You set the scope.
Choose the settings and traffic you’re prepared to change.
Find missed opportunities in your auctions. Put AI to work on the analysis. Evaluate changes with the evidence in view.
14 days free after activation. No credit card to get started.
See pricing and trial details.
Workflow preview · Unattended loop in development.
Responses, timeouts, ad renders, and the settings that actually reached the page.
Built by the Optimize team at BuySellAds.
Working with publishers since 2008.

Review bidder health, late bids, and refresh behavior in one place. Spend less time piecing the auction together.
Requests sent Responses received
See which bidders answer, which contribute, and which deserve a closer look.
AutoResearch connects investigation, controlled testing, and learning. Here’s how a late-bid finding moves through the loop.
Analyze late-bid value before deciding whether a longer timeout deserves a test.
Use a limited test when the auction data alone cannot answer.
Record what happened, then revisit the tradeoff as demand changes.
Inspired by Andrej Karpathy’s autoresearch on GitHub .
Keep your bidders, ad server, and demand relationships. Replace the URL of the Prebid script you already load.
your-prebid.jss.bidtune.net/ YOUR_SITE_ID/p.jsExample URLs. Setup review confirms compatibility and the installation details for your site.
Choose the settings and traffic you’re prepared to change.
Follow what applied, what happened, and what remains uncertain.
Pause the work or undo a change when you need to.
Illustrations of publisher controls
Continue with Google, then submit your site for setup review.
Your free trial starts when your site is activated and collecting data.
Start in observation mode. Explore the work before deciding what to change.
Review bidder activity, response timing, and observed ad delivery as data arrives.
Your starting point: a view of your own traffic.Run on-demand research and review generated proposals alongside the supporting data.
Your next step: an idea with evidence to inspect.Use experiment tools for supported settings, with traffic limits, change history, and undo controls.
Your readout: what applied and what happened.The full unattended research loop is in development. Evidence builds with your traffic, so some tests need longer than the trial to establish an effect.
14 days free after activation. No credit card to get started.
See pricing and trial details. Setup review comes first.