Skip to main content
Metrics type: Key Metrics • Category: Ad Platform

​
At a glance

Pacing check on TTD budgets vs period elapsed. Koa’s auto-bid is well-calibrated (less prone to runaway ramps than peer DSP bidders), but CTV peak hours and DOOH inventory deals can still drive concentrated overpacing. TTD also offers Predictive Pacing (auto-throttle when forecast end-of-period spend exceeds budget), which reduces but doesn’t eliminate pacing alerts.

​
Calculation

Calculated automatically from your The Trade Desk data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.

​
Worked example

The same enterprise homeware brand. Period 01 Apr 26 to 30 Apr 26. Today 18 Apr 26 (60% elapsed). What’s interesting:
  1. Account at 70% pacing vs 60% elapsed = 10pp over, healthy band. Predictive Pacing is doing its job; without it, the NBA playoffs CTV CPM rise would have driven 80, 85% pacing without any human intervention.
  2. CTV Premium PMP at 71% with NBA playoffs running is a quiet success: Predictive Pacing detected the seasonal CPM spike and throttled delivery to maintain monthly cap discipline. Compare to MediaMath worked example (no auto-throttle), where the same NBA event would have driven 83% pacing.
  3. DOOH at 70% has Predictive Pacing disabled by design; DOOH inventory delivery is lumpy (book a billboard for a week, deliver in concentrated bursts) so auto-throttle would create campaign-level disruption. Manual budget management is correct for DOOH.
  4. No campaigns in alert band, the account is well-managed. This is the typical TTD shape; Predictive Pacing prevents most pacing alerts before they escalate.
  5. Compared to peer DSPs: TTD’s Predictive Pacing reduces pacing alert frequency by 60, 80% vs StackAdapt or MediaMath running similar campaigns. The operational quality-of-life difference is meaningful for enterprise teams.
Quick sanity tests:
  • Pacing ≈ time-elapsed ± 5pp = healthy.
  • Pacing > time-elapsed by 10, 15pp = mild over; Predictive Pacing handling.
  • Pacing > time-elapsed by 20+pp = Predictive Pacing isn’t enabled on the campaign (check) OR something happened that overrode it.
  • DOOH overpacing is normal (lumpy delivery); judge across the full month, not intra-week.
  • Pacing < time-elapsed by 15+pp = inventory shortage or bid below floor; raise CPM.

​
Sibling cards merchants should reference together

​
Reconciling against the vendor’s own dashboard

Where to look in TTD: TTD Platform → Campaigns and check the per-campaign budget, pacing indicator, and Predictive Pacing status. TTD’s pacing UI is more polished than peer DSPs, includes forecast end-of-period spend natively. Why our number may legitimately differ from TTD: Cross-connector reconciliation:

​
Known limitations / merchant FAQs

Why am I getting a pacing alert when I have Predictive Pacing enabled? Predictive Pacing reduces but doesn’t eliminate alerts. It throttles within a campaign’s daily budget but doesn’t override hard delivery floors. Common reasons for alerts despite Predictive Pacing: (1) the campaign has a daily floor that prevents adequate throttling; (2) PMP deal commitments require minimum daily delivery; (3) seasonal CPM rise overwhelmed the model’s forecasts; (4) campaign was launched mid-period and Predictive Pacing hasn’t accumulated enough data to forecast accurately. Should I enable Predictive Pacing on every campaign? Almost yes. Exceptions: (1) DOOH where lumpy delivery is intrinsic; (2) PMP deals with minimum-delivery commitments; (3) test campaigns where you want manual control during learning; (4) very short flight (<7 days) campaigns where the forecast window is too short. Why is TTD pacing so much better-behaved than MediaMath? Two reasons: (1) Predictive Pacing actively manages delivery shape, MediaMath has no equivalent; (2) TTD’s reporting infrastructure is faster, so pacing decisions are made on more recent data. The operational difference is one of the harder-to-quantify reasons enterprise advertisers prefer TTD. My DOOH is overpacing, what should I do? Probably nothing. DOOH delivery is lumpy by inventory nature; book a billboard for a week, deliver in concentrated bursts. Judge DOOH pacing across the full month, not intra-week. If end-of-period total is on plan, mid-period overpacing is fine. Can TTD outrun a hard daily cap? Hard caps are enforced; TTD won’t deliver above the daily cap (unlike legacy MediaMath, which had occasional cap-bypass issues). This makes daily caps a reliable backstop on TTD specifically. My pacing dropped sharply, what happened? Common causes: (1) Predictive Pacing aggressively throttling because it forecasts overspend; (2) PMP deal lapsed; (3) inventory partner pulled supply; (4) creative disapproved; (5) bid below floor on key inventory. How does this card handle Q4 / Black Friday peak? Predictive Pacing helps but doesn’t replace seasonal budget planning. For Q4, increase budgets 30, 60% to absorb the CPM rise. TTD specifically benefits from advance budget commitment because Predictive Pacing has more data to plan against. Can I trust pacing on day 1? TTD’s faster ingest makes day 1 more reliable than peer DSPs but still noisy. By day 3, pacing is the actionable read. What’s TTD’s monthly close cadence? Most enterprise contracts close on calendar month. TTD’s invoice typically arrives 5, 10 days after month-end with breakdowns by advertiser, campaign, and inventory source. Reconcile this card’s spent figure against the invoice for the canonical monthly number. Should I set seat-level budgets? For multi-advertiser organisations, yes. Seat-level budgets prevent any single advertiser from exhausting the parent’s budget. This card is per-advertiser; aggregate seat-level pacing requires the seat-level view in TTD UI.

​
Tracked live in Vortex IQ Nerve Centre

Spend vs Budget is one of hundreds of KPI pulses Vortex IQ tracks across The Trade Desk and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.