Every ad-stuffed webpage is a reminder of how meaningless ‘Impressions’ can be as a KPI.
It’s a QA metric at best – a check that everything’s still delivering.
But whether the campaign should still be running is the real question.
Imagine you’re running Awareness activity for a new brand or product – the cost of reaching new users is a fundamental KPI to monitor. Your eCPM might be low, but you could be reaching hardly anyone new.
“What was the reach of our campaign yesterday?”
(log into platform … run report for yesterday … )
“12,452”
“Great. How many of those were people who’d not seen the campaign before?”
“err”
How to figure that out? Using the platform reporting you’d have to run a campaign to date report up to the end of the day before yesterday, take that number, then run the report up to end of yesterday, and take the difference between these 2 numbers. that’s a manual hassle, totally unscalable and inefficient.
But it’s essential if you want to answer the crucial question of whether your ad spend is actually achieving its goal. Imagine you’re running Awareness activity for a new brand or product – the cost of reaching new users is a fundamental KPI to monitor
We think we’ve cracked it – and maybe invented a new KPI along the way.
“Cost per Thousand Incremental Reach” (CPMIR) isn’t the snappiest name, but it tells an important story about the health and efficiency of your campaign.
This isn’t a KPI you can pull from the platforms. At Bright Analytics, we calculate this awkward KPI automatically – and put it in front of planners and campaign teams every day, with zero manual effort.
• It shows how many unique users your ads have reached
• It’s essential for calculating Frequency (and spotting when you’re irritating people)
• Crucially – it shows when your campaign starts burning budget without reaching new users
Because it’s fiddly.
Reach is a unique count of people (really, devices or cookies), and the number changes based on:
• The time period you’re analysing
• The granularity of the data (campaign, ad set, ad, etc.)
Which creates some nasty traps:
• Pull data daily → you’ll overcount Reach (same person, different days)
• Pull data by ad → you’ll overcount Reach (same person, different creatives)
Conclusion: you can’t accurately calculate Reach from your usual delivery data.
You need a different approach.
Say you log into an ad platform, run a report for last week, and look at the Reach figure.
You’re seeing how many people you reached that week. Great! Except… not really.
That number doesn’t tell you how many of those people were already reached in earlier weeks.
That’s a big blind spot. And platform UIs don’t help fill it.
We think there should be a KPI for monitoring the incremental Reach a campaign achieves each day — and we’ve worked out how to automate it.
This is a metric you can look at daily, and see how many genuinely new people your campaign has reached since it launched.
You can chart that against your eCPM and spot when things start to slow down – for example, when the cost of reaching new users begins to spike.
See the chart attached to this post:
A steady eCPM across the campaign, but a rising iCPM – your budget’s still being spent, but you’re no longer expanding your audience.
Where platform APIs allow, we:
• Pull a list of campaign IDs and their start dates
• For each campaign, fetch the total Reach from launch to today
• Calculate daily incremental Reach by comparing each day’s total to the day before
• Store that value in a custom field in our Reach dataset
This gives our customers a daily operational metric – iCPMR – which they can trend over time, or aggregate over any period.
It works for both historic and in-flight campaigns, so you can pinpoint exactly when Reach began to tail off – and why.
And because the process is fully automated (even across hundreds of campaigns), it becomes part of your core reporting – not an afterthought.
Of course, we’re not deduping Reach across platforms (still a dream) – but we’re as rigorous as the APIs allow.
If a campaign is still spending but the cost of reaching anyone new becomes prohibitive, that’s a red flag worth spotting.
If you’ve spent too many hours exporting data and trying to work out Reach in Excel, you’re not alone.
Our self service platform eats complexity for breakfast. With our team of friendly data specialists on hand to help, we will have a solution perfectly tailored to your needs up and running in under a week.
Get In Touch