For kosher device manufacturers: what licensing a complete app layer looks like

Forty dollars. When European regulators forced Google to unbundle its application suite from Android in 2018, the trade press obtained the fee schedule, and the number became public: up to $40 per phone for the Google app suite on high-end devices in the European Economic Area, tiered by country and screen density, with Search and Chrome licensed separately. Manufacturers who had spent a decade treating software as free were shown what the industry's balance sheets already knew: a complete, integrated application layer is one of the most valuable line items on a handset, and the company that owns one sets terms.

Now hold that number next to a different one: zero. That is how many complete application layers have ever been available to license in the kosher device market — a market with roughly half a million subscriber lines in Israel, its own framework anchored in Knesset law, its own certification bodies, a shelf of purpose-built device brands, and twenty years of demonstrated demand. Every kosher device ever sold was assembled: someone else's hardware, a curated slice of someone else's apps, someone else's oversight. There was never a whole software layer built for this market that a manufacturer could simply buy.

That has changed, and this page is the briefing. It covers what actually ships in this market today, what complete software layers cost in the general market, the one cautionary tale every executive should know before signing with any software partner, and a seven-question diligence checklist to run on anyone who pitches you — including KolBo. We build the product this article describes, and we would rather you evaluate it with the full picture than without it. The pitch will stay out of the way until you have that picture.

The market you are actually entering

The kosher device market stopped being a curiosity years ago. In Israel, roughly half a million people use kosher phone lines — an arrangement significant enough that the Knesset anchored it in law, with the Rabbinical Committee for Communications coordinating dedicated number prefixes with every major carrier since 2005. Pelephone, Cellcom, MIRS, and Partner have all shipped kosher offerings. According to the Institute for National Security Studies, some 70 percent of Israel's haredi community carries a phone with no internet access at all — by choice, as a standard, generation after generation. This is not a niche waiting to be convinced. It is a mature customer base that has organized carriers, legislation, and communal infrastructure around a preference it holds deliberately.

In North America, the market has grown into a full commercial ecosystem. There are purpose-built device brands — Fig, Wonder, Pom, Tak, Mind, Qin, MegaLife. There are dedicated operating systems: SafeTelecom's KosherOS ships on current Google Pixel and Motorola hardware; the Mind Phone runs its own MindOS. There are retailers and distributors — KosherSignal, EZ Cell, Planet Cell, KosherCell — and institutional programs like YeshivaLink and TalknSave that place devices in bulk for yeshiva and seminary students every Elul, a purchasing season as predictable as the school calendar that drives it. And there are community certification bodies — TAG, with close to sixty branches worldwide since the 2012 Citi Field gathering, alongside Letaher and regional va'adim — whose approval functions the way a hechsher functions in any other category: it is the difference between a product a family will bring home and one it won't.

Demand seasonality tracks lifecycle moments — a bar mitzvah, a yeshiva enrollment, a chassanah and a new household — and the institutional channel of schools and yeshivos is structural, not incidental. If you manufacture devices for this market, or you are evaluating an entry, none of this is news to you. What may be news — and what this page exists to lay out plainly — is that the software side of this market has just changed shape.

What ships today: a market built by removal

Walk the current landscape and a pattern emerges immediately. As of mid-2026, kosher devices sold in North America fall into three access tiers, and the industry itself uses this vocabulary: talk-only, talk-and-text, and apps-with-limits — a curated handful, typically Waze, Gmail, banking, sometimes a restricted WhatsApp, per KosherSignal's 2026 buyer's guide. Across all three, the browser is deleted and the app store removed. Look at how these devices are made and the pattern deepens:

• The hardware is someone else's. KosherOS ships on Google Pixel 7, 7a, 9, 9a, and 10a, and on the Moto G 5G — mainstream hardware, reflashed. The reconditioned flip-phone tier runs on TCL, Kyocera, LG, Orbic, and Sonim devices adapted after the fact with community-appropriate settings.
• The apps are someone else's. The apps-with-limits tier is, almost universally, a curated slice of Google's catalog: Gmail, Waze, Google Calendar, a banking app. The value the vendor adds is subtraction — deciding what not to include.
• The oversight is someone else's. Certification comes from third parties — TAG, Letaher, a va'ad — and even here the industry acknowledges a gap. TAG of Baltimore states outright that many kosher devices on the market today carry no official certification at all, and that the industry "does not have formal oversight."

None of this is a criticism of the people doing the work. The vendors in this market serve their communities seriously, and the three-tier structure exists because families genuinely hold different standards — a Satmar household, a Litvish yeshiva bochur, and a working baal habayis who needs WhatsApp for parnassah do not hold the same line, and that diversity is a feature of communal life, not a flaw. But from a manufacturer's chair, the structural consequence is unavoidable: when every vendor's product is a different subtraction from the same Google catalog, no vendor has a durable moat, and every device competes on price.

There is a second consequence, and it is the one your customers feel. You cannot subtract your way to a new capability. No amount of removing apps produces a family-location platform. No configuration of deletions yields turn-by-turn navigation that is compliant by design rather than by exception. This is why Waze is the single most common carve-out on apps-with-limits devices: navigation is the pillar families could neither go without nor get in a form built for them — you cannot reach the simcha, the appointment, or the levaya without directions — so the market shrugged and made an exception. And every exception erodes the standard the device exists to hold. The gap between what removal can produce and what families actually need is the market's defining tension in 2026, and it is precisely the gap a built-from-scratch application layer closes. (For the consumer-side view of this landscape, see our 2026 kosher phone guide.)

“Anyone can remove features and call it kosher.”

kolbo.life

“Application layer” is not jargon — it's a line item every OEM already pays

Outside this market, no device maker builds everything. The modern device business runs on licensed software layers, and the reference points are well documented. Four of them will calibrate any licensing conversation you have with anyone:

1. Google Mobile Services — up to $40 per device. The number that opened this article. When the EU forced Google to unbundle its app suite in the European Economic Area, the pricing became public: up to $40 per phone for the Google app suite on high-end devices (up to $20 for tablets), effective on devices activated from February 1, 2019. That is the world's clearest price signal for what a complete, integrated app suite is worth on a handset: tens of dollars per unit, at Google's scale.

2. Microsoft's Android royalties — $5 to $15 per device. For most of a decade, Microsoft collected an estimated $5 to $15 on every Android device sold by major OEMs — reportedly around $12–13 per Samsung device at the 2012 peak, and north of a billion dollars a year from Samsung alone. Manufacturers paid it, year after year, because the per-unit math still worked. Per-device software economics are not exotic; they are the industry's default.

3. Automotive navigation — $25 to $50 per vehicle. Car OEMs have licensed navigation stacks from HERE and TomTom for decades, at an estimated $25–50 per vehicle, under negotiated multi-year agreements bundling map data, software, and updates. A category-defining app, licensed per unit, integrated by the supplier's engineers onto the OEM's hardware — that model is older than the smartphone.

4. Preinstall economics — the direction of payment reveals the value. In the general market, most third-party developers pay OEMs for preinstall slots; the UK competition authority's review of Google's OEM agreements documents per-device activation payments flowing to manufacturers for placement. Software pays for shelf space when it is interchangeable. The payment flows the other way — OEM to supplier — only when the software is the product. GMS, Microsoft's royalties, and automotive navigation all sit in that second category.

Here is the point of the tour: the kosher device market has never had anything in the second category to buy. There was no complete, engineered, community-standard application layer to license at any price. Manufacturers assembled subtractions because assembly was the only option on the table. The build-versus-license decision that every general-market OEM has faced for twenty years arrives in this market only now — which is why, if you are reading this after a pitch meeting, the idea probably sounded both obvious and unprecedented. It is both.

The Cyanogen lesson: why partial layers fail

Before any checklist, one cautionary tale belongs in the file, because it is the strongest argument for rigor in choosing a software partner — and because your team will find it in an afternoon of diligence anyway.

Cyanogen Inc. was the last serious attempt to license an alternative Android software layer to device manufacturers. It raised enormous capital, signed OnePlus, BQ, and Wileyfox, and by December 2016 had shut down its OS and services entirely. The post-mortems converge on two causes worth an executive's attention:

• The layer was incomplete. Licensing Cyanogen meant asking users to live without the services that made a phone useful, with no equivalent replacements. An app layer that leaves daily-life gaps forces the OEM's customers to route around it — and every workaround is a crack in the product.
• Updates broke the economics. Maintaining and updating a fleet of devices in the field cost a fortune the business model never supported, and engineering resources drained away from the platform toward one-off client builds.

Both failure modes translate directly to this market. A kosher app layer that lacks mail, navigation, or messaging will generate exception requests — the same Waze-shaped cracks the market lives with today. And a software partner who cannot fund a real update pipeline will leave your fleet aging in the field, which in this market is not merely a security problem but a communal-trust problem: the device carries your name and, implicitly, the standard it promised the family that bought it.

The diligence checklist: seven questions for any app-layer partner

Run these on KolBo. Run them on anyone. They are ordered by how expensive the wrong answer becomes.

1. Is the layer complete — or will my customers need exceptions? Count the daily-life jobs: calls, messages, mail, navigation, contacts, calendar, camera, photos, notes, weather, an alarm, music, learning. Every job the suite doesn't cover is a future exception request, and every exception erodes both the standard and your support economics. This is the Cyanogen test, and it is pass/fail.
2. Who owns the code? If your partner's suite is a curation of Google's apps, your differentiation is a configuration file — reproducible by any competitor in a quarter. If the partner engineered the applications, the moat is theirs to license and yours to ship. Ask the question app by app.
3. What is the update pipeline, concretely? The general market has formalized this: Google's Android Enterprise Recommended program originally required security updates within 90 days for three years before shifting to mandatory transparency about update cadence. Whatever the paperwork says, the operational questions are: who builds the updates, who pushes them, who pays for the infrastructure, and does the whole suite move together or app-by-app? Fragmented update paths are where fleets go to decay.
4. Can the protection be peeled off? In this market, this is the certification question wearing engineering clothes. A safeguard a determined teenager can defeat is not a safeguard; it is a liability with your logo on it. Ask where enforcement lives — in an app, where it can be uninstalled, or beneath the apps, where it can't.
5. Will the device clear community standards? The certification landscape — TAG, Letaher, the va'adim — is the market's trust infrastructure, and as TAG itself notes, the industry has no formal oversight body. That places the burden on the software's architecture: a device whose compliance is structural rather than configured is a device a certifier can evaluate once and stand behind. Ask your partner to demonstrate compliance on real hardware, not describe it on a slide.
6. What is the support model for the fleet? Devices in this market are bought at lifecycle moments and expected to last. Someone must answer when a device in year three needs help. Is that you, or your partner, and under what terms?
7. Do the per-device economics leave you a margin story that isn't price? Recall the reference points: $40 for GMS in the EEA, $5–15 in Microsoft royalties, $25–50 for automotive navigation. Complete software layers command real per-unit value — and they return it when they carry exclusives. The question is not whether licensed software costs money. It is whether what you license lets you stop selling on price.

How KolBo answers the checklist

Now the part you scrolled for. KolBo's position, in our own words from the homepage: KolBo builds the entire application layer of a kosher device — 22 interoperable apps, engineered in-house, secured before they ship — and licenses it to the manufacturers who intend to lead this market. Here is that claim held against each of the seven questions.

1. Completeness

Twenty-two applications, built in-house, spanning the full daily surface of a device: Safe, Go, Search, Browser, AI, Phone, Text, Mail, WhatsApp, Voice, Keyboard, Zmanim, Calendar, Alarm & Clock, Weather, Contacts, Camera, Gallery, Player, Library, Calculator, Notes. The name is the claim — KolBo means "everything in it" — and the suite includes the categories this market has never had. KolBo Safe is the first family safety platform built for the Jewish world: real-time family location, arrival and departure alerts for school, home, yeshiva, seminary. KolBo Go is the first kosher navigation ever made — full turn-by-turn, built kosher from the first line of code. Not modified. Not wrapped. Made. The Waze exception — the crack running through every apps-with-limits device on the shelf — closes, because the navigation is no longer an exception to the standard. It is the standard: compliant by architecture, nothing to disable, nothing to bypass.

2. Ownership

Every application is engineered in-house, on one platform, one cloud, one security layer — four category firsts among them. That produces things curation structurally cannot. Open a contact on a KolBo device and see every call, every text, and every email with that person in one unified timeline — possible only because one company builds Phone, Text, and Mail on one platform. As the homepage puts it: not even Apple or Samsung ships this. Zmanim isn't an app here; it's a service every other app draws on — the Calendar schedules around it, the Alarm wakes by it. A bundle is apps sitting next to each other. A platform is apps built for each other. No configuration file reproduces that, in a quarter or in a decade.

3. The update pipeline

One pipeline, run by KolBo: the whole suite updates together and sits on the same enforcement layer, and devices in the field stay current, compliant, and protected for their entire life. The Cyanogen failure mode — updates as an unfunded afterthought — is inverted here: the pipeline is part of what you license, not a cost you discover later.

4. Tamper resistance

Protection is enforced at the device-policy level: remove the management layer and the safeguard stays locked. Proven on real hardware, not in a slide deck. Security in the KolBo model is not an app on the device; it is the layer under everything — under everything we build sits security nobody can peel off — with AI sight protection screening images, video, and text in real time.

5. Community standards

Pre-secured and compliant out of the box: every app ships on the enforcement layer, so your devices clear community standards the day they leave the line. Compliance by architecture is also what makes a certifier's job tractable — there is nothing for a user to disable, so there is nothing to re-audit after every family's first week. In a market whose own certifiers say the industry lacks formal oversight, structural compliance is the strongest answer available to anyone.

6. Support

The division of labor is the automotive-navigation model, stated plainly on the homepage: we integrate onto your hardware, run the update pipeline, and support the fleet. You ship complete devices. Engineering-led integration by the supplier, fleet operations by the supplier, hardware and channel by you — under your brand.

7. Economics

Wholesale terms, engineering-led integration — the structure every reference point in this article uses — with the margin story attached: Safe and Go are category firsts, and the manufacturer whose devices carry them isn't competing on price anymore. In a market where every product is a different subtraction from the same catalog, exclusives are the only durable answer to price competition, and Safe and Go cannot be assembled by removal. By anyone. That is the whole argument, and it is also why the first question on the checklist was completeness rather than cost.

Side by side, the two ways of making a kosher device look like this:

What your device becomes, in the homepage's own stack: 22 KolBo applications (including Safe, Go, Browser) · platform services — account, updates, data, everything syncing to one KolBo Cloud · the security layer, always enforced · your hardware, shipping complete.

What the process looks like

KolBo works with a limited number of manufacturers per region — flagship exclusives are only exclusive if they are scarce. The path in is a partnership briefing: write to partners@kolbo.life, and briefings are answered within one business day. Integration is engineering-led onto your hardware; you ship complete devices under your brand, on both iOS and Android, secured on both platforms.

Everything described on this page is built in Brooklyn — every application, every platform service, and every line of the security layer is engineered in-house and proven on real devices before it ships. That sentence is the company's standard, and it is also the honest summary of why this partnership category exists now and didn't five years ago: someone had to build the whole layer first, prove it on hardware, and only then offer it wholesale. (More on the team and the standard on our about page.)

The kosher device market spent twenty years proving the demand: half a million lines in Israel, a law in the Knesset, sixty TAG branches, a shelf of brands, an Elul buying season you can set a calendar by. What it never had was software built for it rather than subtracted toward it. Kosher technology has always been a step behind. That era is over — and the manufacturers who move first will be the reason it stays over.

Frequently asked questions