Reports That Find Profit: Drill-Downs by Source, GEO and Sub ID

Start from money, not clicks

A report sorted by clicks tells you where the traffic went. A report sorted by profit (revenue minus cost) tells you where to act. Before any drill-down, make sure costs are in the tracker for the period you look at; without them every segment looks profitable.

Four columns carry most decisions: clicks, conversions, EPC (revenue per click) and ROI. CR explains why EPC moved; CPC explains why ROI moved.

Drill down in layers

Go from broad to narrow, one dimension at a time, and stop at the level where the numbers become too small to trust.

1. Campaign: which campaigns earn at all.
2. GEO inside a campaign: countries often differ more than creatives.
3. Device and OS: mobile and desktop can be two different businesses.
4. Placement or site (a sub ID): where most of the waste usually hides.
5. Creative or keyword (another sub ID): what to rotate out.

A report builder with two or three grouping levels saves hours here: campaign, then GEO, then placement, in one table with expandable rows. XPNSR Tracker groups by up to three levels and any of 15 sub IDs; other tools have similar views.

Cutting and scaling

Cutting

• Sort placements by cost with zero or negative profit, and cut from the top. The biggest losers matter more than the many small ones.
• Set a spend threshold for a decision, for example two or three times the payout without a conversion.
• Cut at the source when you can. Blocking only in the tracker still pays for the clicks.

Scaling

• Scale segments with stable ROI over at least a week, not the best single day.
• Move budget step by step. Doubling a bid changes the auction you buy in, and yesterday's ROI may not survive it.
• Split a winning segment into its own campaign or stream so you can manage it apart.

A weekly reporting routine

1. Reconcile last week's revenue with the networks.
2. Review campaigns by profit, then the top three losers by placement.
3. Update blacklists and whitelists at the sources.
4. Check running tests against their planned size.
5. Write down what you changed and why, so next week's numbers have context.

Scheduled reports by email or a shared report link for the team make the routine stick: the numbers arrive whether or not someone remembers to open the dashboard.