How Rival Telcos in Ukraine Learned to Share Identity Without Sharing Data

myGaru, an identity-focused ad tech platform, has gotten competing Ukrainian telecom carriers onto one shared identity system without merging their underlying subscriber data — pseudonymizing signals before pooling them, then minting a one-time token per ad request instead of exchanging personal data at all. At Kyivstar, Ukraine’s largest carrier, addressable reach on cookie-based matching topped out around 35%; after the integration, it reached 97%.

01What happened

• The pool, not the merge. MyGaru operates across Ukraine, Vietnam, and Azerbaijan, reaching 115+ million users by combining directly competing telcos’ reach on one platform without combining their data — each carrier pseudonymizes its own subscriber signals before handing them to myGaru.
• The mechanism. When an ad request fires, myGaru’s platform generates a one-time session token for audience matching, then discards it — no raw personal data crosses company lines, and no token outlives the request it was built for.
• Head of ad platform Yuriy Gorokhov, on why that design matters to skittish carriers: “Even if we were to get hacked and all the data is washed away, there’s nothing there to identify a person. No PII whatsoever is stored inside our perimeter.”
• The Kyivstar number is the proof point. Cookie-based matching alone reached roughly 35% of subscribers; myGaru’s pooled, tokenized approach pushed that to 97% — a gap that shows what an individual carrier’s own signal can and can’t do alone.
• MyGaru won’t enter a market below 70% combined population coverage. Gorokhov: “We don’t start in a market unless we see a clear way to getting at least 70% population coverage.” Below that threshold, a pooled identity system doesn’t clear the bar advertisers need.
• The harder problem is trust, not engineering. Gorokhov: “Telcos are very concerned about anything that could hurt their reputation or lead to audiences migrating to a competitor. It’s still a huge challenge” — rivals worry about each other more than about the token architecture.

02Why an operator should read this differently than a headline

MyGaru doesn’t name a Google Ad Manager, SSP, DSP, or PPID integration anywhere in its own account of the deal — it describes itself as platform-agnostic, “just the pipes,” and nothing here is a product a GAM operator can plug into today. What’s worth an operator’s attention is the mechanics, not the vendor: 35%-to-97% is roughly the ceiling-to-coalition gap every publisher running Google’s Publisher Provided Identifiers already lives inside. A single publisher’s PPID coverage is bounded by its own logged-in base; the only way past that ceiling — with cookies gone and no third-party fallback — is pooling signal with parties who were, until recently, competitors for the same inventory.

That reframes identity resolution as a coalition problem, not a per-vendor product decision. The 70%-coverage floor Gorokhov describes is myGaru’s version of a question every publisher building first-party identity should be asking of its own graph: what does “enough addressable reach to matter to a buyer” actually require, and is it reachable from a single seat’s data alone?

Nothing here is plug-and-play for a publisher running GAM today — no ad-serving integration is named, and myGaru's markets don't overlap with most of APH's publisher base.

03What this changes for buyers and publishers

1. Publishers should benchmark their own PPID or first-party addressable reach against the 35%-to-97% swing at Kyivstar. A jump that size comes from adding sources outside one company’s login base, not from tuning a single vendor’s match rate — if a publisher’s own number is stuck in the 30s, tuning won’t be what moves it.
2. This is a template, not a subscription — yet. No SSP, DSP, or GAM/PPID pairing is named in myGaru’s own account, and it operates in three markets most APH publishers don’t serve. Track the mechanism; don’t build a Q4 plan around the vendor.
3. The no-PII, tokenized-per-request design is the credible answer to the objection publishers already raise about pooled identity graphs — that a rival could see their data. Whether or not myGaru is the eventual partner, that architecture is the bar any coalition proposal should be measured against.

04The bottom line

Nothing here is plug-and-play for a publisher running GAM today — no ad-serving integration is named, and myGaru’s markets don’t overlap with most of APH’s publisher base. What it proves is a mechanism: competing telcos pooled subscriber signals through a no-PII, session-token architecture and took one carrier’s addressable reach from 35% to 97%. That’s what identity resolution looks like after cookies, for anyone who has to build it — a coalition of former rivals, not a single company’s dataset.