The real cost of a skewed ROAS: when your budget decisions rest on broken data

The cost lies in the decision that follows

When people talk about dead pixels, double counting or corrupted conversion data, the conversation always centers on the technical problem. How to detect it, how to fix it, how long it takes.

The real cost comes later, from the budget decision you make based on wrong data and only question weeks afterwards.

A Meta pixel that's dead for 72 hours costs the budget wasted during the outage, plus the wrong budget reallocation the CMO makes the following Monday after seeing Meta ROAS collapse.

The four situations that skew ROAS

1. Dead pixel → ROAS collapses artificially

The conversion pixel stops firing. Meta, Google Ads or TikTok no longer receive a purchase signal. Reported ROAS drops to 0 or close to it.

The wrong decision: "Meta has stopped converting, let's cut its budget by 40% and move it to Google."

The reality: Meta is still converting. It's the tracking that's broken. By cutting the budget, you shut down a profitable channel. By raising the Google budget with the "freed-up" money, you may be overinvesting in a channel that performs worse for those audiences.

2. Double counting → ROAS inflated artificially

A dataLayer bug sends two purchase events per transaction. Or the pixel fires both client-side and server-side with no deduplication. Reported ROAS doubles.

The wrong decision: "Meta ROAS went from 3 to 6. Let's raise the budget by 50% to scale."

The reality: real ROAS is still 3. The extra 50% of budget will run at a real ROAS of 3, not 6. If your break-even point is 3.5, you're investing in a loss-making channel.

3. Blocked checkout → ROAS collapses on every channel

The checkout is broken: JavaScript error, API timeout, unresponsive button. No purchase is recorded on any platform.

The wrong decision: "Every channel has a disastrous ROAS this week. Let's cut budgets until we figure it out."

The reality: the channels are working. It was the site that was broken for 2 hours on Tuesday. The weekly numbers are polluted by a one-off technical incident. Cutting budgets amplifies the loss instead of offsetting it.

4. Sponsored out-of-stock → ROAS degraded on the channel

Paid traffic lands on out-of-stock product pages. Clicks are billed, conversions are zero. The channel's ROAS drops.

The wrong decision: "Google Shopping campaigns have a ROAS down 25%. Let's lower the bids."

The reality: ROAS fell because 3 best-sellers were out of stock over the weekend. The bids don't need lowering; the product feed needs syncing.

The cost of a bad budget decision

For a site with a €100,000 monthly paid budget split across 3 channels:

The average cost of a bad budget decision based on broken data: €5,000 to €15,000. That figure covers the wrong decision alone, on top of the technical incident.

The knock-on effect

The problem doesn't stop at the week of the bad decision. There's a knock-on effect:

Week 1: The technical incident skews the data. Reported ROAS doesn't reflect reality.

Week 2: The CMO makes a reallocation decision based on week 1 data. The budget is redistributed.

Week 3: Bidding algorithms adapt to the new budgets. The underfunded channel loses its positions. The overfunded channel shows diminishing returns.

Week 4: The monthly review arrives. The numbers don't add up. The team spends 2 days figuring out what happened. The CMO readjusts, but it still takes 1 to 2 weeks for the algorithms to recover their performance.

In total, a 3-day technical incident can disrupt paid performance for 3 to 4 weeks. The cumulative cost often exceeds €20,000 for a €100,000/month paid budget.

Why the problem is so common

Three structural reasons:

1. The decision cycle is faster than the detection cycle. The CMO makes budget decisions every week. A dead pixel is detected after 3 days on average. The wrong decision is made before the incident is identified.

2. Nobody checks data reliability before deciding. The standard workflow is: open the dashboard → read ROAS → adjust the budget. Nobody asks "are these numbers reliable?" before making the call.

3. Monitoring tools and decision tools are disconnected. Sentry monitors technical errors. GA4 displays conversions. The CMO works in Google Ads Manager. There's no bridge between "the pixel is dead" and "don't touch the budget this week".

How to protect your budget decisions

1. Run a "data health check" before every decision. Did all pixels fire normally? Was there a technical incident? Are the gaps between platforms within their usual ranges? If the answer is "no" or "I don't know", postpone the decision.

2. Always compare with the backend. The order count in Shopify/Prestashop is the only source of truth. If Meta shows 5 conversions and the backend shows 130, the pixel is dead.

3. Automate data anomaly detection. Korvus monitors the four situations that skew ROAS (dead pixel, blocked checkout, degraded performance, sponsored out-of-stock) and alerts you before the next budget decision cycle.

4. Log every incident. Note the date, duration and impact on conversion data. A weekly ROAS that includes 2 days of dead pixel is worthless for decision-making.

Key takeaways

• The cost of a skewed ROAS lies in the bad budget decision that follows the technical error
• A 3-day dead pixel can lead to 3 to 4 weeks of disrupted performance through the knock-on effect
• The average cost of a bad budget decision based on broken data is €5,000 to €15,000
• Double counting inflates ROAS by 50 to 100%, leading to overinvestment in loss-making channels
• The only protection: check data reliability before every decision, not after