How to read Google Ads reports without drowning in numbers
The Google Ads interface shows dozens of number columns, but deciding a campaign's fate genuinely needs three or four of them — the rest create an illusion of control.
Metrics that say nothing about money
Impressions and CTR answer 'is the ad seen' and 'is it clicked', but say nothing about whether the ad paid off. A high-CTR campaign can still produce zero sales if the audience is mistargeted.
Average ad position was historically important but has been replaced in the current Google Ads interface by impression share metrics, which more honestly reflect whether a campaign is losing traffic to a low bid or low budget.
An owner who watches only CTR and position risks approving a campaign that looks great in the report but delivers not a single real lead.
Metrics that actually matter
Cost per conversion is the first metric worth watching, since it directly answers 'how much does one customer cost through this channel' and compares against what the business can afford to pay for acquisition.
ROAS (return on ad spend) matters for product businesses with varied catalogue pricing — it shows whether revenue from the ads pays back the ads themselves, not just lead count.
Site conversion rate, not just the ad's own rate, shows where the buyer is lost: if clicks are cheap but site conversion is low, the problem is the landing page, not the ad setup.
Tying the report to real profit
Track not just cost per lead but how many leads actually closed as a deal — the Google Ads interface only shows the funnel's first step, and the business cares about the finish line.
Average order value and product margin determine what lead cost is acceptable — advertising with a lead cost above what margin allows is loss-making even if every other metric looks fine.
Set up conversion value passback wherever possible — not every lead is equal, and a campaign bringing fewer but pricier leads can outperform one with more cheap ones.
Traps in interpreting reports
Comparing periods without adjusting for seasonality produces false conclusions: a conversion dip in July versus May may be the niche's seasonal effect, not deteriorating campaign settings.
Last-click attribution undervalues top-of-funnel channels including GDN and YouTube — the default reporting model can suggest a channel isn't working when it's actually priming a future conversion through another channel.
A small data volume creates the illusion of a trend: conversion rising from two to four in a week is a doubling in percentage terms but a statistically insignificant change to draw far-reaching conclusions from.
Practice: three reports on a regular cadence
A weekly report on cost per conversion and lead volume per campaign is the baseline account health check, takes five minutes, and immediately flags deviations.
A monthly report on search terms and traffic quality goes deeper but runs less often, for keyword cleanup and structure review.
A quarterly review tied to real CRM sales is the rarest but most important report, because it's the only one that shows whether the ads bring money to the business, not just leads in the interface.