Skip to content
Docs
Browse Analytics

Campaign trends explore

The Campaign trends explore compares paid campaigns' spend, cost per visit, cost per conversion and ROAS across time. What it covers, what to know when reading it, and every dimension and metric in it.

The Campaign trends explore compares your paid campaigns across time. It puts ad spend, paid visits, cost per visit, cost per conversion and ROAS on the days you pick beside the same days 7 days earlier and 52 weeks earlier, what those weekdays typically see, and an expected range.

Good to know

  • Every number is the Campaigns explore's, compared: spend, credit and paid visits are read the same way.
  • Filter Date to pick the days to look at; every comparison follows them.
  • Each comparison is a metric of its own, named after its metric, such as ROAS (7 days earlier).
  • Only complete days are rows. A day is the ad account's own for spend, and your organisation's for visits and credit.

Used on dashboards

Charts on these dashboards are built on the Campaign trends explore:

  • Trend and anomaly: what changed and why?: Week over week, Year over year, Unusual days and What drove it

Dimensions

The Campaign trends explore has 9 dimensions.

DimensionTypeDescription
DateDateThe day being looked at: the ad account's day for spend, and the organisation's for visits and credit. Every metric counts these days, or the days it compares them with. Time intervals: Day, Week, Month, Quarter, Year, Day of the week (index), Day of the week (name).
ChannelStringThe channel of the visits, and of the visit credited; for spend, the channel its source and medium read as, as in the Campaigns explore.
PlatformStringThe ad platform the spend was uploaded from, as in the Campaigns explore.
SourceStringThe campaign's source, lower-cased and trimmed, as spend is matched to visits.
MediumStringThe campaign's medium, lower-cased and trimmed, as spend is matched to visits.
CampaignStringThe campaign, by the name its spend gives it, as in the Campaigns explore.
Campaign IDStringThe ad platform's campaign id, as in the Campaigns explore.
Paid campaignBooleanWhether the campaign has spend uploaded, on any day.
SiteStringThe property, as in the Campaigns explore.

Metrics

The Campaign trends explore has 30 metrics.

MetricCalculationDescription
Ad spendSumWhat the ad platforms charged, from the files uploaded on the Ad spend page, in the organisation's reporting currency, on the ad account's own days. On the days the Period filter picks, in the organisation's time zone, complete days only.
Paid visitsSumVisits that arrived from a campaign with uploaded spend. On the days the Period filter picks, in the organisation's time zone, complete days only.
Cost per visitRatioSpend per paid visit. On the days the Period filter picks, in the organisation's time zone, complete days only.
Cost per conversionRatioSpend per conversion credited to a paid campaign by last touch. On the days the Period filter picks, in the organisation's time zone, complete days only.
ROASRatioRevenue credited to paid campaigns by last touch per unit of spend. On the days the Period filter picks, in the organisation's time zone, complete days only.

Comparisons

Each metric above also comes in 5 versions named after it, such as Ad spend (7 days earlier):

VersionWhat it shows
7 days earlierThe same days 7 days earlier: week over week. Empty while any of those days is in Kontala's first recorded week.
52 weeks earlierThe same days 52 weeks earlier, so each weekday meets its own: year over year. Empty until Kontala has recorded 52 weeks before every one of the days.
TypicalWhat these days usually see: the average of the same weekdays in the 8 weeks before, or in as many of them as Kontala has recorded. For a ratio, the same weekdays in the 8 weeks before, or in as many of them as Kontala has recorded, added up before dividing, so a busy week counts for more than a quiet one.
Expected lowThe bottom of the expected range: 95% of normal periods land between the expected low and high, read from how much the same weekdays varied in the weeks before. Needs 4 weeks of them.
Expected highThe top of the expected range: 95% of normal periods land between the expected low and high, read from how much the same weekdays varied in the weeks before. Needs 4 weeks of them.

Was this helpful?