Adlattice
how ads get there

Delivery models

Same lifecycle, three different shapes. Which one a publisher runs decides who gets first look and where the latency goes.

Waterfall vs Header Bidding

Waterfall

option a

Publisher tries SSPs one at a time. The first that fills above the floor wins.

Pros
  • · Simple to reason about
  • · Cheap to operate
Trade-offs
  • · Sequential calls = latency
  • · Hidden price ceiling
  • · Top SSP gets first look

Header Bidding

option b

All SSPs are asked in parallel from the page header. Winning bid is passed to the ad server as a competing line item.

Pros
  • · All SSPs see all impressions
  • · Higher clearing prices
  • · Price discovery is real
Trade-offs
  • · Client weight (Prebid.js)
  • · Timeout management
  • · Complex setup

CSAI vs SSAI

CSAI

option a

Client-Side Ad Insertion: the player fetches content and ads as separate streams. The classic web pattern.

Pros
  • · VPAID & client interactivity
  • · Easy to track client-side
  • · Standard player SDKs
Trade-offs
  • · Visible buffering at break
  • · Ad blockers can stop the call
  • · Inconsistent on CTV

SSAI

option b

Server-Side Ad Insertion: a service stitches ad segments into the HLS/DASH manifest. The player sees one stream.

Pros
  • · Seamless playback
  • · Bypasses most ad blockers
  • · Works on every CTV
Trade-offs
  • · No VPAID
  • · Tracking through proxy
  • · Manifest manipulation cost

Open auction vs PMP vs PG

Open auction

option a

Anyone can bid. Lowest floor, highest density. Default for unsold inventory.

Pros
  • · Maximum demand
  • · Best price discovery
Trade-offs
  • · Brand safety risk
  • · Lowest priority in ad-server hierarchy

PMP / PG

option b

PMP: invite-only auction with a deal ID. PG: fixed price + fixed volume, no auction — closest thing to a direct buy in programmatic.

Pros
  • · Curated buyers
  • · Fixed price (PG) or floor (PMP)
  • · Higher priority than open
Trade-offs
  • · Lower fill if buyer pulls back
  • · Deal-ID plumbing per partner