AWS re:Invent 2017 has exploded into one of the most significant tech events of the year, swelling to 43,000 attendees with sessions spread across six venues two miles along the Las Vegas Strip.
However, this cloud carnival isn't just physically overwhelming. The relentless barrage of product announcements, feature enhancements, and partnership agreements, along with a string of ancillary news from cloud ecosystem hangers-on exploiting the vast re:Invent publicity makes it impossible to digest all that happens, particularly if you're amidst the stew of noise and commotion that only amplifies the Las Vegas background din.
There's plenty to like in the 61 product announcements in 15 service categories that shows an AWS intent on leaving no customer want unmet, however, taken as a whole they only exacerbate a brewing problem of cloud complexity and the herculean task of incorporating AWS services into application designs and IT strategies.
It's hard to fault a company for putting out too many products and updating them too fast. However, it can lead to a problem familiar to consumer products companies, namely choice overload.
The presence of choice might be appealing as a theory, but in reality, people might find more and more choice to actually be debilitating.
The problem of choice is related to information overload in which,
Having more options to choose from within a category is likely to render the choice more difficult as the differences between attractive options get smaller and the amount of available information about them increases.
By any measure, AWS services qualify as complicated topics replete with technical details, documentation and relevant information.
Product strategy: seeing what sticks
AWS's penchant for adding services and features shows its lineage as the technology arm of Amazon, a consumer retailer with a plan to have something for everyone.
By one measure Amazon directly carries over 12 million products, a number that balloons almost 30-times when including items from marketplace sellers. While having many options when buying a TV or pair of shoes is nice, it can be confusing and counterproductive when trying to select a VM instance or database.
Curation is a foreign concept at both Amazon, and as the explosion of services demonstrates, AWS. The problem is that the customer experience for a retailer versus technology services provider is defined by different things.
Turning AWS into a flea market catering to every whim of long tail customers hinders the more important goal of guiding users and especially the large enterprises AWS wants to win over to its vision of a cloud-based future.
The problem of product overload is rampant in the technology industry where it's easy to make custom SKUs for every need. Product sprawl starts with the best of intentions, more closely matching every customer segment’s needs, but that strategy has unintended consequences. Buyers end up confused while product development and marketing resources end up diluted and fighting for resources.
As one cloud observer noted, Steve Jobs faced this when he returned to Apple. According to the account in Walter Isaacson's seminal biography, immediately set about slashing its portfolio by focusing on one product in each of four segments.
Like weeds, complexity tends to spread throughout an organization and strangle growth. And it keeps coming back unless prevented from doing so. Focused companies have learned this lesson and keep their profit gardens weed-free. They may set absolute targets for new products and services, removing one for every one they add. ... Nearly all periodically review their product portfolios to keep complexity in check.
Aside from contributing to choice overload, adding all those products comes with some significant costs. The Bain column notes that a medical device and specialty pharmaceuticals manufacturer adopted processes to simplify its SKUs and found the reduction in product complexity resulted in a 15% gain in margin. Indeed, he says that,
Companies with the least complex product and service offerings typically have a far better understanding of customers and what they want than companies with more complex offerings. Not surprisingly, the low-complexity companies typically grow almost three times as fast as high-complexity companies and are more profitable as well.
Of course, with revenue growing at 40% annually and 25% operating margins, AWS has room for inefficiency. The more significant problem is what product proliferation says about its understanding, or lack thereof, of its customers and their needs.
Don't expect AWS to change
AWS execs would likely counter that the cloud services business is still young and that the industry and business world writ large are still figuring out the sorts of services they need and how best to use them.
During a Q&A session with analysts, AWS CEO Andy Jassy responded to a question on product complexity by saying it has no intention of slowing its rate of innovation, but that AWS focuses on simplifying consumption.
Admittedly, we see signs of this in the new AWS Systems Manager, Kubernetes for ECS preview, Launch Templates for EC2, IoT Device Management and various machine learning services. No one wants to see AWS stop innovating, but to channel that creative energy into fewer comprehensive, easily used services that are part of a cohesive cloud application design and deployment strategy.
AWS’s focus on individual products, not overall solutions to enterprise and developer problems makes it more difficult in the long term to expand its share of the total enterprise infrastructure and application market. Indeed, of the four IT segments Gartner expects to be affected by a shift to the cloud, application and system infrastructure, i.e. PaaS and IaaS, have the lowest rate of market penetration. In response to Jassy’s answer about AWS complexity, one analyst Tweeted,
That's a shame. Consumption complexity for AWS is the most common cloud question I get; AWS is losing deals by losing customer centricity: a public answer would be useful. #reinvent
— Hyoun Park (@hyounpark) November 28, 2017
A KPMG report highlights the myriad ways that cloud complexity presents a significant barrier to enterprise IT adoption, claiming that the state of online services is far from the vision of the cloud as a computing utility. It notes that "except for most SaaS solutions, this grossly over-simplifies the current state of cloud computing, especially for large, global enterprises." Indeed, the report points out that,
For many organizations, implementing cloud computing and getting value from it are not necessarily synonymous. The diversity of legacy estates, technical debt, and growing number of available cloud services and delivery options (public, private, hybrid) creates a level of complexity in decision-making and implementation that is challenging at best and presents a barrier at worst.
Like Amazon, AWS is executing a strategy of building an online superstore for all an organization's cloud needs, even as IT executives and their advisors are seeing the wisdom of cloud diversification.
Heightened competition among vendors amidst a proliferation of mostly undifferentiated services is causing CIOs and enterprise IT buyers to adopt multiple cloud providers looking for the best product match and pricing for their needs.
Jassy acknowledged to a CIO forum that organizations won't limit themselves to a single cloud. He added that AWS understands the edge that IBM and Microsoft gain by exploiting their close working relationships with enterprise IT organizations, which is one reason he mentioned an initiative to improve enterprise consulting and support. Unfortunately, that message, if delivered at all during re:Invent, got lost among the new product carousel.
AWS is hardly the first company to be afflicted with creeping featuritis,. Apparently, the lesson that layering on products and features trades off flexibility for complexity must be learned the hard way.
Giving customers four different ways to do the same thing isn't doing them a favor in the long run.
Choosing among the 100+ instance types shouldn't be more complicated than buying a server, nor should figuring out which option to use from among five different database services require a project akin to evaluating traditional standalone databases.
While AWS is admirably pushing the state of cloud services, its technology- and the architecture-agnostic explosion of products without an opinionated stance on best practices and implementation misses a significant need for advice how best to incorporate them into an enterprise.