PPC reporting metrics: spend, CPC, conversions and ROAS, explained carefully

Start from 'what did we get for the money'

Every PPC report is a purchase justification. The client spends money on clicks; the report's job is to show what those clicks produced — and where the chain between spend and outcome is weak. Structure the metrics in funnel order: money in, attention bought, action produced, value returned.

That ordering also tells you where to look when results disappoint. Weak outcomes with strong spend point at the click→conversion stages; weak spend with strong efficiency points at budget, not performance — different problems, different conversations.

Spend and impressions: the money in

Cost (spend) is the anchor everything else ratios against. Report it per channel and in total, against the period's budget if there is one — pacing matters as much as the total. A month that spent 60% of budget and hit targets is a different story from one that overspent to reach the same number.

Impressions and reach sit underneath as diagnostics: they answer whether the budget even bought attention. Impression share, where the account exposes it, goes a step further — it says how much of the available attention the budget captured, which is the cleanest 'should we spend more?' signal paid search offers.

CTR and CPC: the efficiency diagnostics

Click-through rate (clicks ÷ impressions) and cost per click (spend ÷ clicks) measure the middle of the funnel: is the ad compelling, and what does the platform charge for it? CTR rises with relevance and creative; CPC rises with competition and bid pressure — and the two often move together in ways that need reading together.

Conversions: check the configuration before the number

'Conversions' in Google Ads means whatever the account's conversion actions define — purchases, leads, page views if someone configured it that way — counted by each action's own rules. A conversion column can include primary actions only or every action; counting can be one-per-click or every occurrence. The same traffic produces different totals under different settings.

This is also why Ads and GA4 conversion figures differ structurally — different definitions, different attribution, different date bases. That gap deserves its own explanation (we wrote a full investigation guide, linked at the end); in the report itself, label each conversion figure with its source so a discrepancy reads as a definitions question, not an error.

Attribution windows bend the number further. A 30-day click window and a 7-day click window report different conversion totals for identical traffic — conversions keep arriving after the click for weeks. When the report compares this month to last, both figures carry their windows; a mid-period window change is a methodology change, and it belongs in the footnotes.

CPA and ROAS: the outcome ratios

Cost per acquisition (spend ÷ conversions) is the lead-gen bottom line: what each conversion cost. Return on ad spend (revenue ÷ spend) is the ecommerce equivalent: what each pound bought back. Both are only as trustworthy as their inputs — CPA inherits whatever 'conversion' means in that account; ROAS inherits the revenue value tracking, which may come from the ad platform's tag, GA4 import, or the store itself.

State which revenue the ROAS uses. A 4.0 ROAS from the platform's own conversion value and a 4.0 from Shopify orders can describe different months. And resist quoting ROAS for lead-gen clients — without revenue values the number is invented; CPA is the honest metric there.

For multi-channel accounts there is a blended cousin worth knowing: total revenue ÷ total ad spend across all platforms (sometimes called MER, marketing efficiency ratio). It sidesteps per-platform attribution arguments — instead of debating which platform 'gets' the sale, it measures whether the whole paid budget is pulling its weight. Per-platform ROAS still earns its place for budget allocation; blended is the honest number for the board question.

The diagnostic layer: search terms, campaigns, devices

Below the funnel metrics sits the detail that explains movement: which campaigns drove the spend, which search terms triggered the clicks, which devices converted. This layer isn't for the headline row — it's for the 'why' when a number moves, and for the work-completed section that shows what you're managing.

Keep it selective. A top-spend-campaigns table and a shortlist of notable search terms carry the story; a full keyword dump is the ad account exported, not reported.

This is also where negative-keyword work shows up — pruning wasted queries is invisible in the headline numbers but legible here. A line naming what you excluded earns its place: it shows the budget actively defended, not just spent.

Compare periods like-for-like

Month-on-month PPC comparisons carry a specific trap: months differ in length and weekday mix, and paid spend follows business days. A 31-day month 'outperforming' a 28-day one can be pure calendar. Equal-length windows or daily-rate comparisons keep the comparison honest; seasonal businesses want year-on-year context instead.

Promotions, sales events and budget changes mid-period deserve annotations in the report — a spike explained beats a spike that invites the wrong question at the review.

A PPC report layout that reads in order

• Headline: spend vs budget, conversions, CPA or ROAS — the verdict row.
• Efficiency: CTR and CPC trends per channel.
• Channel split: each platform's figures labelled by source.
• Diagnostics: top campaigns, notable search terms, device mix.
• Work and next steps: what was managed, what changes next period.

ReportingBee composes this layout from the connected ad accounts — Google, Meta, Microsoft and more in the same branded report, each figure labelled with its platform. See the structure on PPC reporting software, or the channel-specific view on Google Ads reporting.