Real-time bidding, or RTB, is the process by which digital ad impressions are auctioned and sold one at a time as a page loads.
Real-time bidding, or RTB, is the process by which individual digital ad impressions are auctioned and sold one at a time in the instant a page or app loads. Rather than buying blocks of impressions in advance, advertisers bid on each impression as it becomes available, evaluating who the user is and whether the impression is worth their money, all within milliseconds. It is the auction engine that lets programmatic advertising match specific ads to specific users at scale.
The mechanics unfold in the moment a user opens a page. The publisher's system sends a bid request, containing information about the impression and the user, out to an ad exchange. Advertisers' demand-side platforms receive that request, assess it against their targeting and value rules, and respond with a bid. The exchange runs an auction, selects the winning bid, and the corresponding ad is served, all before the page finishes rendering. The entire cycle typically completes in well under a second. Because each impression is evaluated on its own merits, buyers can pay more for users who fit their ideal profile and less, or nothing, for those who do not.
The term joins "real time," meaning instantaneous, with "bidding." It describes the millisecond auctions that powered programmatic advertising from the early 2010s onward, when the infrastructure to evaluate and transact impressions individually became fast and widespread enough to handle enormous volumes. Real-time bidding turned digital advertising from a business of pre-negotiated placements into a continuous, per-impression marketplace.
For a business, real-time bidding matters because it enables precise, efficient spending. Instead of paying a flat rate to appear on a site regardless of who sees the ad, you pay according to the estimated value of each specific impression and audience. This means budget flows toward the users most likely to convert and away from those who are not worth reaching, which can dramatically improve efficiency compared with bulk buying. It also gives advertisers the flexibility to adjust bids and targeting continuously as performance data comes in, rather than being locked into a fixed buy.
The nuances and risks deserve attention. Because bidding is automated and lightning fast, it is only as good as the data and rules behind it; poor targeting logic can spend heavily on worthless impressions. The open marketplace also exposes buyers to invalid traffic, non-viewable placements, and brand-unsafe environments unless verification and controls are in place. There is a layered supply chain between buyer and publisher, and fees along that chain can erode the share of spend that reaches genuine inventory. Advertisers sometimes assume that bidding on impressions guarantees they were seen, when viewability is a separate question entirely. Real-time bidding is best understood alongside the systems it connects: demand-side platforms that place the bids, ad exchanges that run the auctions, and the broader programmatic framework that gives it purpose. Run with sound data and proper safeguards, it is a powerful way to buy attention precisely.
Real-time bidding lets advertisers pay for each impression based on its individual worth, rather than buying blocks of inventory blind. That precision is what makes programmatic efficient.