How I Learned to Live Google-free

This is part of IEEE Spectrum's special report on the battle for the future of the social Web.

If you’re reading this article right now, the chances are good that you’ve recently done business with Google. Maybe you got here via Google Search. Maybe you’re currently signed in to Gmail. Even if you don’t have a Google account, Google Analytics is likely tracking your movements on this site right now.

Google is in the business of data collection, and we, the users, provide the company with its raw source of wealth and power. Google may have entered this world a simple search engine, but the company has grown and evolved to become the core of our online lives.

At a panel on search engines in 2009, a Microsoft executive said that using Google is like smoking cigarettes: "It’s a habit that’s going to be difficult to give up." He was talking about switching search engines, but the difficulties of breaking the Google habit increase exponentially for people who also rely on Google Docs, the Chrome browser, or an Android phone.

Just a few short weeks ago, I was one such person. I used all these products and services daily. This wasn’t planned; my dependence on Google just...happened. One day I signed up for an e-mail account, and a couple of years later, I found myself linking it to an online calendar. Before long, I was using over a dozen services at both work and home. Because I already had one account (to which I was nearly always signed in), it was easy to take new products for a spin. Was I still actively choosing to be a Google customer, or was I effectively locked in for good?

I decided to put my own dependence on Google to the test. The plan was simple—to the best of my ability, I would sever my relationship with Google. I vowed to stop doing business with Google by no longer providing it information. Could I go Google-free without losing my digital quality of life?

My plan was to cut off my data stream to Google from three primary sources: my Google account, my search history, and my trail of browsing across the Web.

Dismantling the All-Encompassing Google Account

One reason this stunt seemed worthwhile is that I expected it to be hard. After all, I was quite content as a Google customer and even recommended many of its services to friends and colleagues. And as I began my research, I realized that I used a lot of services. Early on, I stumbled upon an incredibly handy page called Google Dashboard. At a glance, it showed how much stuff I had entrusted to Google: I had data stored in 16 different services.

There was so much information there that I had no interest in just deleting the account and quitting cold turkey. Instead, I wanted to recover as much data as possible and find out how difficult that would be.

Data portability is frequently discussed in debates about the pros and cons of using any cloud-based service. Once you upload or create content, is there any way to get it back? If so, the data is portable. Google certainly talks a good game about this issue: "When it’s easy for users to leave your product, there’s a sense of urgency to improve and innovate in order to keep your users. When your users are locked in, there’s a strong temptation to be complacent and focus less on making your product better."

That is a quote from Google’s introduction to its Data Liberation Front website, which I happened to run across on the second day. The DLF serves as Google’s clearinghouse for data portability information across all its products and divisions. At first glance the site is really impressive, and it fits Google’s image as a crusader of openness. (I can’t imagine Facebook or Apple having anything remotely similar in the near future.) For each Google service listed as "liberated," the site provides helpful links to instructions on how to download your data.

By following these directions, I found that Google excels at giving back your data whenever there’s a standard format for doing so. For example, it was a piece of cake to get all my appointments and recurring meetings out of Google Calendar —I just exported them as an iCalendar file and then imported that file into iCal on my Mac. It took less than 5 minutes. Similarly, it was really easy to get all my RSS feeds out of Google Reader (with even my folder structures intact). The feed addresses get stored in an Outline Processor Markup Language file, which was easy to import into my new feed reader, the Sage extension for Firefox. Google Docs is exemplary in its export settings: If you want to retrieve all your files, you get a single ZIP file that contains every document. You can even specify conversion settings, so that all documents are in Microsoft Word format or all spreadsheets are exported as comma-separated value (CSV) files.

But if Google succeeds at the obvious aspects of data export, it often falters with the smaller details. For instance, I had "starred" hundreds of items in Google Reader as particularly interesting—stories that I wanted to follow up on or possibly write about in the future. These starred items weren’t handled as part of the standard export process. I had to look beyond Google’s documentation to find instructions for exporting starred items as their own RSS feed, which I could then import into my new feed reader.

Similarly, Google’s more obscure services are minimally supported at best. I had been a big user of Google’s to-do list, called Tasks, adding and checking off items throughout the day. I struggled for a couple of hours to find a way to export my to-do lists, to no avail. I finally had to resort to first, sorting by due date, then printing the tasks to a PDF. At least I then could copy and paste each item without retyping. It’s no wonder that Tasks is not included in the "liberated" list of services.

As I worked on recovering all this data, I realized just how much I had invested in a single Web account. If my Gmail account got hacked, I would stand to lose a lot more than just my e-mails. And I had never really thought to look into the specifics of data portability when I added those additional services.

Replacing Gmail

Of all Google’s offerings, Gmail was always the one I most dreaded confronting. When I told my boss that I planned to get rid of my Gmail account entirely, he laughed and said, "That will wreak havoc." Another colleague asked, "Are you sure you want to do that?" After all, Gmail was my primary nonwork e-mail account, holding more than 2500 e-mail conversations.

Choosing an alternative proved challenging. As I considered the options, I was struck by the way our e-mail accounts have become the linchpins of our online identities. Using my e-mail address, I can reset almost any other user name, password, or Web account I set up online. That means that our e-mail accounts probably deserve the utmost attention to security. But even knowing this, I didn’t consider security the prime factor in selecting a replacement.

You see, I’m a sucker for a $0 price tag. After years of free e-mail service, I now experience cognitive dissonance over the concept of paid Web mail accounts. This is completely irrational, I admit. If I cared about e-mail security as much as I probably should, I would have at least considered paid accounts like those offered by Lavabit or FastMail. Each has a plan for less than $10 a year.

Instead, I went for a free service that had most of the features I’ve grown used to in Gmail: Zoho Mail. With the e-mail account, Zoho also provides an online productivity suite similar to Google Docs. But rather than being supported by ads, Zoho is a "freemium" model, in which users are encouraged to upgrade to a paid account with more features. I found that Google made it relatively easy to transfer my e-mails via Post Office Protocol (POP). But because POP transfers only a limited number of messages at one time, it took more than a day to move all my e-mails. After confirming that everything had migrated successfully, I then set up Gmail to automatically forward any new mail to my new address (and delete its own copy).

At Zoho, I was able to secure the same username I had with Google, so the only change in my address was from "gmail.com" to "zoho.com." Still, it feels weird to send e-mails with the new suffix. People do make value judgments about e-mail addresses. At least I didn’t choose Hotmail.

With my e-mails gone, my Google Dashboard was now as empty as possible (with a couple of exceptions—apparently it’s impossible to delete Google Webmaster accounts). I had managed to retrieve almost all my data.

The accomplishment felt good, even if it had taken over a week to achieve. My original intent was to delete the account altogether, and now that all my data was safely removed, there was nothing stopping me. But I couldn’t do it—I chickened out. I didn’t permanently delete the old Gmail account because it would mean that I would forever give up access to any e-mails sent to my old address. I realized that changing my primary e-mail address would be a serious hassle, not just for me but for all my contacts, which include less tech-savvy family members. I actually think it’s more of a pain than changing my physical address or even my cellphone number.

I’ll admit that extracting my data from my Google account was easier than I expected. I’m now convinced that Google has the right intentions and is committed to data portability. But there are still plenty of glitches in the execution of that vision that should make users think twice.

And they should be aware that retrieving data is different than erasing it. Even though I extracted copies of all my documents, e-mails, and RSS feeds and "deleted" them from my Google account, much of that data is still floating around on Google’s servers. For instance, I can still visit the "recently deleted" link in Google Tasks and see all the to-do items I tried to get rid of. My Gmail account still collects spam messages that it doesn’t forward to Zoho.

Finding a New Search

"Search" is still the verb that’s synonymous with Google, and searching is still the way I begin most of my forays across the Web. It’s such a fundamental and reliable process that it’s easy to take for granted—a minor miracle of technology that becomes invisible to us through routine. But people still remember the days before Google, and that’s why almost everyone I talked to about my little project professed their faith that Google Search is better than any alternative. And the market reflects that belief: Today Google handles 80 to 90 percent of search traffic worldwide, depending on whose metrics you believe. But in my experience, weaning myself from search was one of my first and easiest accomplishments.

When Google came out in the late ’90s, other search engines clearly had room for improvement. They were unreliable, and you could never be sure that the results were really what you wanted. Google smashed the competition by giving users both comprehensiveness and relevance. Over time, people began to trust Google’s algorithms more and more. Most users no longer check the second and third pages of search results because they’ve so often found worthwhile links in the top few spots.

Does that mean Google is really that much better than other search options? No. It just means that satisfied customers don’t see a reason to switch and haven’t had the need to explore competing engines. But several viable alternatives exist. And search quality is no longer such a difference maker.

For instance, Blekko is a relatively new search engine that I’ve used occasionally for research. It lets users create lists of sources that are authoritative resources for certain topics. For example, if you search for "memristor" and append the modifier "/tech," the results will be drawn from a list of top technology websites. In my experience, this makes it a superior alternative to Google News. But the interface is relatively unattractive, and it’s not as intuitive.

I also tried DuckDuckGo, which bills itself as a kind of anti-Google. It doesn’t save your search history, and it doesn’t pass along keywords to the sites you visit. Despite its clever marketing, I was skeptical when I first set it as the default search engine for my browsers. But the more I played around with DuckDuckGo, the more I found to like. It does some of its own indexing, but it also uses external search services to augment its results. By adding "bang" modifiers (like "!ieee" or "!wikipedia," for example), you can even use DuckDuckGo to go directly to results from other search engines.

I’ve been using DuckDuckGo as my search engine exclusively now for weeks, and I can’t say I miss Google at all. I like the fact that I can get good results without leaving a long digital trail, and it’s a fun to be outside the mainstream. I’ll probably stick with it for the same reason most people stick with Google: When you’ve got something that’s working already, switching is just not worth the hassle. I suspect that when it comes to search, laziness begets loyalty, as long as the results seem good enough.

Incidentally, there was one Google service that I found I could just not live without, no matter how hard I tried: YouTube. I was quite surprised that I kept catching myself watching videos, despite all efforts to abstain. It’s easy to take YouTube for granted because it’s been hyped in the press for years. There are other video sites, of course, but the depth, breadth, and ubiquity of YouTube became conspicuous every time I watched another video. As a Web journalist, I use YouTube when researching, writing, and illustrating stories. I run into YouTube videos all over the Web (and all over the IEEE Spectrum site, too). Even on the first day of my experiment, when I was most passionately committed to the project, I just couldn’t stop myself from exploring the " Guile themegoes witheverything" meme I uncovered while editing a story about video games.

Turning Off Tracking

Even if I had avoided conspicuous activities like YouTube and search, Google has plenty of other ways to collect information about me. For one, Google is also the Web’s biggest provider of ads, and it uses these ads to collect information from the visitors who view them. By storing browser cookies, its ads can track users across multiple pages and adapt to serve them ads that are "more relevant."

My interests, according to Google.

I was pleasantly surprised to see how transparent Google is about their ad tracking. It offers the ability to view and even edit your personal ad preferences. I was amused to see the categories of ads that Google had decided I would be interested in. It was like seeing a warped digital reflection of myself. Almost all the categories related to topics we regularly cover here at Spectrum, but my personal interests didn’t seem to make the cut, and Google failed to correctly guess my age.

To its credit, Google provides a way to opt out of this ad tracking. The problem is that opting out works exactly the same way the ads themselves do: by saving a cookie to your browser. To completely avoid tracking, I had to repeat the process on each of the five computers I routinely use. And I would need to do it again whenever I wanted to clear my cookies. Google’s solution to this annoyance is to offer a browser plug-in. It works by restoring the opt-out cookie whenever it gets deleted. The idea of installing a plug-in made by the same company that operates the service you’re trying to avoid seemed weird no matter how I looked at it, so I decided not to use it. But overall, opting out of ad tracking was easier than I had expected.

I was curious to see if the banner ads served to me would suddenly decline in quality. If such a decline took place, it certainly wasn’t noticeable. The experience only reinforced one of my complaints about targeted ads: They’re just not targeted well enough to be distinguishable from generic ads. Just because I write about engineering and technology for a living doesn’t mean I’m ever going to buy test and measurement equipment. Meanwhile, there are plenty of things I’m interested in buying that Google seems to have no clue about.

By this point, I started to notice a pattern emerging. Google can talk a good game about privacy and even provide numerous opt-out tools, partly because it knows the vast majority of its users will never bother to use them. Google’s sheer number of data sources means that it’s not dependent on a single service. Even if you opt out of Google ad tracking, for instance, you’ll still provide browsing information when you visit sites that use Google Analytics (like the Spectrum site). Opting out of that tracking requires a separate process. It seems reasonable that a user who objects to Google ad tracking would also like to limit data collection from other Google services. But there’s no master opt-out switch for users to flip.

Off Google for Good?

Before I began this experiment, I already knew that Google was a pervasive part of my life. But seeing how my Google account connected across platforms and services has made me reconsider the implications. Is it wise to place ever more trust and faith in a single Internet company? That question becomes even more important as Google expands its reach—through the Chrome browser, Chrome OS, Android tablets and phones, and its recently announced Google Wallet.

Although Google seems committed to transparency and openness, it’s still hard for average users to understand just which types of data each service collects and how that data is used. (The Wall Street Journal created an excellent infographic that shows how Google's data usage has changed over time.) Many Google services have their own individual privacy policies. While Gmail uses your e-mail text to serve you relevant ads, Google Calendar does not collect information about your appointments or meetings.

In general, quitting Google was easier than I thought. One of the biggest lessons for me was that Google’s not the best at everything. I’m thrilled to be rid of Google Tasks. I realize now that I was always dealing with its deficiencies; it’s not even supported on Android, and it had a tendency to undo my recent changes. I now use a site called Todoist, which I find vastly superior. I had never bothered to research alternatives before, and I ended up falling for the inferior product out of what I thought was convenience.

It’s easy to get seduced by the lure of a single sign-on. But managing multiple user accounts actually isn’t as much of an annoyance as we think it is. For me, it quickly became clear that my single Google account had mixed and muddled my personal and professional services and data. There are many online services that make sense to link together—but there are plenty of others that don’t. Calendar and e-mail might be a good fit, but do you need to use the same company to manage your social contacts, RSS feeds, and to-do lists? What about your phone and computer operating system? Even in the midst of the experiment, it was hard to remember to sign-out of the Google account; I was signed in by default, just as I’m also often signed in to Twitter and Facebook without realizing it.

Now that I’m done with the experiment, I have a much better idea about which Google products are most worth using. I plan to continue watching and enjoying YouTube on a daily basis. I have no desire to trade in my Android phone for an iPhone (personally, I wouldn’t even consider a BlackBerry or Windows Mobile Phone). Google Analytics remains invaluable for following the ups and downs of traffic to Spectrum’s website, but now I use a shared account that’s not affiliated with my personal content.

I see no reason to go back to Google Search as my primary search engine. I do miss Gmail, but Zoho Mail has been a very functional alternative, and I’m not sure it’s worth the effort to go back. There were many other services that I used so rarely that there’s not much to miss.

I think this experiment was extreme and not something I’d recommend for most users. I’m pretty sure that my defection makes no difference to Google’s bottom line, and as I’m just a single individual, my data just isn’t that valuable. But I think it’s worth considering which services you really need and which you don’t.

This is part of IEEE Spectrum's special report on the battle for the future of the social Web.