This post is a recap of my experience migrating to a Google-free environment. This transition is part of a broader change of habits I made in recent times. It is not just a matter of putting more focus on privacy, but also of escaping enshittified products, and creating some distance from (and not be a customer of) companies whose political and social stances have become more and more questionable.

As part of this change, for instance, I stopped using Amazon altogether one year ago, and I started paying a lot more attention to use products and services that are free from questionable political ties.¹

Beyond political and social concerns, enshittification practices have become more and more frustrating to me personally. No longer passively enduring these changes but rather ditching these products and migrating to alternatives, even if sometimes more expensive or less mature, gives a much better feeling of regaining a customer role rather than being myself the product.

Adding to that, more and more pieces of legislation are being pushed in different countries to put in place mechanisms enabling large-scale surveillance. Chat control is one of the most discussed proposals at the moment, but it only adds to many previous steps that have been eroding privacy and freedom of expression in recent times. Some successful acts of legislation in the opposite direction, such as GDPR, are important milestones but do not feel anywhere near enough. In light of this, allowing companies to get a hold of unnecessary personal information becomes more and more dangerous, as that data can be sold (legally or illegally) to bad faith private entities or hostile government actors.

The DeGoogle movement is not new but it gained significant popularity in the last couple years, likely in connection with recent political developments and the political stances taken not just by Google but also other large IT corporations.

Background

I have been using Android for almost fifteen years, since when I migrated from my last PDA² to my first smartphone. My previous Android devices have been two iterations of the Nexus phone and a Pixel 5.

I am used to custom ROMs and fiddling with system customization, as for a long time in the past I used CyanogenMod (later LineageOS). At the time I had three main motivating factors for using custom ROMs: having access to system features not yet part of AOSP or OEM Android,³ getting faster and longer update support,⁴ and avoiding OEM bloatware.

However, as Android matured, I later settled for vanilla Google Android on Nexus (and later Pixel) devices, as it offered the features I needed with little bloatware compared to other OEMs. Update and support has remained short and underwhelming, but for the rest I found little to no appeal in other Android OEMs.

I was however aware of the trade-off that came with using Google products, and the terrible privacy practices that came with it. In the last few years I came to like less and less the policies adopted by Google,⁵ both in terms of products and general stance, and on top of that I decided that the privacy trade-off was no longer acceptable.⁶ For these reason, I decided to migrate out of the Google ecosystem entirely.

All starts with the email

The obvious first step has been moving out of Google products. I started using the Google ecosystem with Gmail, and ironically email was the first step on the way out as well. Gmail is the only Google service I truly depended on, as for all other Google products I either only had minor and not strictly necessary use cases, or no use at all.

For a privacy-oriented focus, I opted to make Proton Mail my new main email account. I have already had a free Proton account for some time as a secondary email, as I have been planning this transition for a couple years already,⁷ so I switched to a paid plan⁸ and migrated all my data from Gmail.

The plan includes end-to-end encrypted cloud storage and will allow me to also get rid of my current cloud storage subscription,⁹ so the effective cost of this transition, even without accounting for the privacy benefits, is much lower than the label price.

Ecosystem vs single-purpose

While Proton has a strong privacy-focused brand, the choice of email provider is not automatic and arguments could be made about it. One multi-faced aspect is that Proton is no longer just an email provider, but it now offers a cloud suite including file storage, VPN, LLMs,¹⁰ wallet, and now a small but slowly growing office suite.

There would be reasons to prefer a single-purpose focused email provider over a wannabe cloud ecosystem, but in my case some other Proton products happen to fulfill my use cases, so this ends up being an advantage for me.

The could storage is not particularly appealing over competitors in terms of available storage, pricing, speed, and software. The lack of a viable Linux client is a big negative for me, but the recent introduction of an SDK will hopefully fill that gap in the future. The lack of a fast and robust bidirectional synchronization is also limiting.

In my use case, however, I routinely use only a small set of data, while the bulk of data is old items that I do not need to edit and that I rarely access. So I can move the bulk of rarely used data to Proton, without really needing a desktop client nor a mount point for it, and keep the small but frequently used data on Dropbox, which allows me to switch to a lower tier Dropbox subscription due to the now much smaller size requirement.

The office suite is also useful to me, as I have very limited use cases where a Google Docs replacement is handy and saves me from the need for a separate solution.¹¹ The VPN is also handy, it is not an essential tool but I want to have one at hand for technical reasons, and this saves me the need to get it from a separate provider.

Other products in the Proton ecosystem, like wallets and LLM/AI, are not really useful to me.

Choice of phone

I decided to get a new smartphone, which is a rare event¹² and usually requires going through many considerations. Switching to iPhone is not an option, as it implies exchanging one corporate ecosystem for another.¹³ I decided to get a Fairphone 6, for a few different reasons.

Fairphone is an EU-based company that manufactures devices with great lifespan and repairability. It also has a strong focus on sustainability, which is in line with my own values. Fairphone also partners with Murena to offer a Google-free variant of their phones out of the box.

The choice of operating system was one out of several options. GrapheneOS is arguably the safest mobile operating system available at the moment. However it only supports Pixel devices for the foreseeable future, making it a hard pass for me as I am not willing to buy more Google products, not even if just for the hardware. While GrapheneOS’s superior hardening is very interesting, it is an overkill for my personal threat model and not necessary to achieve my narrower goal of getting rid of Google products and services.

In the past I used LineageOS and appreciated it for its simplicity and freedom from bloat. A de-googled LineageOS installation could therefore be a candidate. /e/os is a relatively new system, based on LineageOS, and I decided to give it a chance.

I am aware that prioritising usability, as Murena (the company developing /e/os) does, implies some compromises that could be avoided by going for a more hard-core option like GrapheneOS. I also do not blindly trust the company and look critically ahead to how it will continue its mission in the future. I see this migration primarily as an experiment that might require corrections (all the way up to switching to a different Android ROM).

Switching to /e/os

Compared to other Android distributions, /e/os focuses on privacy and usability. For Fairphone buyers, this also includes a two years software warranty.¹⁴ For instance it ships built-in tracker protection,¹⁵ and it anonymizes some services that still rely on interactions with Google.

MicroG, a free and open source reimplementation of Google Play Services, is pre-installed and configured with reasonable defaults, the App Lounge (/e/os’s app store) allows seamless access to both F-Droid and Google Play Store (either through log-in with a Google account or in an anonymous way) and allows to install apps from there without having to install the Play Store app itself.

I am on the knowledgeable and experienced side of users, so having to configure everything from scratch would not be an issue, but I find it interesting to see a company putting effort into creating a privacy-focused alternative without the hassles that would normally come with it, making privacy accessible to everyone, not just seasoned users.

Speaking of smoothening the user experience, Murena also offers a platform for cloud synchronization, which is something missing from most other alternatives (for good or bad). I have decided to go without cloud for now, as I want to avoid creating more ties with cloud providers, including Murena itself, and see what the experience is like without any cloud integration.

The migration has been incredibly smooth, much less problematic than I expected.¹⁶ All my apps worked just fine, even digital identity and banking apps, which are the ones with the most potential to be fussy. Only one manual settings adjustment was required to get some apps to always work when invoked from links in the browser, I had to manually select in the app settings which app link domains to enable (simply allowing those apps to open links was not enough).

Caveats

While the switch has been surprisingly smooth, there are some unpolished or underwhelming corners that I hit pretty quickly.

Software

The first one is of personal taste (de gustibus non est disputandum and all), but I personally am not a fan of the iPhone-like UI in the default Murena-developed Bliss launcher.¹⁷ Thankfully, however, /e/os is not tied to its default launcher, and for a Pixel escapee like me Lawnchair provides the familiar clean and tidy UI experience with the advantage of a much greater deal of customizability compared to Pixel. Lawnchair is just one among several high quality options for open source Android launchers, so on this aspect there is ample room for choice.

Most stock apps are very bare-bone and some of them are borderline unusable. As an example, at the time of writing, the gallery ignores exif metadata (scrambling the chronological order of imported images) and does not even have a scrollbar, making it painful to navigate if you have more than just a few images. Some software functions feel also less calibrated compared to Pixel (for instance the automatic display rotation being too trigger-happy) and remember me of Google phones from many generations ago.

Thankfully there are excellent open source and privacy-friendly projects to piece the puzzle, like the Aves image gallery. I replaced Google Keep by moving my recipes to Proton Docs¹⁸ and using Fossify Notes for grocery lists and similar.

The built-in /e/os camera app also reminds me of older Android versions, but as an image processing professional I appreciate the large amount of settings, customization, and manual toggles it provides. I never liked the trend of Google to remove manual options and settings from the camera app.

One big absence is Google Pay, which cannot be used through MicroG. This prevents from using any wallets and contactless payment services relying on it. I personally do not use contactless payments on the phone, as they serve no practical purpose (nor give any real benefit nor convenience) compared to directly using a contactless bank card, while on the downside they would require sharing my whole transaction history with a third party (like Google) for no good reason.

This however does not mean it is impossible to use contactless payments without Google Pay. Some bank apps, and also some neobanks or banking intermediaries, implement their own NFC payment mechanism instead of relying on Google Pay, therefore allowing to execute contactless payments from Google-free phones.

Privacy

/e/os ships MagicEarth as its preinstalled maps app, which is not open source. It is considered one of the best alternatives to Google Maps for navigation, and since navigation is my biggest use case it might be a good enough tool. Nonetheless, and regardless of the privacy statement, the lack of openness is a turn down.¹⁹

One also need to be careful about which advanced features one enables, as that is one of the aspects where privacy compromises might pop up. An example is the text-to-speech, that currently relies on OpenAI.

/e/os also ships some pre-installed system apps (browser, maps, notes, etc.) that cannot be uninstalled. None of these seem to be invasive, but I still see this as a downside compared to cleaner systems like GrapheneOS.

Another aspect to keep in mind is that custom ROMs (like LineageOS and /e/os) and OEMs (including Fairphone) can be slower in delivering Android updates in full.

A detailed analysis of shortcomings in /e/os and Fairphone can be found in an official statement from GrapheneOS. In a nutshell, GrapheneOS is presently the best option by far when it comes to hardening and safety. If one is willing to buy a Google Pixel phone to run it, which I am no longer willing to do.

Hardware

The most annoying usability complaints I personally found are hardware-related, however. I personally do not like the fingerprint reader on the power button, as it makes unnecessarily hard to turn the screen on and off without accidentally unlocking the device (even when disabling the always-on fingerprint unlocking). I will have to get used to turning the display off by double-tapping the top edge of the screen (which is not particularly practical). I wish I could also double-tap to turn it on, but always-on display is not supported on /e/os,²⁰ whose absence can be felt if one is used to it.

I also dislike the placement of the up-down volume buttons on the opposite side of the device, being placed too low. It is easy to accidentally click the volume down button while pressing the power button on the other side, and I have already taken way too many unintentional screenshots because of it. While I can understand the choice of bundling the fingerprint reader with the power button to simplify the hardware design, the poor placement of the volume buttons is hard to justify, especially for a sixth generation product.

I will also miss the convenience of wireless charging, and will have to get used again to fiddle with cables (or find a good charging dock), as I do not think I will be willing to deal with the bulkiness of an external wireless charge adapter.²¹ Wireless charging is missing by design, as it is not a power-efficient solution. Even with a wireless charger on the lower end of efficiency, though, the amount of wasted power would be fairly limited in absolute terms, and I could be willing to trade it for the convenience.

Is it really Google-free?

In the end, the straight answer is no. Not entirely. While it is possible to set up a truly Google-free Android derivative (a stock GrapheneOS is a good example), the truth is that many apps have dependencies on Google Play Services, and will break without them.

Applications such as digital identity, banking, and government-related services are becoming more and more important in most societies, approaching the borderline indispensable. Therefore a truly Google-free device (as in, completely free from Google Play Services) would be functionally broken in this context. And even the current approaches to tame Google Play Services, such as using MicroG or by running Google apps in sandboxed mode, feel brittle in terms of long-term viability, as Google could take action to break such solutions, and there would be no one that could realistically stop them.²²

My personal goal for this migration has been to get rid of first-hand dependency from Google services, and limiting interaction with Google APIs to the strict necessary and only for the apps I truly need. Nonetheless, reflecting on this gives a worrisome perspective of how how deeply society has become entangled with private actors like Google and other large IT firms, that have built for themselves a position which is becoming harder and harder to extricate from functional services in modern society, and would require some amount of re-thinking in terms of long term strategy and national sovereignty.

Footnotes