New Dashboard Analytics Now Live
The dashboard now leads with real-time verified-action counts, seven-day rolling averages and a cost-per-verified-action chart, replacing a clicks-first view that encouraged optimising the wrong metric.
Dashboards are opinionated whether they intend to be or not. Whatever number sits at the top, largest, is the number people optimise — and ours was pointing at the wrong one.
What Is New
Real-time verified-action counts, seven-day rolling averages, and a cost-per-verified-action chart that finally makes campaigns comparable at a glance.
Three changes, each solving a specific problem:
Real-time verified-action counts. Previously the prominent live number was clicks, with verified actions arriving later and less visibly. Now the count that represents real outcomes is the one you see first.
Seven-day rolling averages. Daily figures swing for reasons that have nothing to do with your campaign, and verification lag makes any single day incomplete by construction — thorough reviews take longer to submit and longer to verify, so today’s numbers systematically over-represent fast, shallow activity. Rolling averages are now the default view rather than something you had to construct yourself.
A cost-per-verified-action chart. This is the one that changes decisions. Cost per verified action is directly comparable across campaigns in a way cost per click is not, because a verified action means the same thing everywhere. Having it charted over time, per campaign, is what makes reallocation possible.
Why We Changed It
The old dashboard reported clicks first. That trained people to optimise the wrong number, so we rebuilt it around verified outcomes.
The failure mode was consistent enough to be predictable. Cost per click is the most available metric, so people optimised it. Optimising it pushes you toward the cheapest clicks, which are reliably the traffic least likely to do anything. The headline metric improved while results declined — the worst possible combination, because it looks like progress.
A worked example of what the new view makes obvious:
| Campaign A | Campaign B | |
|---|---|---|
| Spend | $500 | $500 |
| Clicks | 2,000 | 800 |
| Cost per click | $0.25 | $0.63 |
| Verified actions | 40 | 95 |
| Cost per verified action | $12.50 | $5.26 |
On the old dashboard, Campaign A looked like the winner. Campaign B is more than twice as efficient at the thing you are actually buying. Advertisers optimising the old headline number were cutting their better campaigns.
How to Use It
The new layout supports a weekly routine rather than daily checking:
- Sort campaigns by cost per verified action. Seven-day average, not today.
- Compare against last week. Direction matters more than the absolute number.
- Segment the worst performer — by region, interest, device, publisher tier.
- Cut the bottom performers and move that budget to the top rather than adding to total spend.
- Make one change, then wait 48 hours before judging it.
That reallocation discipline is what separates campaigns that improve from campaigns that just get more expensive.
What Not to Do With It
Do not read a single day. The chart now makes daily figures easy to see, which is a mild hazard. Verification lag means one day is never a complete picture, and reacting at hour twelve is how good campaigns get killed early.
Do not compare cost per verified action across packages. A traffic action and an engagement action are different units — traffic actions are cheap and shallow by design. Compare within a package.
Do not expect it to explain anything. The dashboard tells you where performance is poor. It cannot tell you why. That answer is in the reviews themselves, which is what the Rewards tab is for.
Reading the Rolling Average Properly
Seven-day rolling averages solve one problem and introduce a smaller one worth knowing about.
They smooth out daily noise, which is the point. But they also lag: a change made on Monday is diluted across the following seven days of the average, so the full effect is not visible until the window has fully turned over. A change that looks modest on day three may look substantial on day eight.
The practical rule: judge a change no earlier than 48 hours, and no later than the point where you have forgotten you made it. Writing down the change and the date is the cheapest way to keep the comparison honest.
If you need a faster read than a seven-day window allows, the answer is more volume rather than a shorter window. Roughly fifty verified actions a week is where week-to-week variation stops dominating; below that, a shorter window will just show you noise more quickly.
What Has Not Changed
The underlying verification is identical. Spend is still only drawn against verified activity, work that fails verification is still not billed, and the definition of a verified action is unchanged. This release changes what the dashboard shows first, not what it counts.
Historic data is intact. The cost-per-verified-action chart backfills from existing campaign history, so you can see how campaigns performed before the change rather than starting a new baseline.
Frequently Asked Questions
Where did the clicks metric go?
It is still available, just no longer the headline figure. Clicks remain useful as a diagnostic — a click count rising while verified actions stay flat tells you traffic is arriving and doing nothing.
Why is my cost per verified action higher than my cost per click suggested?
Because they measure different things. Not every click becomes a verified action, so cost per verified action is always higher. It is also the only one of the two that tracks value.
Can I still see daily figures?
Yes, but treat them carefully. Verification lag means any single day is incomplete by construction, and the careful publishers have not reported yet.
Does the chart include failed verifications?
No. Failed work is not billed and does not appear as an action. What you see is what you paid for.
Related Reading
Reading your analytics covers which metrics matter and which mislead. A beginner’s guide to the ZetPulse dashboard covers the full interface, and setting a budget that actually converts covers acting on what you see.
