Solutions

Delivery shaped around the screen it lands on.

Mobile, CTV, OTT and video put different demands on an ad server. Below are the workflows AdStack is built to handle, and what each one involves in practice.

Mobile

Mobile apps: change delivery without shipping a build.

In-app inventory is the most fragmented of the four. Many app versions are in the field at once, networks vary, and a release can take days to reach users. Keeping decisioning and tracking on the server side means a change to weighting, geography or creative is a configuration change, not an app update.

AdStack serves in-app video and display placements for app developers and publishers, and handles requests coming straight from an app or from a mediation layer.

A smartphone playing a video ad

What a typical set-up covers

  • Ad requests from the app or its mediation layer
  • Server-side decisions on IP, geography and weighting
  • Creative served from hosted CDN storage or passed to a wrapper
  • Signed impression and event tracking

Connected TV

CTV: predictable creative on a screen that cannot buffer politely.

Viewers on a television sit back, use a remote and watch long ad pods. A creative that stalls is far more visible here than on a phone. AdStack is built to support CTV environments, including device and publisher workflows, with creative delivered from CDN infrastructure so files load from a nearby location.

It supports publisher and platform implementations across a broad range of connected-TV environments. How a given device or app runtime behaves still depends on the publisher's implementation, which is why we scope that early.

A smart TV with a remote and a progress bar

What a typical set-up covers

  • Requests from a publisher's CTV app or player
  • Pod-friendly creative delivery from the CDN
  • IP-based geography and policy checks
  • Independent trackers for delivery and quality

Device-level behaviour depends on the app runtime and the publisher's own implementation.

OTT

OTT: a layer beside your player and ad server, not a replacement.

Most streaming services already have a player, an ad server and a set of demand partners. Ripping that out is rarely the right answer. AdStack takes the requests you route to it, applies its delivery logic, and returns VAST-based responses with tracking attached.

It is designed for compatibility with common ad-tech integrations and wrappers, so it can take a particular job, such as controlled allocation or first-party hosting, without disturbing the rest of the chain.

A streaming box beside a TV

What a typical set-up covers

  • Your existing player and ad-server workflow stays in place
  • Third-party VAST wrappers passed through and tracked
  • Weighted allocation across campaigns or sources
  • Event records you can reconcile against your own

Digital video

Digital video: VAST in, VAST out, with the tracking intact.

For web and in-app video players, delivery follows VAST. AdStack serves VAST responses, supports third-party wrappers and external ad-serving workflows, and records start, quartile and completion events so the viewing journey of an ad can be reviewed afterwards. The mechanics are on the platform page.

Wrapper chains are where tracking tends to fray. Keeping an independent tracker alongside the chain gives you a record that does not depend on every hop reporting correctly.

A video player window with a timeline

What a typical set-up covers

  • VAST tag requests from web or in-app players
  • First-party creative or third-party wrapper responses
  • Start, progress and completion event tracking
  • An independent record beside the wrapper chain
  • Third-party trackers fired alongside

CDN delivery

First-party hosting, delivered through CDN infrastructure worldwide.

AdStack can host your ad content itself and deliver it through CDN infrastructure worldwide. Creative is stored once and served from the location nearest the viewer. You control the files, the paths and the replacement process instead of relying on someone else's host.

Hosting first-party is optional. If a campaign comes through a third party, the same delivery path can pass a wrapper through untouched.

An origin server connected to edge servers around a globe

What a typical set-up covers

  • Hosted creative with stable, absolute URLs
  • CDN delivery to viewers worldwide
  • Replacement and rotation without changing tags
  • Pass-through for third-party wrappers

Instant ad serving

Served from the nearest PoP, so the ad starts before the viewer notices.

An ad that takes two seconds to begin is an ad many viewers have already left. The fix is rarely a faster server. It is a shorter distance.

A large hub node and smaller nodes sending light to a phone and a TV
  1. Request and decision

    The player asks for an ad. AdStack applies geography and allocation rules and returns a VAST response whose creative URL points at the CDN.

  2. Routed to a nearby PoP

    The creative request is steered to a point of presence close to the viewer: a smaller edge PoP in or near their city where one exists, or a major PoP in the nearest hub.

  3. Cache hit

    If the file is already held there, it starts streaming at once. No long-haul trip to the origin, no waiting on a distant server.

  4. Cache miss

    The PoP asks a larger upstream PoP. If that one also lacks the file, a single request goes to the origin. Each PoP on the way keeps a copy, so the next viewer nearby gets a hit.

Smaller PoPs

They sit closer to viewers, often inside regional networks, which gives the shortest path to the screen. Each holds less content, so a file that is rarely requested may be fetched from one level up.

Major PoPs

Larger hubs with more capacity and a deeper cache. They absorb misses from the smaller PoPs around them and keep repeat requests away from the origin. Exact coverage and tiering depend on the CDN in use.

Blue fibre optic strands radiating outwards in the dark
Shorter distance, faster start.

Video streaming

Video that starts on the first chunk, not the last.

For video, the creative is delivered in a streaming-friendly form: either a progressive file or short segments over HLS or DASH. Playback begins once the first segment arrives instead of after a full download. Which form is used depends on the player and is agreed during scoping.

How quartile tracking works

What streaming delivery covers

  • Progressive files or short segments, depending on the player
  • Several bitrates, so the player can match the viewer's connection
  • Files requested ahead of launch where the CDN supports it, so the first viewers hit cache
  • The same signed trackers whichever delivery form is used

Decisioning

Decisioning: who sees what, and from where.

Before an ad goes out, AdStack looks at the request. MaxMind IP intelligence supports IP detection and geographic or policy decisions. Weighted allocation divides delivery between campaigns in the proportions you set. Fraud signals can limit or flag requests that look wrong.

Each control is configurable per integration, and the logic is described in detail on the platform page.

How the platform decides

A routing switch with three branching paths and a shield

What a typical set-up covers

  • IP detection and geographic rules
  • Weighted split between campaigns
  • Fraud-signal checks at request time
  • Policy outcomes recorded with the event

Who it fits

Different teams use the same layer for different jobs.

Publishers and OTT platforms
Run delivery and tracking beside an existing player and ad server, with first-party hosting where it helps.
App developers
Serve in-app placements and adjust delivery server-side, without waiting for release cycles.
CTV operators
Deliver creative reliably to connected-TV environments and keep a verifiable record of what played.
Ad networks
Divide delivery between sources in set proportions and reconcile against an independent tracker.
Advertisers
Place campaigns on chosen screens and regions, then inspect signed delivery and event data.
Ad-tech teams
Take a specific component, such as signing, hosting or geography checks, and fit it into your own pipeline.

How an engagement runs

From first conversation to live traffic.

The steps are short on purpose. Most of the effort sits in mapping your current ad stack correctly, not in the build.

  1. Scope

    We agree the environment, current ad stack, expected monthly ad request range and what has to stay untouched.

  2. Map the workflow

    Which requests route to AdStack, where creative is hosted, which trackers fire and who reconciles them.

  3. Test on a limited path

    Run a controlled slice of traffic, check signatures and counts against your own records.

  4. Go live and watch

    Widen the traffic, with fraud signals and tracker comparisons reviewed along the way.

Not sure which workflow matches your set-up?

Describe your environment and we will tell you plainly whether, and how, AdStack fits.

WhatsApp us