I participated in a watch party of the Oracle AI World event with colleagues Brent Leary, Jon Reed and Paul Greenberg. We were listening in as new Oracle Co-CEO Mike Sicilia spoke with top executives from several different Oracle customers.
Those customer interviews were progressive with each adding new insights into how Oracle technologies, especially AI, could be of value to them and other firms.
Listening to the speakers, it was clear that each is adopting AI in one of three familiar patterns.
The first customer described how they were using the new tech to further automate a pre-existing process (or parts therein). The process won’t be materially different. It will just work faster and be less error-prone. This is a basic efficiency or productivity play and is one the most common use cases that vendors and customers alike mention.
Interestingly, a number of the customers/prospects at the show would definitely love to get a lot of this efficiency and productivity value. They still have too many spreadsheets in use, manual or broken integrations, inefficiencies left over from poorly integrated acquisitions, tons of technical debt and other challenges that are keeping their firm from operating at best-in-class levels. In this value proposition, the technology only has to fix a small portion of a broken process to deliver value.
The second customer speaker described a new level of capabilities that made a series of work steps run better and be more automated. Specifically, the speaker described how the new tech could automate a string of events and also deliver outsized value. These changes are often process or workflow wide. They cover many steps of a process and enhance the overall effectiveness and efficiency of an entire stream of work. As impressive as the improvements can be, they don’t necessarily reflect a radical reimagination of the process.
The third speaker described how they are using/may use Oracle’s vectoring technology to quickly and precisely identify DNA fragments and other factors to better prescribe the correct antibiotics to patients. Vector technology is mostly used as a way of codifying the contents of databases, documents, etc. by assigning values to the things it finds in the data. For example, it’s great for finding/identifying objects in pictures as the vector co-ordinates of two different images of cats in two different pictures have a number of similar features in their overall shapes.
That third example is intriguing as it is an example of how a technology designed for one set of use cases can also be used in new novel ways that create all kinds of new capabilities and value for its users. This is what transformational change is all about.
In fact, as this customer described this vector use case, I immediately saw how other organizations could use this. For example, organizations that are trying to identify gene or genome variants that give plants immunity to specific pathogens or pests would love this.
These three ways of utilizing new technology remind us that Fubini’s Law remains valid and helps us understand the uptake of new software technologies.
It’s time to dust that off.
The first part of Fubini’s Law (i.e., people use tech to what they do now) is exactly where a number of software shoppers are today. They are experimenting/piloting/testing new technologies to see how they can either replace or supplement parts of existing workflows and processes. They want their people and processes to be more efficient, more cost effective and less error-prone.
These approaches allow buyers to gain familiarity with new technology in a safe bounded use case. This makes a lot of sense as several customers reported end-user unease with new AI technologies. Specifically, we heard of users concerned about AI hallucinations, data privacy concerns, etc. This is actually a key step in the deployment of new technology as tech decisions are not just about code, platforms, UIs, functions and features. Technology deployment decisions are multifaceted and involve financial, human, change, policy, scheduling, audit, staffing, legal and other concern areas. While some executives might want to ramrod new technology into a firm, the world is full of stories of failed IT projects and many of those are due to an inattention to the non-technical aspects of tech projects.
The second part of Fubini’s Law (i.e., users gradually begin to use the new tech in all-new ways) is also correct. The company using vector mapping is a perfect example of this phenomenon.
And an outgrowth of the second part leads to the third part (e.g., the new tech eventually leads to changes in life-styles and work-styles). We’ve all seen and experienced how smart phones changed numerous aspects of our lives including how we buy most anything, are always mobile accessible, conduct cash-less commerce globally, etc.
AI is changing (or will soon change) work, work-styles, needed worker skillsets, and more. AI will run large chunks of processes in the background and only surface anomalies and items requiring review to users. AI Agents will move data along to different work steps, processes, users and other agents. This change is well underway.
We’re still in the early stages of AI adoption, as per Fubini’s Law. It will take years before more firms get to the fourth level and some of the all-new capabilities that AI will enable (in the second and third stages) will be eye-opening. It’s going to be fascinating to watch these evolutions emerge.
My unscientific conversations with Oracle customers showed definite near-term interest in AI solutions solving part one of Fubini’s Law and awareness of future advanced technologies to use in the long-term. AI uptake may start off modestly and then hit a crescendo quickly.
Who’s buying this?
In short, there seemed to be buyers for the new tech everywhere. Oracle has already got a massive ecosystem of partner firms building agents to work with Oracle’s applications and tools. Oracle also has an impressively deep and broad set of applications (i.e., Fusion Apps) that are already using AI capabilities. And don’t forget they also possess the platform, the cloud computing infrastructure, alliances/partnerships with key LLM providers, data centers and much more.
Oracle may be ready for the business world to adopt AI solutions, en masse, but are businesses equally ready to buy? Readers should note that not all businesses adopt technology at the same pace/tempo. Different software buyers have differing demographics, competitive needs, financial constraints, etc. Software buyers are not, and have never been, a homogeneous lot.
The definitive book re: tech buyers (Crossing the Chasm) was actually penned a couple of decades ago by Geoffrey Moore. His book introduced Silicon Valley and others to five kinds of tech buyers and the technology adoption life cycle (TALC). Here’s a quick refresher of those items:
The companies that will be deploying AI in major, bold and enterprise-wide ways in the next couple of years will likely be those on the left side of Moore’s TALC. These firms will look at AI as a means of achieving a measure of competitive advantage and will likely invest in the skills, technology, etc. to make big ideas come to life. Additionally, some of these buyers are going to really stress the capabilities of potential partners (e.g., systems integrators) to deliver powerful extensions, reimagined vertical solutions, slick workflows, deeply skilled personnel and more.
Companies on the right side of this bell curve are more likely to ease into AI and take their time doing so. They will embrace more pre-packaged solutions whether they come from Oracle or one its partners.
Oracle’s AI World show had software shoppers from around the world and in most every vertical slice imaginable. Not all of these are ready to embrace AI in its totality now but make no mistake, they are looking at these technical improvements will an eye to buy.
That’s a great sign for Oracle as customers/prospects are soaking in the new Oracle offerings, exploring how these tools could help their firm and what it will take to bring them to life. That’s actually a great buying sign for Oracle and these are buyers that want to bring change to their firm. They are also collecting tips/pointers and are making connections with Oracle, other Oracle customers and with Oracle partners. Bottom line: Oracle should expect a significant amount of new business. The timing of when each buyer will utilize the new tools may be the key variable.
The other (big) complicating factors
It was also clear that customers know that they must complete a lot of prep work prior to launching their new AI initiatives. Specifically, customers:
- Must have great, clean, current and accessible data.
- Are realizing that piles of best of breed applications are harder to integrate into a great AI solution than having a full suite of applications from one vendor (a positive for Oracle).
- Need data with consistent data meanings, units of measure, etc. will make for easier and better implementations.
- May need to develop all-new approaches to using other data types (e.g., video for the development of training), images, sounds, other languages and new combinations of all of these.
- May have material IP (intellectual property) challenges. For some applications (e.g., sales productivity tools, collateral development tools, estimating tools, etc.), access to complete, current materials is essential to having AI tools deliver highly relevant and accurate work products.
- Could get better AI results if they license external data, external LLMs (large language models), etc. Oracle is already helping in this regard as it currently supports a number of the major LLM providers.
Change management concerns factored large in customer conversations. While people are generally a bit change averse, the best run project teams must definitely realize the materiality, import and significance of change management in these newer AI projects. Profound changes in technology, workflows, the nature of work, systems and metrics will likely be part of all AI projects, even some of the more modest ones. If you want your AI effort to be a success, get on top of these matters!
My take
Sometimes the real headlines of a software conference aren’t found in the numerous press releases issued in conjunction with the event’s conduct. The real stories are often in the interactions with the customers and prospects in attendance. And while the noise, chaos, scheduling, etc. of these events makes it hard to get great, statistically significant information, it’s the sentiment and observations that really paint a picture.
What I noticed is that:
- existing customers will likely stay with Oracle and possibly extend their footprint with Oracle
- existing customers and prospects want to implement AI, use a number of the Fusion apps, and, take advantage of Oracle’s numerous platform capabilities
- customers and prospects definitely know they have material pre-work ahead of their AI ambitions
- change management, skills shortages, risk concerns, etc. are all issues that are front of mind for these executives
Finally, in a book I wrote a few years back (i.e., Digital With Impact – on Amazon), I suggested an alternative to Fubini’s Law. I called it Sommer’s Law and it posited:
- “People now use new technology to do what they do today – but differently.
- Their use of new technology makes businesses and others change how they interact with and understand people, organizations and systems.
- The new technology may be combined with other technologies to create transformative solutions that affect entire industries.
- The transformative technologies trigger new waves of innovation that eventually change businesses, work and economies
…and these set the stage for the next wave of transformation to come.”
I believe I got that right.
We’re in the early innings of big changes and it should be interesting to see how things change…
link