Scaling can be tricky for consumer internet startups, especially when it comes to hardware and infrastructure.

Many startups have failed in part because they couldn’t successfully scale effectively and sensibly.

I’ve argued that scaling is a significant problem for some of today’s most popular internet startups, including Facebook, and that it rapidly poses significant risks, especially for ad-supported businesses.

So it was with interest that I read Gordon Haff’s article on SmugMug’s use of Amazon’s S3 hardware-as-a-service offering:

“Amazon S3 provides a simple web services interface that can be used to store and retrieve any amount of data, at any time, from anywhere on the web.

“It gives any developer access to the same highly scalable, reliable, fast, inexpensive data storage infrastructure that Amazon uses to run its own global network of websites. The service aims to maximise benefits of scale and to pass those benefits on to developers.”

In other words, Amazon S3 does for storage what Salesforce has done for CRM software.

SmugMug provides an interesting case study of storage-of-a-service because it’s a rare example of a “Web 2.0” company that has implemented a highly-successful business model that actually makes money.

Despite competition from a plethora of free online photo services, SmugMug has built up a base of 350,000 customers who pay between $39.95/year and $149.95/year to use its service.

It offers none of its services for free (outside of a 14 day trial), hasn’t raised a cent of VC funding and is profitable to boot.

All of this makes it an oddball in a Web 2.0 world where VC funding is equated with validation and where advertising is the default business model.

Clearly, SmugMug is run by some savvy people and their decision to leverage Amazon S3 was made with the bottom line in mind:

  • The company wanted to reduce the number of servers it was having to buy, which was creating significant space and power issues.
  • Its expansion of storage capacity in-house resulted in increased capital expenditures, which had a significant impact on cashflow and profitability. Using S3 significantly eliminated many of the negative financial aspects of this type of expansion.
  • SmugMug essentially gets backup storage capacity for free. As noted by Gordon Haff, “Were SmugMug to host all this storage in-house, it’d actually have to buy more like 1.2 petabytes because it’d need enough to support any growth spurts and enough for backup, as well as primary storage.
  • The storage cost for S3 is 15 cents per GB, which was astonishing to SmugMug’s management when the price was revealed to them for the first time.

While S3 obviously offers an enticing value proposition, there are storage needs that SmugMug continues to keep in-house due to special requirements.

And there are always risks associated with using a third-party, especially when it comes to storing valuable customer data.

S3 has experienced some outages which caused customer unhappiness but SmugMug is now apparently “seeing better uptime from Amazon than it could deliver in-house“.

I personally think the future is bright for services like S3 (and Amazon’s cloud computing service). Internet entrepreneurs would be wise to give serious consideration to the benefits hardware-as-a-service offerings can provide to their companies, especially as the quality and reliability of these services inevitably increases over time.

As the market for hardware-as-a-service matures and more specialised and flexible offerings are introduced, internet entrepreneurs may find that one of their biggest challenges just got a whole lot less worrying.