Four mobile gaming publishers recorded revenue growth of as much as 14 times through Amazon Publisher Services during a beta of the new APS mobile Software Development Kit bridge, according to material published by InMobi and Amazon Publisher Services. The gains came from inventory the publishers were already selling, with no changes made to their mediation setups.

The figure sits behind one of six capabilities Amazon Publisher Services introduced at its annual summit on Thursday, May 28, 2026. The bridge itself was described at that event in fairly compact terms: a mechanism allowing third-party bidders to compete for mobile app inventory using on-device auction transparency, signals and rendering. The beta numbers, published separately in co-branded material from InMobi and APS, put a commercial figure on what that mechanism did for a small group of app developers.

Numbers of this shape deserve careful reading. Fourteen times is presented as a ceiling rather than an average, the sample is four publishers, all of them in gaming, and no baseline revenue, absolute dollar amount or observation window is disclosed. What the material does specify is the condition attached to the result: growth arrived on existing inventory, without publishers reconfiguring their mediation stacks. That condition is the part app monetisation teams are likely to examine most closely, because reconfiguration cost is precisely what has historically slowed the adoption of new demand paths inside mobile applications.

What the bridge changes inside the app

Mobile app publishers have long accessed programmatic demand through a stack of bidder SDKs, each integrated separately into the application binary. According to Amazon Publisher Services, those SDKs once ran independently of one another, which limited visibility and made it harder for demand sources to compete on equal terms. The structural consequence is familiar to anyone who has run an app monetisation operation: each bidder sees a partial picture of the auction, rendering behaviour varies by partner, and adding a new demand source means shipping a new app version.

The APS mobile SDK bridge addresses the second and third of those problems. According to the co-branded InMobi and APS material, the bridge activates native SDK rendering, which allows publishers to unlock rich ad formats from dozens of new demand partners and hundreds of advertisers through InMobi. Rendering is the operative word. A creative served through a generic web view behaves differently from one drawn by a partner's own rendering layer, and format richness has historically been the trade-off publishers accepted when they routed demand through server-side paths rather than through a bidder's own SDK.

According to Amazon Publisher Services, the bridge gives publishers access to more than 1,000 advertisers and 28 additional bidders without additional integration work. PPC Land reported the full slate of APS Summit capabilities on May 28, 2026, including the Signal IQ expansion to the full OpenRTB bidstream, the Publisher Supply AI assistant scheduled for open beta over the summer, and the extension of Shopping Insights deals from streaming television to web and mobile.

The bridge enables publishers to retrieve device parameters and construct bid requests for real-time transactions, according to the company. That description matters technically. Device parameters passed at request construction time determine what a bidder can evaluate before pricing an impression, and a bidder operating with a thinner parameter set generally bids more conservatively or declines to bid at all.

The beta result and its limits

The co-branded material describes a recent beta conducted with gaming publishers. Four publishers using the new bridge saw immediate revenue growth, as much as 14 times, on existing inventory, and made no changes to their mediation setups, according to InMobi and Amazon Publisher Services.

Several things are worth separating out. The metric is APS revenue specifically, not total application revenue. A publisher whose APS channel represented a small fraction of monetisation before the bridge could post a large multiple on that channel while total revenue moved by a far smaller proportion. The material does not state where the APS channel sat in each publisher's revenue mix beforehand.

Gaming is also not a neutral test environment. Game applications run interstitial and rewarded formats at volumes and completion rates that content applications rarely match, and the format richness the bridge unlocks maps closely onto what game inventory already supports. A news application running standard banner units would be exercising a narrower slice of the same capability.

The phrase covering the sample is "as much as," which describes the best observed outcome rather than the central tendency across the four. No distribution is given.

How the InMobi integration is documented

InMobi's public support documentation for Amazon Publisher Services, last updated on July 21, 2026, describes the mechanics of the partnership in operational detail. According to that documentation, InMobi is available as a bidder on APS through Transparent Ad Marketplace (TAM) and Unified Ad Marketplace (UAM), using a server-side integration built on the OpenRTB protocol that does not require SDK implementation.

That description sits in interesting tension with the bridge announced in May. The server-side bidder path and the native rendering path are different routes to the same demand, and the documentation as of July 21, 2026 continues to describe the server-side route as the standard integration for both mobile web and in-app inventory. The bridge does not replace it. Publishers reading both documents encounter two answers to the question of whether an SDK is involved, because both answers are correct for different configurations.

Placement setup and a video format constraint

The documented configuration path runs through the InMobi Publisher Dashboard. According to the documentation, publishers navigate to Inventory and then Add Placement, select App, enter app store details, set child compliance and block settings, choose the ad format, then create a placement with Amazon TAM selected as partner and Audience Bidding switched on. The Tag ID and App ID obtained from APS are entered in the respective fields before the placement is saved.

One constraint carries practical weight for mobile publishers. According to InMobi's documentation, mobile publishers create only in-stream placements for any video ad type, including interstitial and rewarded video, because placements created with interstitial or rewarded video ad units will only work for static ads. Rewarded video is among the highest-yielding formats in mobile gaming, so the distinction between an in-stream placement and a rewarded video placement is not a cosmetic setting.

Supported formats are listed by environment. Mobile in-app inventory covers banner, interstitial, video in all formats, and native. Web inventory covers banner, in-stream video and native.

Slot mapping, account identifiers and ads.txt

The connection sequence begins with a request initiated on the APS dashboard or through an APS account manager, which appears on InMobi's partner dashboard under Supply, Publishers and In-Progress, according to the documentation. Publishers then download a slot mapping file, in which the Ad Slot Name corresponds to the Tag ID and the app ID or site ID carries across directly. Each Tag ID and App ID is mapped to the corresponding InMobi placement ID and app ID, and the documentation states that these identifiers ensure correct request routing between APS and InMobi.

Every placement requires a unique Tag ID or ad slot name, and duplicate values across placements are ruled out. According to the documentation, APS requests arriving at InMobi carry the App ID and Tag ID, which is how the corresponding InMobi placement ID is identified. Routing accuracy in this model is therefore a function of identifier hygiene at setup time rather than of anything the auction can correct later.

A third step covers account and ads.txt identifiers. The account number field takes InMobi's account ID, and the ads.txt ID field takes the InMobi seller ID, also described as the payment ID. Once mapping is complete, publishers mark the integration ready to enable and approve new apps and sites under a pending review queue.

Seller identifier declaration is not an administrative footnote in current market conditions. Authorised seller files turned ten years old in April 2026 amid proposals to replace them with a richer format, and buyers increasingly validate supply paths before bidding. Display and Video 360 began enforcing app-ads.txt for app inventory years earlier, declining to bid where a publisher ID is absent from a file that exists.

Fallback placements

For applications carrying multiple placements, the documentation describes a fallback mechanism. A fallback is created as an Audience Bidding placement for Amazon TAM with the fallback box checked, with one fallback placement permitted per app per placement type. A fallback placement does not require a Tag ID.

Where the bridge sits in the wider market

The engineering problem the bridge addresses has been attacked from several directions over the past two years, and the pattern is consistent enough to be worth naming: supply-side infrastructure providers are trying to make mobile app inventory behave more like web inventory in terms of auction mechanics and transparency.

Amazon opened its Prebid adapter in global open beta on January 22, 2026, connecting APS demand to existing Prebid.js implementations on the web side without requiring publishers to rebuild their auction frameworks. The mobile SDK bridge extends the same logic into app environments where Prebid has thinner penetration.

Six days later, a different approach reached the market. CloudX opened mobile app inventory to omnichannel demand-side platforms without SDK integration on January 28, 2026, bundling its own rendering SDK into the standard implementation so that buyers built for web and connected television could compete alongside SDK-based bidders under identical fee structures. The two designs share a premise: the rendering layer, not the demand relationship, is the bottleneck keeping buyers out of app auctions.

Open-source infrastructure has moved along the same axis. Prebid Mobile SDK 3.0 delivered CPM gains while the 4.0 roadmap targeted publisher control, with plugin support and modular architecture positioned to reduce integration friction and creative standardisation addressed as a buyer confidence problem.

The demand-side pressure has been running longer still. Display and Video 360 stopped bidding on multiple waterfall calls for mobile app inventory on October 23, 2024, making real-time bidding the primary purchase method across mediation platforms. Waterfall mechanics have been in retreat since, and every subsequent infrastructure release has assumed a unified auction as the target state.

InMobi's position across mediation platforms gives the bridge a wider reach than a single-partner integration might suggest. The company appeared among the bidding partners named in AdMob mediation updates published in July 2025, alongside ironSource, Unity, Liftoff, Mintegral, Moloco and LINE, and was among six partners added to Kochava's measurement certification programme on March 14, 2026.

Why this matters for the marketing community

For app publishers, the operative claim is not the multiple. It is the absence of engineering work. Integration fatigue is a genuine constraint on smaller studios without dedicated ad operations staff, and the calculation that governs whether a new demand source gets adopted usually turns on release cycles rather than on projected revenue. A demand expansion that requires no new SDK, no binary change and no mediation reconfiguration removes the argument that normally kills the proposal.

For buyers, the bridge widens the surface on which Amazon's demand and its 28 named third-party bidders compete. That has consequences for supply path planning. IAB Spain's first supply-side platform guide, published in April 2026, described Transparent Ad Marketplace and Google Open Bidding as server-side alternatives to Prebid.js and framed supply path optimisation as the primary structural response to fragmentation, citing benchmark data showing activated domains and applications more than doubling since 2023 while 90.3% of impressions concentrated on roughly 3,000 sites. Adding bidders to an app auction cuts against consolidation logic unless the added competition demonstrably raises clearing prices.

The commercial backdrop is a supply-side business attached to an advertising operation growing faster than its peers. Amazon's advertising services segment generated $17.2 billion in the first quarter of 2026, up 24% year on year and crossing $70 billion on a trailing twelve-month basis. APS operates inside that stack, as does the infrastructure beneath it: AWS launched a dedicated network for real-time bidding workloads in October 2025, offering single-digit millisecond latency for the transaction volumes these auctions generate.

Gaming inventory is where the beta result was produced, and gaming is a segment that has attracted disproportionate platform attention. Mobile gaming has drawn expanded advertising capabilities across major platforms through 2025 and into 2026, with rewarded video formats central to that expansion. A capability tested in gaming and then offered across app categories generally performs differently outside its test environment, and the co-branded material offers no data from non-gaming applications.

The open questions

Three gaps remain in the public record. The first is distribution: four publishers with a stated ceiling of 14 times says nothing about the median outcome or the floor. The second is duration, since immediate revenue growth measured in the days after activation can reflect a demand influx that normalises once bidders recalibrate. The third is category breadth, because rich format support is worth considerably more to a game developer running rewarded video than to a publisher whose inventory is predominantly banner.

Signal measurement may eventually close some of that gap. The Signal IQ expansion announced at the same summit moves into an A/B testing phase during the summer of 2026, projecting revenue lift in dollar terms tied to specific signal investments. Signal IQ first appeared in beta in 2024 alongside the extension of Transparent Ad Marketplace to streaming television publishers, initially covering third-party ad identifiers. Applied to mobile app inventory flowing through the bridge, the same methodology would give publishers a measured figure rather than a beta headline.

Until then, the 14 times figure functions as a ceiling observation from a four-publisher gaming beta, published by the two companies with a commercial interest in the integration. The integration mechanics, by contrast, are documented in detail and independently checkable: placement types, identifier mapping, format constraints, fallback rules. Those details, rather than the multiple, describe what the bridge actually does.

Timeline

Summary

Who: Amazon Publisher Services, the Amazon Ads division supplying monetisation infrastructure to web, mobile app and streaming television publishers, and InMobi, the Singapore-headquartered mobile advertising company operating as a third-party bidder on the APS mobile SDK bridge. Four unnamed mobile gaming publishers participated in the beta.

What: Co-branded material from InMobi and Amazon Publisher Services reports that four gaming publishers in a beta of the APS mobile SDK bridge saw revenue growth of as much as 14 times through APS, on existing inventory and without changes to their mediation setups. The bridge activates native SDK rendering, unlocking rich ad formats from additional demand partners. According to Amazon Publisher Services, it provides access to more than 1,000 advertisers and 28 additional bidders without additional integration work.

When: The bridge was introduced at the APS Summit on Thursday, May 28, 2026, alongside five other capabilities. InMobi's supporting integration documentation carries a last-updated date of July 21, 2026. The Signal IQ A/B testing phase and the Publisher Supply AI open beta are scheduled for the summer of 2026.

Where: The capability applies to mobile application inventory. InMobi's documented integration covers mobile in-app formats including banner, interstitial, video in all formats and native, plus web banner, in-stream video and native, routed through Transparent Ad Marketplace and Unified Ad Marketplace.

Why: Bidder SDKs inside mobile applications have historically operated independently, limiting auction visibility and preventing demand sources from competing on equal terms, according to Amazon Publisher Services. Adding demand has typically required new integrations and application releases. A bridge that adds bidders and rich format support without engineering work removes the operational cost that has slowed adoption, while the reported revenue multiple remains a ceiling figure from a four-publisher gaming beta rather than a measured average across categories.