward-looking indicator of revenue, not backward-looking vanity | Weekly |
| Brand search volume trend | Leading indicator of mental availability building over time | Quarterly |
That's six. One page. Every number connects to a decision. Everything else is a diagnostic metric you check when something goes wrong, not something you watch every morning.
The most dangerous phrase in marketing meetings
"The numbers look good."
This phrase ends more productive conversations than any other. It anchors the room to whatever metric the speaker has in mind (usually ROAS or total leads). It triggers confirmation bias across the whole team. And it shields underperforming campaigns from the scrutiny they need.
A better phrase: "The numbers look good. What are we missing?"
Rory Sutherland, Vice Chairman of Ogilvy UK, makes a broader version of this argument in Alchemy. He calls it the "arithmocracy": the belief that every business question has a numerical answer, and that the person with the most data wins. "The fatal issue is that logic always gets you to exactly the same place as your competitors." If everyone is optimising toward the same dashboard metrics, everyone converges on the same strategy. The businesses that break out are the ones asking questions the dashboard wasn't designed to answer.
Questions like: Why do 65% of our leads never get a response? Why do people who find us through brand search convert at 3x the rate of generic search? What would happen if we spent 20% less on ads and invested that money in something the dashboard can't measure?
These are the questions that change growth trajectories. None of them show up in a standard marketing dashboard.
What This Means for Your Business
If you're an Australian SME spending $2,000 to $10,000 a month on advertising, here's the uncomfortable exercise.
Step 1: Calculate your real profitability. Take your ad spend. Add your agency fees, your team's time, your software costs. Subtract it all from the gross profit (not revenue) generated by ad-sourced customers. If the number is negative, your 4x ROAS is a fiction. Step 2: Cut your dashboard to six metrics. Use the table above as a starting point. If you can't explain why a metric is on your dashboard in one sentence, remove it. Step 3: Add context to every number. For each remaining metric, add a forecast (what you expected this month) and a benchmark (what's normal for your industry). A number without context is a number without meaning. Sam Tomlinson's metric trio framework prevents isolated numbers from anchoring your judgment. Step 4: Schedule a quarterly "what are we missing?" review. Not a performance review. A gap review. What questions can't your current data answer? What decisions are you making on incomplete information? This is where the real breakthroughs hide.The goal isn't to stop measuring. It's to stop being measured by your measurements. Your dashboard should serve your strategy, not replace it.
The businesses that grow aren't the ones with the most data. They're the ones who know which data to ignore.Further Reading
- Rules for Dashboard Revolutionaries - Avinash Kaushik's framework for building dashboards that drive action, not just display data
- Reporting Needs a Revolution - Sam Tomlinson on why most marketing dashboards are 80% noise and how to fix them
- The ROAS Trap: Why High ROAS Isn't Always Profitable - A detailed breakdown of how ROAS masks true campaign profitability
- Thinking, Fast and Slow - Daniel Kahneman's foundational work on the cognitive biases that shape every business decision
- Transform Data's Impact: Pick the Right Success KPI - Kaushik's ROAS-to-POI accountability progression explained
Dream Outcome is an Australian digital marketing agency helping SMEs grow through Google Ads, Facebook Ads, and Email Marketing.