What is an ad network?

Robin Lim

An ad network is a company that collects ad space from many publishers, such as websites, mobile apps and AI apps, and sells it to advertisers in one place. The network matches each ad to the right placement, serves it, tracks the results, and pays the publisher a share of what the advertiser spends.

Key takeaways

  • Ad networks sit between publishers, who have audiences, and advertisers, who want to reach them.
  • Advertisers usually pay per 1,000 impressions (CPM), per click (CPC) or per completed action (CPA).
  • Publishers earn a share of that spend without having to sell ads to each brand themselves.
  • Networks differ by channel (web, mobile, native, AI apps), by audience (general or one industry) and by how they price (CPM, CPC, CPA)
  • Ad networks for AI apps are a newer type that places a labeled recommendation next to an AI assistant's answer.

How does an ad network work?

An ad network does five jobs that a publisher and an advertiser would otherwise have to handle on their own.

  1. It gathers inventory. Publishers add the network's code or SDK to their site or app, which tells the network where ads can appear and who is likely to see them.
  2. It signs up advertisers. Brands and agencies set a budget, a target audience or context, a pricing model and their ad creative.
  3. It matches ads to placements. When a page loads or a user takes an action, the network picks the ad that best fits the placement and the advertiser's goals.
  4. It serves and measures the ad. The network shows the ad and records impressions, clicks and conversions.
  5. It pays the publisher. The network bills the advertiser, keeps a margin, and pays the publisher the rest on a regular schedule.

How do advertisers pay on an ad network?

Pricing modelWhat the advertiser pays forBest for
CPM (cost per mille)Every 1,000 times the ad is shownReach and brand awareness
CPC (cost per click)Each click on the adWebsite or app traffic
CPA (cost per action)Each completed action, such as a purchase or sign-upSales and lead generation
CPV (cost per view)Each view of a video ad, usually after a set number of secondsVideo campaigns

Publishers often compare networks using eCPM, the effective revenue they earn per 1,000 impressions across all pricing models. It lets a publisher compare a CPC campaign and a CPM campaign on the same scale.

What are the main types of ad networks?

TypeWhat it doesExamples
General (horizontal)Sells inventory across many sites and topics to reach a broad audienceGoogle AdSense, Media.net
VerticalFocuses on one industry, such as travel, finance or gamingIndustry-specific networks
PremiumSells inventory from a selected group of high-quality publishers at higher pricesNetworks run by large media groups
Mobile appPlaces ads inside mobile apps, often as banners, interstitials or rewarded videoGoogle AdMob, AppLovin, Unity Ads
Native and content recommendationShows ads styled to match the surrounding content, often as "recommended" linksTaboola, Outbrain
Affiliate and CPAPays publishers only when a referred user completes a sale or sign-upCJ, Rakuten Advertising, Impact
AI app ad networksPlaces a paid recommendation next to an AI assistant's answer, matched to what the user askedStartups such as Aria Ads

What is the difference between an ad network, an ad exchange, an SSP and a DSP?

These four terms describe different parts of the advertising supply chain, and they are often confused.

TermWho uses itWhat it does
Ad networkPublishers and advertisersPackages inventory from many publishers and resells it to advertisers, usually at prices it sets
Ad exchangeNetworks, SSPs and DSPsRuns an open marketplace where inventory is bought and sold in real-time auctions
Supply-side platform (SSP)PublishersSoftware that offers a publisher's inventory to many buyers at once to raise the price
Demand-side platform (DSP)Advertisers and agenciesSoftware that buys inventory across many exchanges and networks from one dashboard

For a fuller comparison, see ad network vs ad exchange and what is a supply-side platform.

What are the pros and cons of using an ad network?

For publishers, an ad network brings in advertisers without a sales team, fills unsold inventory, and handles billing and payment. The trade-offs are a smaller share of each dollar, less control over which ads appear, and a dependence on the network's rules and payout schedule.

For advertisers, a network gives access to many publishers through one contract and one dashboard, with targeting and reporting built in. The trade-offs are less visibility into exactly where ads run on some networks, and exposure to invalid traffic if the network does not filter it well.

How do you choose an ad network?

  • Fit with your format: a network built for web banners will not suit a mobile game or an AI chat app.
  • Demand: ask how many active advertisers the network has in your categories and what fill rate to expect.
  • Pricing and payouts: check the revenue share, payment terms and minimum payout.
  • Controls: look for the ability to block categories and brands and to cap how many ads appear.
  • Traffic quality: ask how the network detects and filters bots and invalid clicks.
  • Privacy: check what user data the network collects and whether it shares any with advertisers.

What is an ad network for AI apps?

An ad network for AI apps places ads inside AI assistants, chatbots and agents built by independent developers. Instead of a banner, the ad is usually a single sponsored recommendation that appears next to the assistant's answer when the user's question relates to what a brand sells. Because people describe what they want in full sentences, these networks can match ads to stated intent instead of cookies or browsing history.

The largest assistants are building their own ad systems. ChatGPT began showing ads to US users on its Free and Go tiers in February 2026, and Google shows ads in AI Overviews and AI Mode. Most other AI apps do not have the scale to run their own ad business, which is why ad networks for AI apps have appeared. Aria Ads is one of them: it matches each prompt to a relevant brand, labels every placement as sponsored, and lets advertisers buy on CPM, CPC or CPA.

Four things set these networks apart from web and mobile networks:

  • Ad Format: Ads need to feel native to the user experience without breaking user trust.
  • Latency: the ad request has to run alongside the model's answer without slowing it down.
  • Disclosure: the ad must be clearly separate from the assistant's answer so users can tell the two apart.
  • Privacy: the network needs the context of the conversation to match an ad, so publishers should check that conversations are never shared with advertisers.

FAQ

What is an ad network in simple terms?

An ad network is a middleman between websites or apps that have space for ads and brands that want to buy that space. It collects the space from many publishers, sells it to advertisers, shows the ads, and pays each publisher a share of the money advertisers spend.

How do ad networks make money?

Ad networks keep a share of what advertisers pay and pass the rest to publishers. The share varies by network and is set out in each network's publisher terms. Some networks also charge advertisers fees for targeting, data or managed campaigns.

What is the difference between an ad network and an ad exchange?

An ad network buys inventory from publishers and resells it to advertisers, usually at prices it sets. An ad exchange is an open marketplace where buyers and sellers trade inventory in real-time auctions. Many networks now buy and sell through exchanges as well.

What is an ad network for AI apps?

An ad network for AI apps places sponsored recommendations inside AI assistants, chatbots and agents. It reads the context of a conversation, matches it to a relevant brand, and shows one labeled recommendation next to the answer. Developers earn a share of what advertisers spend, and advertisers reach people at the moment they ask for advice.

Are ad networks safe for publishers?

Most established ad networks are safe when publishers check their terms. Look for clear revenue share and payout terms, controls to block categories and brands, filtering of invalid traffic, and a privacy policy that limits what user data is collected and shared with advertisers.