The SPO Mini Exercise: One DV360 Report to Clean Your Supply Path in 30 Minutes

by

Supply Path Optimisation (SPO) doesn’t need to be complicated – especially if you’re running campaigns in DV360 and you just want a practical hygiene routine.

In SEA, one of the most common blind spots isn’t “fraud dashboards” or fancy supply-chain diagnostics. It’s the long tail of low-volume inventory that doesn’t surface clearly in reporting – which effectively becomes “unknown” in the sense that you can’t explain exactly where a portion of spend went.

This post gives you a 30-minute SPO mini exercise using one DV360 report. The goal is simple:

  • Understand your main supply paths (exchanges)
  • Audit your top disclosed placements (App/URL)
  • Quantify the long tail (low-volume inventory that’s not meaningfully disclosed)

What you need: One DV360 report

DV360 report setup (single report)

Dimensions (columns):

  • Environment (to split Web vs App)
  • Exchange
  • App/URL (contains both sites and apps)

Metrics (columns):

  • Impressions
  • Media Cost (Advertiser currency)
  • (Optional) Clicks, CTR
  • (Optional) Conversions / CPA (if you have Floodlight)

Date range: last 14 days (quick signal) or last full month (more stable).

Why include Environment? Because App/URL includes both websites and apps. Environment is the cleanest way to filter your App/URL rows into Web vs App views without needing separate reports.

The 30-minute SPO mini exercise (3 checks)

Check 1: Quantify your “low-volume inventory” long tail

This is the core of the exercise. If a meaningful share of spend sits outside your top disclosed placements, your plan becomes harder to explain and harder to optimise.

How:

  1. Sort the report by Media Cost (descending).
  2. Filter to Environment = Web and note the share of spend captured by the Top 20 App/URL rows.
  3. Repeat for Environment = App and note the share of spend captured by the Top 20 App/URL rows.

What you’re measuring: How much of your spend is in a long tail of low-volume inventory (i.e., not meaningfully “disclosed” when you review placements).

Rule of thumb: If your Top 20 App/URL rows capture less than 70% of spend for Web or App, you likely have a large long tail. That doesn’t automatically mean it’s “bad” – it means it’s harder to control, explain, and optimise.

Actions:

  • If long tail is high, tighten targeting and inventory controls to improve concentration into disclosed placements.
  • Consider consolidating supply sources (next check) to reduce accidental fragmentation.

Check 2: Exchange concentration (fragmentation and accidental dependency)

Pivot/group by: Exchange • Sort: Media Cost (descending)

What “healthy” often looks like: your top 3 exchanges represent roughly ~60%+ of spend because you chose that on purpose.

Caveat (important): If one exchange is taking ~90%+ of spend and it’s not intentional, that’s also a problem. You may be accidentally over-dependent on one path (pricing leverage, delivery quirks, blind spots, and less flexibility when something breaks).

Questions:

  • Are we too fragmented (spend scattered across too many exchanges)?
  • Or are we accidentally over-dependent (one exchange dominates without a clear reason)?

Actions:

  • If fragmented: consolidate intentionally to fewer exchanges you trust.
  • If over-dependent: test a second path on purpose (even 10–20%) so you have benchmarks and optionality.

Check 3: Placement sanity (focus on what you can see)

Filter: Environment = Web (review top App/URL rows), then Environment = App (review top App/URL rows).

Questions:

  • Do the top App/URL placements look legitimate and relevant?
  • Are any top placements “why are we here?” candidates taking meaningful spend?

Actions:

  • Exclude obvious junk / irrelevant top placements.
  • Use exclusions and tighter targeting to reduce spillover into the long tail.

The only 3 actions you need (keep it simple)

  1. Reduce long tail exposure (improve Top 20 share for Web + App)
  2. Balance exchange concentration (avoid both fragmentation and accidental dependency)
  3. Exclude bad top placements that you can clearly identify

Pro tip: Don’t try to “solve SPO” in one day. Run this exercise monthly. Your goal is steady improvement: more spend in placements you can name, fewer surprise long-tail pockets.

Want to take this further? Outcome optimisation with Custom Bidding

Once you’ve reduced obvious long-tail spill and tightened your main supply paths, the next challenge is: how do you optimise more effectively when you can’t manually review every placement?

This is where Custom Bidding can help. Instead of optimising to a single default metric, you can tailor bidding logic toward the outcomes you care about (for example: stronger post-click behaviour, higher-quality conversions, or value-weighted actions).

If you want support: reach out and I can help you think through a Custom Bidding approach that fits your measurement setup and campaign goals – especially for SEA campaigns where signals and inventory quality can vary a lot by market.

Final checklist (copy/paste)

  • ✅ Run the DV360 report (Environment + Exchange + App/URL)
  • ✅ Check your Top 20 share for Web and App
  • ✅ Review exchange concentration: top 3 should be meaningful, but one exchange dominating unintentionally is also a red flag
  • ✅ Review top App/URL rows → exclude obvious junk
  • ✅ Re-run in 2–4 weeks → track whether the long tail shrinks