“Proprietary bidder” is one of those phrases that appears on every adtech website and gets explained on almost none of them.
It sounds like a technical detail. It is actually the single biggest determinant of what a programmatic company can and cannot do for its partners. Everything downstream, the pricing model, the optimisation logic, the transparency, the speed at which anything can change, is decided by whether a company owns the thing that makes the bidding decision or rents it.
This is the full explanation. What a bidder is, what it does in the 100 milliseconds it has, what “proprietary” changes, and what to ask before you believe anyone’s claim about it.
What a bidder actually is
A bidder is a server that answers one question, several hundred thousand times per second: what is this specific impression worth to me right now, and should I bid?
When a user opens an app, the SSP or exchange packages what it knows about that moment into an OpenRTB bid request: the app’s bundle ID, the ad slot’s size and format, the device, the geo, the operating system, whatever identifiers and consent signals are available, the floor price. That request goes out to every connected buyer.
Each buyer’s bidder receives it, evaluates it, and returns a bid response with a price and a creative, or returns nothing. The exchange runs the auction. The winner’s ad renders.
The entire round trip has to complete in roughly 100 to 150 milliseconds, including network transit. The bidder’s own decision window is a fraction of that, often 30 to 60 milliseconds. Miss the deadline and your bid is not late, it is absent. You do not get told you lost. You simply were not in the auction.
So a bidder is not one thing. It is four things running against a stopwatch:
Step 3 is where the money is made or lost. Everything else is plumbing that has to be excellent.
What “proprietary” means, precisely
It means the company wrote and operates its own bidder rather than licensing someone else’s.
That is a bigger distinction than it sounds, because the licensing model is genuinely reasonable. Building a bidder is slow, expensive, and produces no revenue for a long time. Licensing a white-label platform gets you to market in months. Most companies take that path, and for good reason.
The difference is what you control afterward.
Optimisation logic. With a licensed bidder, you configure. You do not decide. If your partner needs bidding tuned toward a downstream event the platform does not model, you file a feature request and wait. With your own bidder, you write it.
The data. Bid landscapes, win rates, creative response, retention signals. This is the raw material for every model you will ever build. On a licensed platform it is generated by your traffic and lives in someone else’s infrastructure, on their retention schedule, subject to their export policy. You cannot train on what you cannot hold.
The economics. White-label fees are usually a percentage of media spend. At $100k a month, 5 percent is $5,000 and nobody cares. At $5m a month, it is $250,000 a month for infrastructure you will never own. The fee scales with your success and buys you nothing new.
The roadmap. Your platform provider improves their product for their business, on their timeline. Your competitive differentiation is capped at whatever they ship.
Why it matters more in mobile than anywhere else
Generic bidders are built to serve everything: web display, video, CTV, in-app. Mobile in-app has characteristics that a general-purpose bidder averages away.
Signal is different. There is no cookie. Attribution runs through an MMP with a delay measured in hours or days, which means the feedback loop that trains your model is fundamentally lagged in a way web display is not. Your bidder has to learn from evidence that arrives late.
Session context is different. An interstitial between levels of a game is a different psychological moment from a banner on a news page, and the value gradient across placements inside a single app is steep.
Fraud surface is different. Click injection and SDK spoofing have no web equivalent.
A bidder tuned for these specifics beats a bidder configured for them. That is not a marketing claim, it is the difference between a model trained on your distribution and a model trained on everyone’s.
The part nobody advertises: it is hard
A proprietary bidder means owning latency budgets, capacity planning for traffic that spikes without warning, model retraining pipelines, and the unglamorous reality that at high QPS you are paying for bandwidth on requests you will never bid on. It means an engineering team on call for infrastructure that must not go down, because a bidder that is down is a bidder that silently stops appearing in auctions.
It also means that when the IAB Tech Lab updates the OpenRTB spec, you implement it. Nobody does it for you.
This is why “we built our own” is worth verifying rather than accepting. The phrase is used loosely, and a thin layer of custom logic on top of a licensed core is not the same thing.
How to verify the claim
Five questions. The answers are diagnostic.
Where AdSpin sits
AdSpin built its exchange and its oRTB bidder in-house. Not because building is romantic, but because we could see the ceiling. Every capability we wanted to offer eventually ran into something a licensed platform would not let us change.
Our bidder is tuned for mobile in-app specifically. It is not a general-purpose engine licensed to hundreds of companies with our logo on it. It runs against supply we own and operate and against direct publisher integrations, which means the signal reaching the model is clean before the model ever sees it. Owning both ends of that, the supply and the bidding, is what makes full supply-path visibility possible rather than aspirational.
When a partner needs optimisation toward a specific event, an LTV signal from their MMP, a particular audience definition, we build it. We are not waiting for a platform update.
The question worth asking is not “what can this platform do today?” It is “what will I be able to change in two years, when my scale is ten times what it is now?”
White-label gets you to market. Proprietary is what lets you compete once you are there.
Explore how AdSpin’s proprietary exchange and oRTB bidder power transparent mobile programmatic at adspin.io.
