### The answer, according to this calculation: 12-24 million, but only 650,000 generate annual sales of more than \$1,000. Read how this online entrepreneur calculated those figures.

We’re already in the holiday season of 2014, and a question we get asked a lot at DandyLoop is ‘how many online stores are there?’, not only in the US, not on a specific platform, but rather all online stores. So, especially for the coming holiday season, we checked:

The number of online stores isn’t directly monitored by anybody, so what we can do is calculate. Since there are many inaccuracies and assumptions, and yet to answer this questions as accurately as possible, we use 2 existing analyses and utilize two angles of attack.

1st angle of attack: The Power law

In a remarkable post, Dinesh Raju took the US top 500 retailers with their respective revenues, and plotted a power rule from which we can deduce the overall number of online stores based on their revenues even if they’re not one of the top-500.

We can learn that in 2013 (the year in which that post was written) there were 576,000 online stores in the US with more than \$1,000 annual revenues. However, the field is growing fast – so how many stores are there now? Applying the 2012-2013 growth rate (13.5%) will result in 654,033 online stores in 2014 with more than \$1k annual revenues.

But wait, this is based on US top 500 retailers, not the worldwide top 500. Does this mean there are 654,033 online stores in the US alone? Well, being the dominant market, there is not much difference between US top 500 retailers and the world top 500 retailers – the vast majority of them are Americans (plus, the power rule calculation ignores the top 10 stores so Alibaba is not calculated). This means we can assume the worldwide power law is close enough to US’s for us to take them as one. Meaning there are 650,000+ revenue generating (\$1k annually) online stores.

Want to know how big your store is compared to all the stores in the world? Use our calculator

2nd angle of attack: The Survey

According to a survey conducted by Tom Robertshaw, we learn there are about 47,000 online stores in the Alexa top 1 million sites, which is 4.7% of the million biggest sites on the internet.

Does that 4.7% ratio of the top million sites continues to all internet sites? First, let’s talk about what does ‘all sites’ mean: The Pareto principle (80% of the effects originate from 20% of the causes) is extremely distorted once we go online – trailing the top websites, there is an extremely long tail with almost no traffic. For example, online advertisers have very little interest in sites with an Alexa rank higher than 50,000 (over 99.995% of sites). The long tail of websites’ size according to their Alexa rank starts very soon in the count and by time it reaches 1,000,000 there’s very little change in the size of sites and the graph is practically flat. This leads us to assume that the 4.7% ratio of ecommerce to regular sites we see in Alexa top 1m is kept for Alexa 10m and also 100m count (i.e. – 4.7% of all websites are eCommerce sites)

Let’s go mathematical and validate this ratio in front of the power law (from our first angle of attack) using our own data: From what we know from DandyLoop’s clients: stores with annual revenues of around \$20k, have an Alexa rank of 1,500,000-1,900,000 (an average of 1,700,000). Crunching \$20k annual revenues in the power law results in these stores being the 72,000 largest eCommerce sites. 72,000 online stores spread across 1.7 million sites result in a ratio of 72,000/1,700,000=4.24%, pretty close to the survey’s 4.7% estimation which we were trying to validate.

Now the question is how many online sites are there worldwide. According to TekEye there is anywhere between 108,000,000 sites to 860,000,000 sites worldwide. TekEye explain the variance in that the higher figure counts the sub-domains (e.g. xxx.google.com), and the 108m counts as the number of hosts, making the real number of sites likely to be much closer to 108m than to 860m.

Based on the 4.7% ratio, 108m sites gives us 5m online stores and 860m result in 40m online stores. I can try to speculate that the result is most likely somewhere in the low-middle, meaning there are probably 12-24 million online stores, from which only 650k have an annual revenue higher than \$1,000k – this will make sense given the long tail nature of online stores in the top 650k.

Want to know how big your store is compared to all the stores in the world? Use our calculator

Summary:

Based on the power law pulled from US top 500 retailers we calculated there are slightly more than 650,000 revenue generating (>\$1k annually) online stores worldwide.

Based on a survey for the number of online stores in the Alexa top 1 million, the number of websites on the internet, and a sample of stores which use DandyLoop, we calculated that  the overall number of online stores is 12-24 million.

We see eCommerce is growing in an incredible rate. With the proliferation of tools that lower the barriers for opening online stores – this growth of eCommerce can be shared by anyone, not only the big players.

Assumptions used for the calculations:

1. US top 500 retailers ≈ WorldWide top 500 retailers (not including the top 10 in each list)
2. The power law that applies for the top 500 retailers is effective for rest of the list.
3. The ratio of webSites/webStores in the Alexa top 1 million is the same for all webSites/WebStores.

DandyLoop is a network of online stores that expose their customers to each other’s e-commerce sites.