A low advertising cost can be encouraging and still answer the wrong question. An inexpensive impression is not automatically a useful visit, and a useful visit is not automatically a profitable customer. CPM, CPC and CPA become more informative when you know which event each describes and which decision you are trying to make.

Start by separating the price you agree to pay, the objective you ask a system to optimize and the results you calculate afterward. They can use similar names while referring to different things.

Separate the bill, the bid and the result

The billing basis identifies the event used to calculate media charges. CPM pricing uses thousands of impressions; CPC pricing uses clicks; CPA pricing uses a defined action. The Campaign Manager 360 media-cost documentation illustrates these distinct cost methods. The applicable agreement and reporting definitions determine what is billable.

The bidding objective guides how a platform tries to achieve an outcome. For example, Google Ads describes target CPA as a desired average cost per conversion used by its bidding strategy. That target is not a promise that every conversion will cost that amount, and it is not the same as a fixed CPA payment agreement.

The reporting metric summarizes observed cost relative to a recorded event. You can calculate an effective cost per click for an impression-billed campaign. That calculation helps evaluate the result; it does not change the basis of the invoice.

Read the denominator before the number

The formulas are simple. Keeping the underlying populations consistent is the harder part. Use the same campaign scope, currency, cost basis and reporting period for the numerator and denominator. Identify whether conversions are attributed to that same advertising activity.

Effective costs calculated from an observed campaign report
Metric Calculation Question it answers
eCPM Cost ÷ impressions × 1,000 What did a thousand recorded impressions cost?
eCPC Cost ÷ clicks What did each recorded click cost on average?
eCPA Cost ÷ attributed actions What did each defined, attributed action cost on average?

The “effective” label signals a calculated result, although report labels vary. Google’s average CPC definition likewise uses total click cost divided by total clicks and distinguishes the result from a maximum CPC bid.

When the denominator is zero, the rate is not a meaningful finite number. If a campaign spends money but records no purchases, showing a purchase CPA of zero would incorrectly suggest free purchases. Keep the cost and zero outcome visible, with the ratio marked unavailable.

Define what the action is worth

“Conversion” is a reporting category, not a universal business result. A completed purchase, a newsletter signup and a request for a sales call can all be tracked actions. Google’s conversion-measurement overview describes several such actions. Decide which one belongs in the CPA you are using.

Write a definition a colleague could reproduce: completed paid order, excluding test orders, using a specified reporting source. Keep lead submissions separate from qualified leads if qualification matters to the business. A cheaper submission may still require more work or produce fewer customers.

Counting settings matter too. Google Ads supports one or every conversion per ad interaction for a conversion action. Those choices can change the denominator. Neither setting should casually be described as a count of unique people.

Record the attribution model, window and treatment of repeated actions beside the metric. Our attribution guide explains why two reports can assign different credit to the same business outcome.

Work through a complete comparison

This example is fictional and is not a benchmark. Imagine two campaigns for the same product, with equally mature measurement windows and the same purchase definition. Each spends $300. The figures below illustrate arithmetic, not a controlled experiment or a claim about any platform.

Two invented campaign reports with the same media spend
Measure Campaign A Campaign B
Media cost $300 $300
Impressions 60,000 30,000
Clicks 600 300
Attributed purchases 12 15
Effective CPM $5 $10
Effective CPC $0.50 $1
Effective purchase CPA $25 $20
Click-through rate 1% 1%
Attributed purchases ÷ clicks 2% 5%

Campaign A buys more impressions and clicks for the same spend. Campaign B records more purchases and a lower purchase CPA. Both have the same click-through rate, calculated as clicks divided by impressions. None of those statements contradicts the others.

The purchase-to-click ratio here assumes the stated matching reporting scope; it is not a universal platform conversion-rate definition. Before preferring B, compare order value, margin, cancellations and whether these observations are stable enough to support the next decision. The table alone does not establish that a campaign change caused the difference.

Combine totals instead of averaging rates

Averages can hide unequal volumes. In a separate fictional example, one group spends $100 for 100 clicks and another spends $900 for 300 clicks. Their effective CPCs are $1 and $3. Taking the simple average gives $2, but the combined result is $1,000 divided by 400 clicks: $2.50.

Add compatible costs and events first, then recalculate the ratio. The same principle applies to CPM and CPA. Combining incompatible actions or overlapping reports will not become valid just because the arithmetic is correct.

Also name the cost included. Media-only CPA and CPA including creative production or service fees answer different planning questions. Keep the narrower media comparison available, and document any broader cost allocation instead of silently mixing the two.

Use a change to guide the next investigation

When a rate moves, inspect its components. A higher effective CPC could accompany more expensive impressions, a lower click-through rate, or a different mix of placements. Those are investigation paths, not automatic explanations.

Compare like-for-like segments before declaring a trend. Device, placement, offer and audience differences can change the aggregate. Check whether recent conversions have had enough time to be reported and whether the cost report has been adjusted.

If inexpensive impressions are the main attraction, assess the quality and measurability of that inventory. If clicks look healthy but purchases do not, review the landing experience and conversion recording. A price metric alone cannot confirm either traffic validity or a usable destination.

Build a scorecard around a decision

Keep the working report compact enough to use:

  • Primary outcome: name the action and why it matters.
  • Cost scope: identify currency, included charges and reporting period.
  • Volume: show spend, impressions, clicks and actions beside the ratios.
  • Measurement rules: record counting, attribution and the comparison window.
  • Decision: specify what evidence would justify continuing, investigating or changing the campaign.

Connect that review to a budget plan and an interpretable testing process. The useful outcome is a clearer decision about the next investment, supported by numbers whose meaning everyone understands.

Sources & further reading

Explore advertising with Promoweb.
Meet our self-serve platform