Where is your company on the Artificial Intelligence Value Chain?

AI is moving at “break-neck” speed, leaving many business and government agencies wrestling with “make-buy” decisions that are unaffordable or impossible to deliver.  In this blog post, we present a model that frames the challenges and provides insights to make better decisions.  As with any business, there are business drivers that require changes to the business in order to grow.  Once those drivers are understood, the business can assess market trends and then develop a list of initiatives or projects to move the business forward.  This evolution holds true for everyone in the business ecosystem.  In order to simplify the discussion, we can look at it from a technical perspective, or the “Anything as a Service” (XaaS) model.  This approach offers a way to access and use technology resources as a subscription-based service rather than owning and maintaining them, such as Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). 

For representation purposes, let’s use the OSI model and build from the bottom up starting with Infrastructure as a Service (IaaS) where basic facility resources such as “Ping”, “Power”, and “Pipe” are found.  IaaS is a cloud computing model that provides on-demand access to computing resources such as servers, storage, networking, and virtualization.  The market for providing IaaS can be expensive and if it isn’t part of your core business offerings, can be too expensive to develop and implement.  What makes outsourcing IaaS attractive is that acquiring computing resources to run applications or store data the traditional way requires time and capital, where buying or leasing may offer more affordable options.  There are many companies who have already invested in IaaS as part of their core offerings, such as Telecom companies like Lumen Technologies and AT&T, as well as hyperscalers such as Google, Meta, AWS, Microsoft and others who have invested billions of dollars in new fiber networks and AI data centers.

Once the infrastructure is in place, the next investment is the Platform as a Service (PaaS) where the cloud application Infrastructure sits on top of the IaaS to provide compute resources.  PaaS is a cloud computing model that provides a complete development and deployment environment for applications. PaaS allows developers to build, test, and manage apps without having to worry about infrastructure, software updates, or hardware maintenance.  This is a key component of the AI data centers and requires a significant investment to build and involves creating a robust, scalable, and accessible environment for developers to build, deploy, and manage applications, encompassing infrastructure, middleware, and development tools. It requires careful planning of the architecture, selection of technologies, and design for manageability and automation.  Once built, the PaaS provides the servers, processors, and compute environment for the next level of services.

Software as a Service rides on top of the PaaS and allows users to connect to and use cloud-based apps over the Internet. Common examples are email, calendaring, and office tools (such as Microsoft Office 365).  Implementing SaaS involves a multifaceted process, from initial planning and vendor selection to configuration, training, and ongoing optimization, ensuring a smooth and effective transition for users.  Companies must define the objectives they intend to achieve with SaaS, analyze existing systems, develop a sound business case, determine the products needed, implement and deploy the solution, migrate existing data, test the implementation, train users, and maintain the environment if the implementation is in house.  This is an expensive undertaking and has led to a development of SaaS providers such as SalesForce, Microsoft, Google, Adobe, Zoom, and many others.  With all of this infrastructure, hardware, and software there is a hierarchical value stream that leads somewhere.

The IaaS, PaaS, and SaaS are meant to serve the business needs of customers.  These business needs are the internal business processes that companies use to perform the primary functions such as healthcare and banking systems for example.  Let’s refer to these as Business Process as a Service (BPaaS).  These businesses have unique functions but can benefit from similar services.  Let’s take transaction processing as an example.  In healthcare there are many transactions that take place, such as data analytics on patient data, protocol analysis, and Electronic Healthcare Records update and management.  There are also billing and payment processes that need to be managed.  In the banking industry, there are also many transactions (deposits and payments), financial analytics, electronic funds transfers, and electronic data interchanges.  Imagine having to determine the control points to run an effective fraud, waste, and abuse program in these two examples.  So where is the right place to engage your current business?  Let’s take a look at a model that may help you find your “sweet spot”.

As a business, it is important to understand the cost verses value in your market space.  As the graphic represents, it is much more costly to engage the customer at the IaaS level, as the investment typically makes it too expensive to enter, and the customer doesn’t place much intrinsic value on IaaS, looking at it more as a commodity component of their business.  Moving up the stack into the PaaS, it becomes a little more affordable, assuming you have an infrastructure to build it on.  Even the assets at this level can be cost prohibitive for many companies to focus on, as the total cost of ownership of the assets is not only in the hardware and software, but the administration of the assets as well (yes, even in a cloud environment).  Most customers again look at this as a commodity and are not willing to pay for all of the PaaS resources they actually need, at least not on a full-time basis.  The next level is the SaaS, and if your key product is a piece of software that the customer needs, you are now entering the level where a smaller company can gain an advantage and deliver real value to their customers.  Customers know they need the software to execute their business processes and need to be moving from a commodity to a business solution.  Which brings us to BPaaS, where customers like the healthcare and banking industries need not only the hardware and software to perform their business tasks, but now really need industry professionals who understand the business, legislation, technical solutions, process maps, interfaces, and such.  They need subject mater experts, and this is where they are more willing to spend their business dollars, and typically try to develop their own company talent from these expert engagements.  As a business you have an opportunity to figure out where you want to jump in.  Let’s take the model above and flip it on it’s side for the next part of the discussion.

As the model shows, your business investment is pretty high to acquire a facility and cloud environment, to which the customer typically places little unique value on.  As you move to the right, down the investment curve and up the customer value curve, you transition from a purely technology company to a business solutions company.  A lot of small companies start engaging the customer value curve at the SaaS level where they are beginning to incorporate their knowledge of the customer business processes, while other smaller companies have a strong play beginning at the BPaaS level.

As the model demonstrates, you want to be able to determine at which point along the customer value chain your company can engage.  Simply having the IaaS and PaaS does not solve the customers problem.  Understanding where the customer is regarding SaaS and BPaaS is also very important.

Taking the model one step further, suppose you are a business supporting a customer who delivers PaaS or SaaS.  To engage these companies, the investment and value curves tend to be a little steeper.  This is where partnerships, joint ventures, and financing can be the difference between success and failure.  If you customer is in need of a PaaS solution to run their entire operation on, and you have a partner ship with a PaaS provider, you will see the high value the customer will place on those resources and where the profitability intersections occur.

Affordability has always been a differentiator in business. Companies like AWS are able to cover most of these elements but can’t be the savior to all businesses nor should they want to.  There is a theoretical point of diminishing returns and thus the demand for subject matter experts and companies who can more efficiently leverage companies like Amazon and Lumen Technologies capabilities.  The old model was to build it from scratch, to have all of the layers within one company, but as technology evolved and business financial models changed, there was a new model evolving.

Companies began to realize they could no longer afford to do it all and began buying infrastructure and platform support, outsourcing nearly every aspect of their company except the intellectual business capital they had.  With the continued development of Artificial Intelligence (AI), we can expect a continuing evolution and movement away from an army of engineers and architects to a workforce who visualizes a need and leverages outsourced services to create, implement, and manage their new business processes. 

Will the final step be for hyperscalers to consume it all, to eliminate businesses by the thousands?  We will see.  But I predict that, much like the mobile phones and internet, new markets and capabilities will drive other needs not yet realized.