The Windows Maintenance Challenge: Part 2

Fred Langa

Can commercial, third-party maintenance software outperform Windows’ built-in, free tools?

This is the conclusion of a two-part series that will help you determine which PC maintenance tools — free or paid — yield the best results on your specific system.

PC maintenance suites: Free versus paid

This two-part series came about because of an ongoing debate over PC scan-and-repair tools. We’ve all seen the ads for paid Windows-maintenance tools. They typically claim to correct slowdowns, improve system-startup times, prevent crashes, fix all sorts of hidden system errors, and more.

But do they really deliver as advertised? Do they do more than the free tools built into Windows? Over the years, numerous tech publications have attempted to answer those questions. But, as I noted in Part 1 of this series, all those reviews have had one major flaw: the test results on one or two systems can’t be applied generally to most Windows PCs. Windows configurations vary enormously, and what works (or doesn’t work) on one setup might not work (or work just fine) on another setup.

So the Windows Maintenance Challenge takes a different tack to answer the long-standing free-versus-paid question — it shows you how to test maintenance tools yourself and find what works best for your specific Windows configuration.

Part 1 of this series discussed how to build and check a thoroughly cleaned, baseline system using the free tools that come with Windows. Part 2 will tell you how to run commercial Windows-maintenance tools on the baseline system and find out whether they make any significant improvements over what you get for free.

As with Part 1, this article is designed to let you safely work along with me. You don’t have to test the same group of paid products discussed below — you can use the following steps to test whatever maintenance tools you wish. When you’re done, you’ll know which tool yields the best results for you.

Of course, you can also simply read these articles and use my results as a general guide to free and commercial maintenance tools and techniques. You don’t have to do anything to your PC if you don’t want to!

But by participating, you’re sure to have a cleaner and more secure system. Why not give it a try?

Ready? Let’s begin.

Some final steps before testing paid products

Again, today’s article builds on the information, procedures, and preparations discussed in Part 1. To get up to speed on the topic, or to verify that your PC is ready for today’s tests, please review the Aug. 14 Top Story, “The Windows Maintenance Challenge: Part 1.”

(Note: Before starting this process, you’ll want to be comfortable with creating and restoring complete system backups. And you’ll want to ensure that your backup/restore system really works.)

With the preparatory tasks in Part 1 done, I was ready to start testing Windows-maintenance tools. Here’s how I ran the tests — and how you can, too.

I began by making a complete, up-to-the-minute system image. I prefer system imaging, but you can use any backup tool you’re comfortable with, as long as it can completely restore your entire system from scratch.

This system image is in addition to any other system image you’ve already made. Its only function is to create an initial, standardized, baseline system for each maintenance tool tested.

While the imaging was taking place, I used that time to organize some of the data gathered in Part 1 into a simple table format.

Your table can include whatever metrics you want to track. Those shown below in Table 1 are just examples — a possible starting point.

Baseline system

Table 1. The baseline system's recorded stats. The stability and security results are based on Windows' built-in Reliability Monitor and represent a clean system.

Here’s why I picked these specific examples.

  • Restart: Some commercial maintenance tools promise to automatically improve boot times. Tracking the length of a shutdown/restart cycle lets you examine that claim.
  • File size: Tracking the aggregate size of all files on C: will let you verify claims of superior junk-file removal.
  • Registry size: Some tools boast of enhanced Registry cleaning and compression. Noting the original and final Registry sizes can help gauge that claim.
  • Stability and security issues: Assuming that Windows’ built-in Reliability Monitor can find any true system errors, the baseline configuration has no known defects. (Nearly every PC has minor errors that don’t affect system performance or reliability.) That “zero-defect” condition provides the baseline for paid maintenance tools that claim to discover hidden stability or security issues.

Again, these are just a few examples. Your table can include any measurements you consider important.

As with the testing metrics, you can test whatever maintenance tools you wish. I picked Reimage, Iolo’s System Mechanic, and PC Pitstop’s PC Matic mostly because of their high visibility — and because they offer either free diagnostic tools or money-back guarantees on the full commercial tool.

Note. Keep in mind that these products were tested on pre-cleaned baseline systems. After Part 1 was published, a reader asked why I hadn’t used a system known to have issues. The short answer: there’s no way to create a representative “dirty” Windows setup — the results would have been meaningless. And using a known-good setup provided information on the biggest problem with many automated Windows-maintenance tools: false positives. (More on that below.)

By luck of the draw, Reimage was up first.

Reimage: An automated maintenance system

As with many commercial maintenance tools, Reimage offers a free diagnostic scan of your system. But if you want the software to make repairs, you have to pay. I won’t post a long laundry list of Reimage’s features and options; that’s all available on its homepage.

Using Reimage starts with downloading and running the Reimage Repair Setup Wizard software. Once the installation process finishes, the program immediately begins scanning your system. Reimage first compiles and displays a system profile and then provides a reliability report based on information stored by Windows. It then scans for malware.

Once the scan is done, Reimage displays a relatively simple summary report. On my baseline system (freshly restored from the cleaned baseline image), Reimage reported 10 “stability issues” and one “security threat.” Surprisingly, it assessed the test PC’s Windows Damage Severity as “high” (see Figure 1). Remember, it was scanning a system that had already had a thorough cleaning.

Reimage scan results

Figure 1. Reimage's free report on my clean baseline system's condition. It rated system damage as high.

The Security Threat was alarming. Reimage reported that a Windows system file contained a virus (see Figure 2). If true, that meant the operating system itself was compromised!

Reimage security report

Figure 2. Reimage's description of an infected Windows file

I was dubious about the supposed infection. Before the baseline image was created, the system had been scanned and found to be 100 percent clean. And Reimage was installed and run immediately after I restored the image. There should have been no opportunity to become infected. Moreover the “infected” file — searchfilterhost.exe — looked okay; its file properties appeared identical in every way to the same file on other Windows systems.

Was there a virus? To find out, I scanned searchfilterhost.exe using six different and well-regarded anti-malware tools available online or built into Windows:

Because none of the six scans flagged the “infected” file, I concluded that Reimage’s malware report was a flagrant false positive.

Next, I looked at the 10 reported “stability issues” and cross-referenced Reimage’s reports with incidents recorded by Windows’ built-in Reliability Monitor (see Part 1).

None of the reported “stability issues” was serious; none required action or attention. For example, at some point during system prep, I’d started a backup but then decided to do a full system image. So I manually canceled the backup.

Reimage apparently dug this harmless “backup did not complete” datum out of Windows’ internal reliability logs and reported it as “Windows Backup has crashed.” Reimage also offered to “fix the problem.” Given that I was the problem (I had aborted the backup), I wondered what Reimage’s “fix” would be.

The more I looked at the 10 “security issues,” the more convinced I became that they were false positives, mistakes, or overreactions to routine, trivial, and harmless system events.

If I were using Reimage to clean my day-to-day PC, I probably would have bailed out then and there. As far as I could tell, Reimage wasn’t flagging anything substantive, and the seemingly erroneous or exaggerated “issues” didn’t inspire my confidence.

But for the Maintenance Challenge to be thorough, I needed to complete my test of Reimage. I ordered a Reimage license (U.S. $70 per year) and authorized the software to start repairs.

Surprisingly, the repair process took about two hours to complete, including a reboot in the middle. Reimage’s work was mostly opaque, providing only the vaguest on-screen indication (e.g., “Repair damage”) of what it was fixing.

But the real surprise came after the reboot. Reimage spent a very long time downloading many dozens — possibly hundreds — of Windows Update files. I have no idea why; the baseline system was already 100 percent up to date.

When Reimage finished its work, it presented a summary stating that my system had been “missing” three percent of its operating system files — and that another 12 percent were “damaged or incomplete.” I don’t know what that means, and Reimage provided no additional details.

But I do know that Windows’ own System File Checker didn’t notice any missing or damaged system files, and Windows Update reported that no updates were needed. Also, the baseline system was derived from a machine that had been working fine for about four years.

I could only conclude that Reimage’s file-replacement activity was unnecessary.

With Reimage’s scan and repair done, I next took some sample measurements, shown in Table 2.

Reimage metrics

Table 2. Reimage's scan metrics varied only slightly from those of the baseline system.

As you can see from the numbers, the clean/repair process made no significant changes to the baseline setup. Boot-up time got marginally longer, while the file and Registry sizes became a bit smaller. Reimage’s junk-file and Registry processing might be slightly more effective than Windows’ free tools.

It was those stability and security numbers that really stood out. While playing up minor system hiccups might be a minor transgression, incorrectly flagging a system file as infected is, to me, unforgivable. At best, it suggests some faulty analysis by Reimage.

During testing, I jotted down notes about the experience of using Reimage, especially the non-numeric factors: things such as ease of use, speed, interface, and so on.

My notes mostly tell a story of overly aggressive repairs — such as its apparent replacement of system files. For example, after my first run with Reimage, the test system wouldn’t boot from the hard drive or via USB. My guess is that Reimage did something fatal to system drivers and/or to the boot records. (Again, Reimage provided few details about its repair actions and processes.)

By booting from a DVD, I was eventually able to restore the baseline system image and get my PC working again. I then tried to run Reimage again to finish the tests. But my bought-and-paid-for unlock key wouldn’t work. Apparently, any major change to a system — a system restore, restoring a backup or image, upgrading Windows, or reinstalling Reimage itself — invalidates the unlock key.

I acquired a new key from Reimage tech support, but that took four days, three phone calls, three trouble tickets, and having a Reimage support tech crawl around in the test PC via remote access. I did finally get Reimage to work again, but it was a huge hassle.

So on my test PC, Reimage was effectively a bust.

Of course, you might have an entirely different experience. As is the point of this series, you should do your own tests, record your own measurements and experiences, and draw your own conclusions.

If Reimage (or any other tool in these tests) works for you, then great!

However, make note of any serial numbers, unlock keys, or similar information related to your maintenance-tool purchase. You’ll probably need that info if you reinstall the program, ask for tech support, or request a refund.

On the other hand, should you decide not to keep Reimage — or any other commercial Windows-maintenance application — don’t bother uninstalling it. These tools make far too many changes to your system. Uninstalling them will not put everything back exactly the way it was. Restore your pre-test system image and go on to the next test.

(Note: I’ll go through the next two products much faster, focusing primarily on how they differed from the Reimage tests.)

Iolo’s System Checkup/System Mechanic

Iolo offers the free System Checkup (site) and the $40 System Mechanic (site).

The two tools work in series; you scan with the free tool to uncover problems and then, if you wish, pay for the full version of the software to make fixes.

For my Maintenance Challenge data points, I ran both tools, one after the other, on a freshly restored baseline system.

As Table 3 shows, System Mechanic trimmed a few seconds off the test systems’ restart time, and it also squeezed a few megabytes out of the Windows Registry. But as with Reimage, the improvements were small.

System Mechanic results

Table 3. Iolo's System Mechanic produced slightly better boot times and a modestly smaller Registry.

On my system, when compared to Windows’ free, built-in tools, System Mechanic wasn’t meaningfully better at reducing junk files and saving space — the test system’s overall file size was unchanged.

Like Reimage, System Mechanic also seemed to suffer from over-reporting. For example, it said it had found “10 core data conflicts within the Windows communications infrastructure.” Because there were no additional details offered, there’s no way to know what that really means.

Because the test system was working fine and Windows’ own tools reported no known defects, I assume the “core data conflicts” were the same sort of seemingly minor issues that Reimage reported — i.e., transient conditions and unimportant errors that really require no attention or repair.

System Mechanic was much faster than Reimage, completing its scan-and-clean tasks in about 25 minutes. System Mechanic’s license key is a lot more user-friendly, too. After restoring a system image on my test system, I successfully used the unlock key to reinstall System Mechanic.

As with Reimage, do your own tests and draw your own conclusions.

PC Pitstop’s comprehensive PC Matic

PC Matic is also a two-part system, with a free diagnostic component and a paid-for ($50 per year) repair component. Unusually comprehensive, the program even includes full-time antivirus protection. See PC Pitstop’s site for more info, a full features list, and download links.

Once again, I created a fresh copy of the baseline setup and immediately ran PC Matic’s free diagnostics. It completed the process quickly — just a few minutes — and produced a list of 25 items it said needed repair or improvement. It also listed one suggested “security adjustment.”

To its credit, PC Matic provides real information on its findings; it also tells you in advance what steps it will to take if you buy the full version. This is a refreshing change from the “black box” tools that reveal almost nothing about what they’re doing.

When I reviewed its list of fixes, I saw that PC Matic had recommended replacing several drivers on my system (see Figure 3).

PC Matic scan report

Figure 3. PC Matic is nicely transparent about what it has found and what it plans to do.

But if your hardware is working properly, there’s no reason to replace drivers. In fact, the act of needlessly churning drivers is an excellent way to screw up a perfectly good system. (See the Feb. 21, 2013, LangaList Plus column, “How and when to update your system’s drivers.”)

Because the test system’s hardware was working fine, there was no need to replace drivers. Deciding not to tempt fate — and possibly destabilize the test system — I opted not to buy and run the repair half of PC Matic. (Replacing the drivers was simply not acceptable to me, so I also took a pass on the other proposed fixes.”)

But again, you’ll likely get different results on your test system. Run PC Matic and decide for yourself.

Calling the shot: Only your results matter

Once you’ve tested a few paid maintenance programs and analyzed the results, it should be easy to select the best tool for your Windows system. You can also decide what’s most important to you: ease of use, features, information provided, and so forth.

Table 4 highlights the results of my tests. As you can readily see, there are no significant improvements in system start times, overall file size, or Registry size. Again, I tested an already clean system — according to Windows’ built-in maintenance tools. These numbers might be quite different if you have a poorly maintained PC — or they might not. Every system is unique.

Summary results

Table 4. The overall results were remarkably close and consistent. The best results are highlighted in green, but the differences are very small.

It’s the subjective areas where the paid tools come into question. In my tests, there were few positives and some glaring negatives — such as Reimage’s false-positive malware flagging, its boot failure, and licensing hassles. For me, the best choice is to stick with Windows’ built-in and free tools. Using the paid products would not only cost money I don’t need to spend, I might end up wasting time researching whether a reported problem really needs fixing. And in the worst scenario, the paid product might crash my system.

But that’s me. By doing your own tests, you’ll know whether any of the paid products significantly improves your system’s performance, security, and stability. It might be tempting to use a commercial product to scan your system for free and then use Windows’ built-in tools to fix any found problems. That might work with an informative product such as PC Matic, but it’s not feasible with tools such as Reimage and System Mechanic.

If you decide to stick with a commercial maintenance tool, make a full system backup before you install and run it for the first time. Then make another full backup of your newly cleaned system — with the commercial tool installed and set up.

If you decide not to keep any of the commercial maintenance products you tested, just restore your pre-test system image; you’ll be back exactly where you started — with a PC that’s been thoroughly cleaned with the Windows-based tools discussed in Part 1.

Either way, you win! Commercial tools or free — your PC is now leaner, cleaner, more secure, and closer to error-free than it was before you took the Maintenance Challenge!

Subscribe to our Windows Secrets Newsletter - It's Free!

Get our unique weekly Newsletter with tips and techniques, how to's and critical updates on Windows 7, Windows 8, Windows XP, Firefox, Internet Explorer, Google, etc. Join our 480,000 subscribers!

PC Drive Maintenance (Excerpt)

Subscribe and get our monthly bonuses - free!

Your hard drives store photos, books, music, and film libraries, letters, financial documents, and so on. This ebook is aimed at helping you understand your hard drives, expand their capacities and length of life, and recover what you can from them when they fail. We're offering you a FREE Excerpt! Get this excerpt and other 4 bonuses if you subscribe FREE now!

= Paid content

All Windows Secrets articles posted on 2014-08-28:

Fred Langa

About Fred Langa

Fred Langa is senior editor. His LangaList Newsletter merged with Windows Secrets on Nov. 16, 2006. Prior to that, Fred was editor of Byte Magazine (1987 to 1991) and editorial director of CMP Media (1991 to 1996), overseeing Windows Magazine and others.