“Can you help me boost my Alexa ranking fast?”
The novelty was her motivation made sense: “I need an Alexa rank under 1 million to put a query up on HARO.”
More often, small business owners fixate on this number for no clear reason at all. The reasoning seems to be that, because good ranking indicate website popularity (true enough), it should be a goal in itself (which makes no sense at all).
But did I have an answer for her? Honestly, it just wasn’t my jam.
The best I could suggest was to install the Alexa browser extension and get back to promoting content to real readers: “You might as well let Alexa see you load your own website. Then get on with the rest of your business.”
The short answer is no.
How does that work? They insert a tracking code into a wide range of browser extensions: the branded Alexa add-on I mentioned just before, and thousands more. It’s not too different to how ratings agencies have long tracked TV audiences.
Some websites also choose to install an Alexa tracking script. That means Alexa can measure how their browser extension data skews and correct for it.
What this all amounts to is an independent estimate of total traffic volume. It’s not perfect, but for popular websites, it’s actually not too bad.
Still, analytics software is always a better, more direct measure of your own website’s traffic.
What’s more, few businesses rely much on total traffic volume. What matters is traffic quality: those particular click paths that bring buyers.
So why even care about Alexa rank? Because others look at it.
It can help attract new advertisers, and it can help interest the busy bloggers and PR types who matter to your content marketing.
It makes sense that installing an extension could visibly boost an Alexa ranking in the millions.
When you’re actively promoting a website, you regularly load it up to check things, monitor comments, add new content, copy and paste URLs, that sort of thing. It can only help if Alexa sees these page views.
Still, I was surprised to hear how effective it was. In a week, her ranking had jumped more than a million places to under 2 million.
Still, she hadn’t yet hit her target. “I'm going to try one of these traffic exchanges to get me over the line.”
I begged her to reconsider. I’d honestly no idea what to expect, but these schemes can backfire.
The best I could get though was a pinky promise that she’d only take part for an hour or so, and then wait a few days to see how it went.
If one tracked user could boost you a million places, 200 would skyrocket it, right?
After a few days, her rank actually dropped a negligible 2000 places.
It looked that way. Or was she far too timid? When my curiosity gets tickled, I can’t help but poke around.
It should be easy enough to take a look at what’s going on here. The central premise of a traffic exchange is that the participants visit each other’s websites. So scoping out the other users should be as easy as joining in and watching where it sends us.
That’s how I found myself signing up to Rankboostup. First, I ran it in a virtual machine and watched with my own eyes.
This uses a browser extension to surf the web automatically, first checking that the Alexa add-on is already installed so it’s tracked. About every minute, it loads up a new URL in the next tab. The longer you leave this, the more credits you earn for others to visit your website in the same way.
I also watched it load Google, search for a phrase then click on a result. I’m guessing this is an attempt to game Google’s engagement metrics by boosting the click-through rate.
Watching this run, I noticed two websites appear regularly. One, a garishly designed news aggregator for cryptocurrency and blockchain with an Alexa rank of 3,218,108. News aggregators need to pull a fair amount of traffic to be worthwhile, so that’s honestly quite unimpressive.
I was happy to watch this for a few minutes, but all day? Time to write a script.
Using Python and Selenium, I was able to run this Rankboostup extension headlessly in the cloud and record where it goes. (Hat tip to Recycled Robot for the amazing help when I got stuck with part of it :)
My first effort revealed that, while this autosurfer runs happily for a while, you can’t leave it unsupervised forever. Every so often it hits a Google captcha and stalls.
I tweaked the code to handle this by closing the browser and lying low for a while before resuming. I left it to run like this for a weekend and came back to about a day’s worth of data.
Here were the top 20 most visited websites in this time:
Two things about traffic exchanges are now clear: they’re popular with stoners, and no guarantee of an impressive Alexa ranking.
Techquark.com performs best here. My hunch is that this rank comes from somewhere else. With almost 2000 articles published over many years, it might even be from real traffic.
The most damning statistic? You can be in the top 20 most visibly active users and not rank anywhere in Alexa’s top 30 million at all.
That doesn’t look good for Rankboostup. But I’d now earned many hours worth of credits. Let’s give it one last chance.
I have a website from an old, long abandoned project in my back pocket, which I’ve kept just to test things on. With an Alexa rank of around 9 million, it really shouldn’t take many visits to move it up significantly. It’s been dormant long enough that there’d be no other reason to see a short, sudden ranking improvement.
So let’s jump right in.
Was there traffic? Yes. Thousands of hits started arriving the moment I turned on the tap.
But would it move anything? I waited and waited and waited. Days passed. A week. Nothing.
This looked conclusively busted.
Doing this definitely stretched my simple Python skills. But with a little help from Reddit, it was totally feasible for me – a freelance writer who likes computers – to see what was going on inside this traffic exchange.
So how do you think it would be for a veteran Silicon Valley analytics team backed by a trillion dollar company? Would they struggle to keep up?
The entire concept of a public traffic exchange is flawed. How do you get total strangers on the internet to visit each other’s websites without disclosing where they’re going? And if you open that to everyone, what’s to stop Alexa?
It wouldn’t be difficult to unravel the exchange from the other direction too. You might build honey pots: quick and cheap websites on brand new domains whose only purpose is to lurk in these traffic exchanges. As the hits roll in, the participating browsers shine bright like lights on a Christmas tree.
But would Alexa even need to infiltrate? This software generates an utterly monotonous, regular pattern of traffic. That can be profiled.
Real human traffic is irregular. People click internal links, leave comments, upload memes, all sorts of things. There are short and long periods of inactivity: you make a cup of tea, stay on long articles for a while, switch to different applications or get distracted by a boss, phone, spouse or cat.
Alexa has so many options for filtering out this traffic that it’s hard to say which they’re using.
Across facebook groups, internet forums and Fiverr, there’s a grey market of services that also claim to boost your Alexa rank. Are these any better?
On Fiverr, I found a US$5 gig that promised to send “10,000 real, human visitors” to a website. At this price, how “real” or “human” can these visitors be? Still, the ad does promise to lift your Alexa rank, and the vendor has a perfect 5 star rating from hundreds of customers.
By this point, a couple of weeks had passed since my experiment with the traffic exchange; any sudden, sharp boost would surely be the Fiverr gig.
When I pulled the trigger on a Monday morning, it ranked at 9,477,457. By Wednesday, it was 9,463,359.
Do we just need to give this more of a chance? Friday rolled around: 9,473,489.
Yeah. I’m calling this a total bust too.
Actually, I wouldn’t be surprised if someone had something that worked.
But how do you know who’s got the goods? This vendor had glowing reviews. Would someone shady enough to sell fake traffic be more scrupulous about their reviews?
If you buy something that doesn’t work as advertised, there’s always the refund process. But that’s one more hassle. For the sum involved, I couldn’t be bothered. I wonder how many other people decide the same thing?
In a real business situation, you’re (hopefully) also actively promoting your website to a real audience. With that in the mix, you might not even know where your boost comes from.
In any case, these services are available to the public. Alexa can at any time buy them and profile them. Even if it works, it has a shelf life.
This was giving off a vibe. If I wanted to pull this off, some secrecy wouldn’t hurt. What’s more secret than coding it yourself and telling nobody?
So I built a little bot, again using Python and Selenium, and put him in the cloud. This little dude lives on a server in the Netherlands where he drives a headless Firefox browser with the Alexa toolbar installed.
Instead of just firing off a whole lot of hits to the target website, I built a backstory – I had him crawl around Wikipedia for a couple of hours each evening, spending a few minutes on each page, a bit like I do now and then. I also got him to browse Google News in a similar way.
3 days in, I added the next bit: I got him to lurk around on the blog section of the test website, again spending a few minutes on each page.
The result? Success!
In just two days, it had jumped 9,474,764 to 6,620,738 – nearly 3 million places!
As I kept it running, it kept climbing.
With a little thought to stealth and operational security, Alexa rank can be gamed after all.
It’s definitely possible, if not quite as straightforward as pressing a button.
I see big downsides though.
What I got away with here is fake traffic at the very smallest scale. It makes a meaningful difference if your ranking is in the millions. That’s useful for the newest or smallest operations.
If you’re in that situation and have the Linux, coding and internet savvy to put this together, I suppose there’s not a lot to stop you.
But how many small operations have these skills in house? Without them, it’s a frustrating thing to outsource; it’s hard to judge the work you get.
What if you scaled it?
Well, my lone robot assassin managed to pass for human for a few days. But hiding a single bot in an entire public internet for a few days is playing this game on the easiest setting. If you left many identical instances running for months, they’d be vastly more vulnerable.
I mean, they should definitely all browse the web a little differently. Each might have its own interests: different hobbies and news stories that they follow, holidays in different locations to plan, checking opening hours of shops in different areas. It’s a lot of scripts to manage.
Real humans also log into accounts. So your bots should do that too. Many websites won’t be all that happy about that. That means it’s not just Alexa you’re gaming here.
You might even want some of your bots to look like they’re doing a bit of banking. It’s hardly suspicious if just one or two tracked users never bank online. Maybe they prefer the app on their phone. But if a much larger tracked audience doesn’t? Then they’re either all New World Order conspiracy loons or they’re bots.
And as all these websites change their layouts periodically, you’re going to have to update your scripts.
Of course I’m wildly speculating. I don’t actually know where the red lines are. But neither do you, and you won’t figure it out without a lot of testing and effort.
Alexa has so many options for profiling fake traffic. How do you anticipate which they pick next? Guarding your investment means protecting against as many as possible.
And then what about all the browser extensions to juggle? If you’re sending all your fake traffic through the one extension that has their name on it, ignoring their 25,000 other extensions, do you think that it might look suspicious?
Unsurprisingly, they’re not crystal clear about what these extensions are. You’d need to track them down yourself. You might automate that, perhaps by writing scripts to install and run extensions and then monitor which servers they communicate with.
You’re going to need to stay on top of every part of this as Alexa evolves.
That’s not impossible.
It’s just quite a project.
Clearly I felt comfortable testing these things. But to promote a real business like this?
If you fake your Alexa rank for PR purposes, you might rationalise that you’re not lying directly. After all, isn’t the whole point of PR to put the best spin on things? That’s a nice justification for what’s still dishonesty.
It’s arguably worse to fake your Alexa rank to dupe advertisers because money changes hands.
If you game your ranking too shamelessly then some advertisers, surely, would smell a rat. In any case, once the ads are running, they’ll see how many impressions, clicks and purchases it was really worth.
The real reason to resort to things like this is to cut corners on your content.
That means your website will suck harder than it should. Talented and motivated people will be less excited to be involved with what you’re doing.
This is the big one. All that time and effort spent to please your new robot masters does nothing to satisfy the audience that matters most: paying customers.
Imagine if you put this might actually click on, engage with, upvote and share. Do you think they might also help your Alexa rank? Of course. As a bonus, you’re developing a genuine, engaged audience and a brand, and learning what your audience responds to.
How many of your bots are going to buy anything from you or your advertisers? They don’t even have credit cards.
Sure, some of your fellow marketers look at Alexa rank. But none of them ever got far by outsourcing all their judgement to this number.
Alexa rank is most useful as a broad filter to work out which websites merit further attention. It opens doors rather than closing deals.
Sure, you could game your Alexa rank. But it’s a lot of effort for a very narrow objective.
And each time you come back to this crack pipe, it’s going to take a bigger and bigger hit to get you where you want to go.
After all this, I find myself returning to my original advice: just install the browser extension and move on. Get straight back to crafting substantial, engaging content and promoting it to real people.
Installing the extension can nudge the number in the right direction, and it’s even useful sometimes.
The best thing though is that it’s dirt simple and takes little of your time. So you can get on with the thing that matters most: bringing actual humans to your website.