Our last post on the number of online stores in the U.S. turned out to be quite a hit. So we decided to update it this year with Internet Retailer’s new Top 500 Guide for 2013 to see if there was anything new we could discover. Would we find the same power law? Would there be more stores this year? We were curious to find out.

We dug into the problem the same way as we did last year (see how we did it here). A quick trip to the U.S. Census Bureau also let us know that the total Ecommerce sales for 2012 grew to $225 billion (a 13.6% increase from $198 billion in 2011).

Crunching the numbers, we plotted a new graph of the top retailers and their revenue for 2012.

Some interpolation of the data points to figure out the new curve (hat tip: Excel) gave us the following:

Interesting… we got a different power law function (in red) this year but it still fit the data points really well with an R-squared value of 0.99.

With the new formula, we also had a way of figuring out how many stores had a least some given amount of sales revenue for 2012.

We put these new numbers alongside the ones for last year to get:

Yearly Sales of at least | Number of retailers in 2011 | Number of retailers in 2012 | Year-on-Year Growth |

$12,000 | 90,501 | 102,728 | 13.5% |

$25,000 | 54,686 | 61,728 | 12.8% |

$50,000 | 33,983 | 38,157 | 12.3% |

$100,000 | 21,118 | 23,587 | 11.7% |

(As a sanity check, we also added the year-on-year growth rates and they were roughly in line with the overall ecommerce sales growth rate of 13.6% for 2012. Nice!)

While admiring our handiwork, one kink in the graph stood out. Most of the derived power law curve fit the actual Top 500 list quite nicely. But as it started approaching the 500th store, it looked like the actual revenue numbers were starting to dip away. We saw this trend last year as well. Was this cause for concern?

Well, probably not. Here’s why. It would be pretty easy to pick out the top 10 or 50 retail stores in the U.S. since everyone would have heard of them. Once you got out to the 500th store though, it would start getting tricky.

According to the Internet Retailer Top 500 Guide, the 500th store makes $18,690,000 in yearly revenue. But since not all stores need to disclose their revenue, there might be a store somewhere that made more than $18,690,000 in 2012 which the Top 500 Guide missed out. And if it was included in the list, the curve of actual store revenues would dip down at a slower rate and would be a closer match to the power law curve.

So it looks like there are more online stores now and ecommerce is growing! The future certainly looks exciting.

_____

*Read next: The Biggest Opportunities In Ecommerce Right Now, As Revealed By Retail Experts [Infographic]*

Isn’t 23,587/21,118-1 equal to 11.7% growth?

Fixed. Thanks for catching that!

Hi, can you please explain how did you use the power equation?

The power equation (http://en.wikipedia.org/wiki/Power_law) lets us understand how the revenue of an online store varies with its rank. We then used it to predict the rank of a store with a specified amount of revenue.

Thanks. It’s still unclear to me, but I will try to understand.

WHAT IS THIS POSTI I DONT UNDERSTANDS,, PLEASE SEND ME THE DETAILS..

http://cikyanzone.blogspot.com/

WHAT IS THIS POSTI I DONT UNDERSTANDS,, PLEASE SEND ME THE DETAILS..

http://suffianzone.blogspot.com/2013/10/gempa-bumi-di-pulau-bohol-filipina.html

Apologies for the simple sounding question but are the numbers inclusive? For example, does the 102,728 stores making more than $12,000 include the 23,587 stores making more than $100,000? Or is the 23,587 specifically stores making between $50,000 to $100,000? Many thanks again for this – it’s very interesting!

Yup, it’s inclusive- any store making more than $100,000 is also counted as making more than $12,000! 🙂

This is a great Calculation. Thanks.

I’m trying to find out, of those numbers you came up with, what is the marketplaces’ share (e.g. AMAZON, eBay, ETSY), compared to all other, more open, solutions’ share (Magento, Shopify, Independent). And even more than the distribution of stores, what is the distribution of revenue, which is something you don’t relate to here.

Hey Otto! We haven’t really explored that yet, you’re right.

What I can tell you though is that Amazon’s share of the pie (which is expanding, of course) is absolutely massive. I believe the Top 500 retailer list shows Amazon’s global revenue to be 61 billion USD last year, while less than 10 retailers make 1 billion or more. It’s really a “winner-takes-most” phenomenon.

But we’ll try and explore this in greater depth!

Can you recommend on some of the resources you use? I’ve been exploring this, and can’t find even an estimation. Most likely because there’s no clear distinction where the line draws between the ‘closed’ platform such as Amazon and eBay, and the ‘open’ ones such as Magento and Shopify. My final goal is to realize what is my potential market share when approaching ‘open’ platform stores – that have the ability to use our tools.

I am trying to calculate the humber of US online retail stores with $1M and over in revenue and the same for those that are $5M and over in revenue. Can you tell me how to build the same formula in excel?

Hi Kieran,

You would be able to get the rank of the store making $5M in revenue by using the following: ((2*10^11)/5000000)^(1/1.441)

Hope that helps!

Hello, Dinesh – this is a very interesting post, however, when I try to replicate your findings in excel, I come out with other values for the power law formula you have above. For example, I come up with 12,475 stores at the 100K mark. Here’s the formula: (2*10^11)/S15^1.441 with s15, in this case, equal to 100,000. I am rather puzzled by this, and would appreciate hearing back from you.

Thanks,

Owen

Hi Owen, I’m coming out with the same results as you…essentially a more steep power law slope than the numbers Dinesh came up with. Ex. annual sales of $100,000 = 10,376 retailers. Any other insight since your post?

Correction: $100,000 = 12,475 retailers. 10,376 was using 2011’s 1.457 number instead of 2012’s 1.411.

Hi Owen and Tim,

Thanks for checking out the article!

It looks like you’re getting a different result since you’re setting the ‘x’ value of the power law equation to the revenue amount instead of the ‘y’ value. Since the y-axis of the graph above tracks revenue, to get the rank of the store making $100k in revenue, you would need to invert the equation and use the following: ((2*10^11)/100000)^(1/1.441)

Hope that’s clear! Let me know if you have any other questions.

Hi Dinesh,

Very useful article for a model I’m working on. Owen Brown who posted below and I coincidentally are finding the same results and are posting hours apart from one another. My results produce a more steep slope than yours.

Ex. $100,000 = 12,475 retailers, and $12,000 = 264,807 retailers.

Would you please clarify my assumptions and take a look at my formula (basically the same operations as Owen) and help me obtain the same results as your figures?

Assumptions:

y = No. of retailers in 2012

x = yearly sales

Excel Formula:

y=2*10^11/x^1.441

Thank you,

Tim M.

Hi Tim, I added a reply to the comment below. Let me know if you have any other questions I can help you with!

Thanks for the fix! Got it now.

Hi Dinesh, how would you project the number of retailers going forward, say from 2013 through 2018?

Also, any insight on how to project number of retailers at a certain annual revenue level? For example, number of retailers with yearly sales of at least $5MM from 2013 – 2018.

Hi Dinesh,

Just a small clarification I need as I have number of retailers doing $1 Million and number of retailers doing $2 Million but how can I calculate the range. For example how many retailers are doing between $1-$2 Million?

Hi Dinesh,

I have a question regarding how to calculate number of retailers in certain revenue range. For example, how many sellers are having revenue less than $100,000?

Can you tell me how many online retailers there are at $100 million and above

This was great! Can you separate the data out by state to get the number of online stores in each state?

Hi there – first of all, sorry for the late comment – I just came across this and it’s almost exactly what I was looking for so thanks for this great analysis! Second, I wonder if you can explain your methodology. The link in the post suggests that you explain it in 2012’s analysis, but that’s a dead link (http://privateblog.referralcandy.com/2012/08/14/how-many-online-stores-are-there-in-the-u-s/) Is it a simple linear regression? How did you translate the rank to absolute numbers of retailers in each category? And why did you choose to break up the revenue into the categories you did? This does a great job at showing the volume of the long tail, but $100k yearly revenue is really low for the many big guys out there so you’re overgrouping a lot of significant players in a big bucket. Anyway, thanks again and I hope you’re still monitoring this post 🙂

Oops, thanks for pointing out the broken link! Fixing it now. Here’s the original post: https://referralcandy.com/blog/how-many-online-stores-are-there-in-the-u-s/

Hi Referral Candy!

Great Article. My question is how you are defining “online store” does it include ecommerce sites selling services or content, or does it only refer to retailers selling physical product?

How many “Fashion” [apparel, footwear, home, accessory, kids] etailers are there Globally between 10K-500m and what is the combined gross revenue? and Growth Rate?