Google and the "Snowball Effect" (exponential site traffic growth)

[aristotle]

Most people hope that the traffic to their website will grow over time. The natural way for this to happen is for people to discover the site, like it, and spread the word about it in various ways. Of course such growth can happen at different rates, depending on the site's niche and the degree of general interest in the topic. The "snowball effect" usually refers to cases in which the growth is exponential. But even exponential growth can happen at different rates, even rather slowly. Of course a site's traffic can't grow exponentially forever, since the population of the earth is finite.

Occasionally someone here at WebmasterWorld mentions that Google's algorithm might take account of the rate of backlink acquistion in proportion to a site's traffic. For example, A site that gets very little traffic can't be expected to rapidly acquire lots of new "natural" backlinks [This assumes that no artificial "link-building" takes place]. This also applies to social button signals, etc. So if everything happens naturally, there should be a correlation between the amount of traffic a site gets and the rate of backlink and social signals growth.

Exponential growth in traffic (the snowball effect) will occur if visitors to a site tend to spread the word in ways which, on average, bring in a steady number of new visitors. Thus the more visitors it already has, the more new visitors it will gain, in proportion.

In most cases Google can use its Chrome browser to get a good estimate of a site's traffic and rate of growth. If traffic is growing, the cause needs to be determined. For example, traffic growth due to increased Adwords spending should to be discounted.

Snowball-type growth is probably more likely to happen with some types of sites than with others. Also, there's a big complication, because changes in Google's own rankings for a site can have a strong feedback effect on its traffic, since Google is so dominant in search. On the other hand, if a snowball effect is occurring, Google's algorithm might take it as a positive signal. Then the resulting rankings improvement would reinforce the snowball effect, and maybe even accelerate it.

[aakk9999]

there should be a correlation between the amount of traffic a site gets and the rate of backlink and social signals growth

I would imagine some part of Google algo is looking at this correlation, probably in not so simplistic way.

[adder]

From the anecdotal evidence and conspiracy theories (like this one: [siliconbeachtraining.co.uk...] ) you may gather that there is a correlation but I'm yet to hear it from a real Google employee (ex ones don't count)

I don't think Google uses Chrome signals as a ranking factor. And even if it does, it must be a weak ranking signal. Otherwise, I could build a network of 1000 people, make them install Chrome, visit my site, spend a lot of time on it and then visit the competitors' sites and do a quick bounce. And then delete the cookies, change the IP address and do it all over again.

One of the sites I look after for somebody has a very low percentage of Chrome users (45% use IE) and the site hasn't got Analytics installed, yet does particularly well in organic search. Google has not got a foggiest idea as to how much traffic this site really gets but we are sort of digressing here :)

I think to sustain 'Exponential growth' you need to be on a constant lookout for a new channel for driving traffic. You tend to exhaust them in the long run.

since the population of the earth is finite

Ha, interesting thought. I would argue that it is infinite!
Think about this:

• Last year 583,000 new users joined the internet every day! That's half a million new users every single day!
  
• Here in the UK, one in eight people don't use the internet. They will have to start catching up sooner or later.

The challenge is understanding how these new users are using the internet and how you can enhance their experience. And once you crack it, Exponential growth is a possibility.

[samwest]

@ aristotle : This is a great topic and one I've pondered for the last five years and have experienced myself. My site experienced a very nice and natural rise in traffic and sales between about 2001 and 2010. The graph I have for traffic and sales between that time frame was up, up, up! It was a beautiful sight for any business owner. I loved Google! The marker for the coming decline started EXACTLY on May 18th 2010 a day that suddenly had ZERO sales. It was as obvious as a bomb drop. Since then, every year has declined slowly and had a "melting snowball" effect. The site suddenly lost all traction (no pun intended) and continues to slip.
The negative effect of this has had a trickle down effect. I eventually (recently) had to stop spending on Adsense and even buying new software and office equipment. I had to refi my house and sell off a ton of assets. As our sites continued to drop, the search engine began making obscene amounts of cash. Now it doesn't take a genius to identify what is happening, Just follow the dollars. If websites are making 20% less and a search engine is making 20+% each quarter to the tune of billions. It's pretty easy to find your problem.
Because the search engine runs a black box culture and algos, we are prevented from knowing for sure. My bet is that they use many signals to detect and prevent the snowball effect from ever happening. Unless of course that is a very well funded snowball. It may be sickening, but it's the present American way. Many in the webmaster community have lost trust, respect and love for the once great search engine.

Consider this quote from Mr Schmidt : "…someone, somewhere in a garage is gunning for us. I know, because not long ago we were in that garage. Change comes from where you least expect it."
Because Google wants to always remain on top, it is my opinion that they will always be watching and wary of the next big thing (snowball) then shoot it down or quickly copy it. It's what they do.

[brotherhood of LAN]

since Google is so dominant in search

I think Google will have a good handle on what % of visitors to a site are delivered by Google itself, for any site that has a reasonable volume of traffic, i.e. signals that indicate it's not entirely reliant on organic results.

>Chrome

There's also their public DNS too, which serves 400 billion requests a day. [en.wikipedia.org...]

[frankleeceo]

Exponential growth exist (percentage wise) to a certain point. Then it's capped, either by natural competition, google, or some sort of invisible hand.

For example...2~10% increase week over week for 1~2 years or more, then stops. Then slow decline. After noting it on 10+ sites.

Maybe's it's just how people function?

[aristotle]

One signal that Google's algorithm can detect is when ever-increasing numbers of people type the site's name into Google's search box. This would be a reliable indicator of growing interest in the site.

samwest -- Don't you think it would be ironic if Google's algorithm penalized a site that has rapidly growing traffic. In the typical case I had in mind, the initial traffic growth will happen without much help from Google.

[samwest]

Aristotle - In my own case, the site growth occurred well before Google had mastered any method of throttling, so there is was no irony. It was purely natural growth. I do find it ironic that so few sites can show 20+% quarterly growth or even go viral these days unless it's very well funded...or about a scandal.

[samwest]

What you're proposing is that site traffic could take off like wild fire. Even if you have a field of bone dry switch grass, 40 mph winds a can of gas and a match, your fire won't burn if it's hosed down with fire retardant foam. Google and their zoo full of penalties is that foam. You need a natural, unfettered medium. G is no longer that medium.

When you start factoring in back link profiles and algorithmic feedback, you are no longer playing on a "natural" field. I personally doubt Google uses high search traffic, not even high direct traffic as a positive signal. In my experience, I've seen high traffic periods quickly capped as if I hit some invisible glass ceiling. Others have noticed the same. Throttling has pretty much been observed and accepted by all here. This never happened prior to 2010. A good question is then "why would they throttle"? (sound of can of worms being opened)

To snowball, you would have to rely on the news media (TV and radio) to spread the word. People would have to find the site by directly entering the URL, then LOVE it to make it roll. Personally I would love to see this happen with a new search engine. Perhaps we should all pool our resources and start an experiment to see if your snowball theory can hold water?
If G read this, they could have foamed it down already. :)

Great topic tho...

[Awarn]

I do find this very interesting. I too had consistent 25% growth for years. Mine went up until May 7, 2012. Now no matter what we do the homepage will not rank on page one. I can get an interior page to page one. I can get other sites to page one but not the main site. I believe Google works in cycles. I suspect the great mobile algo that is so publicized might be a correction or possibly the removal of some penalties that they know will shake up the SERPS so lets blow smoke its mobile related. If traffic is lost on one site it has to go to another. The Adwords were not so fabulous that people were jumping at those so what happened to the traffic? I suspect it is more that Google is not fully indexing sites. Oh sure the WMT shows they are indexed. But that may mean they know the page exists. There is a huge difference between a quick skim and fully digesting a page. I have been watching Google digest structured data for the past few months and I find it extremely telling of how much they actually read pages. It is definitely time for a new search engine.

[aristotle]

samwest -- You make some good points. And I agree that it's very difficult for a site to enjoy rapid traffic growth without any help from Google. But I don't think it's impossible.

My question is how does Google's algorithm react when it discovers a site that is experiencing rapid traffic growth without its help. You seem to be saying that Google will somehow try to stifle it. To me, that's illogical. Instead, I think that Google will have to acknowledge that people like the site and thus throw its support behind it by sending additional traffic to it. That's what I tried to say in my first post.

[samwest]

My question is how does Google's algorithm react when it discovers a site that is experiencing rapid traffic growth without its help. You seem to be saying that Google will somehow try to stifle it. To me, that's illogical.

It's illogical to me too! Until Google releases the official book of algorithmic specifications, all we can do is speculate. I spent most of my career troubleshooting systems. Many without a manual. For a black box, you have to go by the observed input vs. output, consider any other external influences then come to some kind of deduction. I'm not hating on Google, I'm only looking at the obvious and making my own personal evaluation.

In my own vertical, I have witnessed one site post 2010 go literally viral and snowball as you describe. That site started out with 50M in VC, then received over 150M in additional VC. It now sits in the top 500. All than in 4 short years. Anything is possible with the right size briefcase. Some might call that gaming the system, others call it business as usual.

[edited by: samwest at 4:38 pm (utc) on Apr 14, 2015]

[brotherhood of LAN]

>Google's algorithm react when it discovers a site that is experiencing rapid traffic growth without its help. You seem to be saying that Google will somehow try to stifle it. To me, that's illogical.

Stifle aside, there are still reasons why it would "rate-limit" exposure of a SERP. Namely whether the hosting is capable of a surge in traffic, whether the server is responsive from locations farther afield. I'm not saying that's how they do it, but it's a plausible reason for all the positive ranking increase signals not being directly transferred into actual ranking increases.

[samwest]

BTW - the site I'm referring to claims their success is based solely on the three "C"'s Content, Commerce and Community. I tried that, didn't work in the long run. I forgot there was one more C....cash.

For all these hot start ups, the key is "free". While I made money from day one, these guys dumped in amounts that any sane person could retire on for life before they even made a dime. At my peak in 2010, I was over six figures. A one man band. Since then, I'm back to Ramen Noodles. Take that as sage advice. Get a team, get financing and start free. It won't always work, but today it's pay to play. IMHO, that free internet thing is long gone.