I sat in on a quarterly review a few months ago where a client's marketing lead pulled up a slide showing CTR was up 40% year over year. Everyone was happy to see a green arrow on the dashboard. It was an accurate performance metric, but it was also irrelevant to any business goal.
Impressions, CTR, engagement rate, etc. are unlikely to ever look bad enough to force a difficult marketing strategy conversation. A number that only trends in one direction, or that has no benchmark a client would recognize as a failure, isn't communicating results. They're also completely disconnected from pipeline and revenue in most cases.
Why these metrics survive
Platform dashboards default to whatever's easiest to compute, not whatever's decision-relevant. Impressions and CTR sit at the top of every native reporting view because Google Ads outputs them automatically, not because they're the right numbers to lead a report with.
There's also a literacy gap on the client side that reporting doesn't always account for. Without an explicit, shared definition of what "good" means for a given account, any number moving in the right direction is treated as progress, whether or not it had anything to do with pipeline.
Now, these metrics are not totally worthless. Impressions, CTR, and other metrics can be useful directional signals for performance. The problem is specifically their promotion to headline KPI status on a client-facing report, where they take up space that would otherwise go to something that could justify a strategic or budget decision, like pipeline or revenue generated.
The test
Ask yourself this before adding a metric to a report: has this number, in the actual history of this account, ever changed what the team did next? Not whether it could in theory. Whether it has, on record, triggered a budget shift, a paused campaign, a rewritten ad, or an escalation to the client?
Vanity metrics to eliminate and replace
Impressions
Impressions measure reach, not intent or outcome. An account can double its impressions by broadening match types or raising bids on low-relevance terms, and the number will look great while cost per lead gets worse. A low impression count doesn't obviously mean "bad," and a high one doesn't obviously mean "good."
Replace it with spend and conversion volume broken out by intent tier (branded, high-intent non-branded, top-of-funnel). That framing forces the same underlying data to answer a question that actually matters: where is the budget going, and is that where the pipeline is coming from?
Click-through rate
CTR is as much a function of ad position, match type, and SERP layout as it is of message quality. A client with no benchmarking context can't tell a genuinely strong CTR from one that's just riding a favorable auction. Reported on its own, it invites the wrong conclusion in both directions: a rising CTR gets read as "the ads are working," a falling one gets read as "the ads got worse," and neither is necessarily true.
Instead, report CTR paired with the conversion rate segmented by funnel stage. This gives you a deeper understanding of ad performance and your customer journey.
Engagement rate and bounce rate
Engagement rate doesn't map to anything downstream. A session can be "engaged" by GA4's definition (10+ seconds, a conversion event, or two-plus pageviews) without the visitor moving any closer to becoming a lead. Engagement rate can climb while pipeline stays flat, and there's no way to tell from the metric alone whether that's happening.
Report conversion rate by channel and by landing page instead. If you have a specific reason to care about on-page attention (a content-heavy page where dwell time is a real proxy for read-through), scroll depth or time-on-key-section is a better version of the same idea; just attach it to a specific hypothesis rather than reporting it by default as a win.
Takeaways
A shorter report where every metric has, at some point, changed a decision is more useful than a longer one where none of them have, even if the longer one looks more thorough. Thoroughness isn't the goal of a client report; the ability to act and communicate value is.
Caveats
- This is about what faces the client, not what you monitor internally. Impressions, CTR, and other metrics remain useful for day-to-day pacing and QA; the fix here is about what gets promoted to a headline KPI on a client report.
- Cutting a metric a client is emotionally attached to, such as total clicks, needs to be handled as a conversation, not a silent removal from the next deck next month.