Discover more from The Ontologist
By addressing these risks proactively, the service can better position itself to capitalize on its innovative approach while safeguarding against potential pitfalls.
There is a classic conundrum that any CEO, CTO or IT director has to face: when looking to produce a new product, how do they assess the risk of bringing that product to market? Executives often talk about how deep the moat is that protects their business from competitors, but by this, they mean that a competitor would need to take inordinate risks (including capital expenditures) to develop an analogous product. To assess those risks, they would need to understand the comparative maturity of the components that go into that product and walk a fine line between working with technology that is too immature to be stable and easily reproducible.
Simon Wardley addressed this particular conundrum in 2016 as he was determining what strategies his companies should take to remain a market leader. Being of a decidedly analytic mindset, he broke down the nature of strategy and realized that such strategies required not only a purpose but also an awareness of both the internal factors that a company faced and where that company existed relative to larger economic factors. Like a game of chess or go, business strategy depended upon being aware of the placement of the pieces at any given time and their influence on achieving specific objectives that made that board more favourable for the business owner over time.
Taking a sabbatical, Wardley refined these ideas until he had developed a maturity diagram “language” for describing risk in the form of a set of diagrams. These diagrams soon became known as Wardley Maps, with Wardley then penning a book (https://learnwardleymapping.com/book/) to explain how best to use them. These maps are beginning to appear in board rooms, investors’ forums and even project management venues as ways to perform risk analysis and establish strategy through storytelling.
The Anatomy of Wardley Maps
The typical Wardley map is a two-dimensional grid. The x-axis of that grid indicates the maturity of a given component, from novel (unstable, few instances, very risky) all the way to commodity (universally present, easily available, generally inexpensive, and with low risk), tied to a scale from 0 to 1. A new attention optimization technique based upon a white paper written two months ago would be considered highly risky, and might be given a maturity of 0.1, while the use of the electrical power grid would be considered as a commodity or utility, and be in the 0.8 to 0.9 range.
The y axis of a Yardley Map is indication of its visibility to the anchor points (those likely to utilize the technology in question). A visibility of 1 means that the technological component is very visible, such as users having to invoke the component by manually running a python process in a command line, while a visibility of 0 means that the component is deep in the background (the aforementioned electrical grid, which you don’t think about until it fails). Visible processes are fault tolerant - they are almost guaranteed to fail, and fail frequently, but when low visibility processes fail (you lose power, for instance) it can represent a complete shut down of everything that is dependent upon that.
An example of such a Wardley Map might be something like the following:
In this example, I used as an example creating a generative AI dating business that would let you create your own avatar and interact with other avatars to do dating online. The anchor in this case is the business, showing that it has dependency upon all of the other technologies (you could think of this as the domain ontology of this particular map).
There are two very new technologies - the Generative AI Matching Engine, which helps do the comparisons between individuals and recommends potential matches, ,and the Avatars component, which handles the actual rendering of the characters avatars. These are both very immature technologies, though far enough along to be beyond the Genesis stage (hey, I have an idea!) to the Custom Built or Proof of Concept stage (eh, that works, but why does he have six fingers?). They are also both highly visible to the customers.
Avatars have a dependency upon an older Chatbot Interface which is somewhat more stable in that it has a reliable industry wide API, just as the matching engine is dependent upon contemporary data analytics tools that are themselves more battle tests. the matching engine also makes use of a combination of email and social media feedback tools which are generally very well known and frequently out of the hands of the product owner.
If either of these failed (say the social media company is bought up by a deranged billionaire who then runs the company into the ground to soothe his ego) the impact could be widespread, but the chance of that happening is small (how likely would that really be, eh?). Similarly the company itself has dependencies upon Marketing and Customer support, which are similarly commoditized (very low risk, high impact. The vertical aspect, visibility, is less important here, though most Wardley Maps do run from upper left to lower right as the older technologies are also generally less visible to the average customer.
Wardley Maps and OWM
Note that any of the components involved have their own Wardley maps, with older components being compressed farther to the right. The business, however, does not necessarily have dependencies upon every component in this space, meaning that should those components fail, it is unlikely to impact the business much at all.
This particular map was created online at https://onlinewardleymaps.com (and the map itself can be referenced here). While there is a drag and drop interface, the other reason why such maps are useful is that they make use of a particular domain specific language (DSL) called OWM (online wardley maps). This can be used to fine tune the map, add annotations, and provide styling. The above map can be described vie the following OWM:
title Generative AI Dating Service
style wardley
# Anchors
anchor Dating Service Business [0.91, 0.48]
# Components
component Avatars [0.75, 0.31] label [-70, -5]
component Generative AI Matching Engine [0.70, 0.15] label [15, -4]
component User Profile Management [0.60, 0.53] label [5, -10]
component Chatbot Interface [0.67, 0.45] label [-76, -22]
component Data Analytics and User Behavior Modeling [0.61, 0.34]
component Secure Data Storage [0.33, 0.81]
component User Feedback System [0.40, 0.70]
component Mobile Application [0.50, 0.75]
component Marketing and User Acquisition [0.40, 0.60]
component Customer Support [0.35, 0.66]
# Links
Dating Service Business->Avatars
Dating Service Business->Generative AI Matching Engine
Dating Service Business->User Profile Management
Dating Service Business->Chatbot Interface
Dating Service Business->Data Analytics and User Behavior Modeling
Dating Service Business->Secure Data Storage
Dating Service Business->User Feedback System
Dating Service Business->Mobile Application
Dating Service Business->Marketing and User Acquisition
Dating Service Business->Customer Support
Generative AI Matching Engine->User Profile Management
Generative AI Matching Engine->Data Analytics and User Behavior Modeling
Generative AI Matching Engine->User Feedback System
Generative AI Matching Engine->Avatars
Avatars->Chatbot Interface
Chatbot Interface->User Profile Management
Mobile Application->User Profile Management
# Annotations
annotation 1 [0.70,0.12] Continuous improvement of the AI Matching Engine is vital for successful matchmaking
annotation 2 [0.30, 0.79] User privacy and data security are critical
annotation 3 [0.32, 0.62] Effective marketing is necessary to grow the user base
annotations [0.4,0.15]
The Wardley Map is itself a graph, connected through the graph Links given above. The coordinates given above are of the form [visibility,maturity] with both going from 0 to 1. This can readily also be expressed as a semantic graph. For example, the Avatars component may very well look like the following as Turtle:
Component:Avatars a Component: ;
Component:dependency Component:ChatbotInterface ;
Component:maturity 0.31 ;
Component:visibility 0.75 ;
Component:description "Real time animated figures that reflect the desired projected appearance state of the user"^^xsd:string.
Component:label [
Label:text "Avatars"^^xsd:string;
Label:xOffset -70 ;
Label:yOffset -5
]
.
This opens up the possibility of analytics on Wardley maps, such as the ability to determine the full dependency graph of a component by doing a transitive closure on the Component:dependency
predicate or identifying potential complementary (or competitive maps based upon the combination of overlap and maturity.
Wardley Maps and ChatGPT
Another very useful tool is the Wardley Map Analyst which is a custom GPT on ChatGPT located here, and that may be available as part of the ChatGPT store when that debuts in the near future. The WM Analyst makes it possible to interpret Wardley Maps and ask questions of them.
For instance, I uploaded the Wardley Map (as a PNG) by attaching it to the prompt, then included the prompt message “Analyze this Wardley Map”. The analyst examined the map then gave a breakdown of each component by analyzing the image itself. When it was done, it also made a number of strategic observations and recommendations, as follows:
Strategic Insights:
Security and Privacy: The positioning of Secure Data Storage and the related annotation suggest that data security is a high priority, critical for maintaining user trust.
Market Positioning: The presence of effective marketing and user acquisition components suggests that growing the user base is a strategic goal, as indicated by the annotation.
Innovation Focus: The service is investing in innovation, particularly in the AI Matching Engine and Avatars, to create a unique value proposition.
User Experience: The User Feedback System’s evolution level indicates a focus on user experience, which is crucial for product improvement and customer satisfaction.
Recommendations:
Prioritize AI Development: Focus on the continuous development of the AI Matching Engine as it's the key differentiator for the service.
Invest in Data Analytics: Enhance data analytics and user behavior modeling to refine the matching process and improve success rates.
Enhance Security Measures: Given the critical nature of privacy in a dating service, ensure that secure data storage solutions are robust and up-to-date.
Monitor the Evolution: Keep an eye on the evolution of the components, especially those in the genesis and custom-built stages, to maintain a competitive edge.
Expand on Standard Services: For components like User Profile Management, consider innovations that could add value beyond the standard offerings in the market.
The map is a dynamic tool and should be updated regularly as the service evolves and adapts to market changes, user feedback, and technological advancements.
The Wardley maps can also provide a proxy into the company itself that can be used for marketing and funding. For instance, consider the following exchange on the Analyst that can cut down on the tedium of building a good elevator pitch for the company:
You
You are a founder seeking investment into this company. Can you put together an elevator pitch for why you think this is likely to be a good investment?
Wardley Map Analyst
Certainly! Here's an elevator pitch for the Generative AI Dating Service:
Introduction: "Good [morning/afternoon/evening], I'm the founder of a groundbreaking Generative AI Dating Service that’s poised to revolutionize the way people connect and find love. We leverage advanced AI to offer a deeply personalized matchmaking experience."
Problem: "In the crowded market of dating services, users are often overwhelmed with matches that don't resonate with them. The lack of deep personalization and meaningful connections leads to user fatigue and churn."
Solution: "Our proprietary AI Matching Engine uses generative AI to understand the nuance of user preferences and behavior, creating avatars that facilitate matches with unprecedented depth and compatibility. This isn't just another swiping game; it’s about creating real connections."
Product: "We’re at the forefront of the custom-built phase with our AI, actively learning and evolving with each interaction. Unlike other services, our platform offers a unique chatbot interface that allows users to interact naturally, as well as a continuous feedback loop that refines user profiles over time. Our mobile application blends sophisticated technology with ease of use, ensuring accessibility for all users."
Market Potential: "The global online dating market is expected to grow significantly, and with our advanced AI capabilities, we are well-positioned to capture and expand this market. Our commitment to user privacy and security with state-of-the-art data storage solutions sets us apart in an era where trust is paramount."
Business Model: "We operate on a freemium model with multiple revenue streams, including premium features, in-app purchases, and partnerships. Our data analytics offer valuable insights, which can open doors to tailored user experiences and strategic advertising opportunities."
Growth Strategy: "Our marketing and user acquisition strategies are data-driven, ensuring we target the right audience efficiently. With a scalable model and a clear path to profitability, we aim to grow our user base exponentially."
Ask: "We're seeking an investment to scale our AI development, enhance our data analytics capabilities, and accelerate our user acquisition efforts. Join us on this journey to reshape the future of dating."
Closing: "Imagine a world where finding a match is as profound and personalized as the connection itself. With your support, we can make this a reality. Let's create a new paradigm for love in the digital age."
This pitch is designed to highlight the innovation, market potential, and the strategic vision of the company, aiming to capture the interest of potential investors within a brief encounter.
Similar role-playing can be done from a project management and DevOps perspective to get an idea about potential risk points from a more technical standpoint and to use such Wardley Maps as roadmaps for identifying and developing key technology infrastructure and areas requiring integration. Wardley maps could also tease out Agents and Actions as those move online throughout the next couple of years. Finally, they can be used to create at least rough estimates of the risks and consequent costs of any particular venture, such as the following:
Given this Wardley Map, provide a risk assessment analysis of this particular approach.
Wardley Map Analyst
A risk assessment analysis for this Generative AI Dating Service, based on the provided Wardley Map, will consider various factors such as dependency risks, market risks, technological risks, operational risks, and security risks. Here's a detailed breakdown:
Dependency Risks:
AI Matching Engine:
Being in the genesis stage, the AI Matching Engine is both a unique value proposition and a significant risk due to its innovative nature.
Dependency on cutting-edge AI means any setbacks in AI research or development could impede the service's core functionality.
Avatars:
High innovation also means high unpredictability. If avatars do not resonate with users or fail to enhance the matchmaking process, the user experience could suffer.
Data Analytics and User Behavior Modeling:
These components, while more evolved, rely on large volumes of quality data. Poor data quality or insufficient data could lead to inaccurate analytics, affecting the matching process.
Market Risks:
User Acquisition and Marketing:
The success of the service is heavily dependent on effective marketing strategies to grow the user base.
There is a risk that acquisition costs may be higher than anticipated, particularly if the market is saturated or if user retention rates are lower than expected.
Technological Risks:
Mobile Application and Chatbot Interface:
These are closer to the commodity space and might not provide a competitive advantage.
The risk of technological obsolescence is present if competitors introduce more advanced interfaces or applications.
Secure Data Storage:
As a commodity, the assumption is that this is a solved problem. However, any breach or failure in data security would have catastrophic effects on user trust and the company's reputation.
Operational Risks:
Customer Support:
As the service scales, the need for robust customer support will increase. If not scaled properly, it could lead to user dissatisfaction.
Over-reliance on standardized customer support processes without considering the personalized nature of the service may lead to inadequate support.
Security Risks:
Data Security and Privacy:
Handling sensitive user data, particularly in a dating service, is fraught with privacy concerns and regulatory compliance risks.
There's a risk of data breaches, which could lead to significant legal and reputational damage.
Recommendations for Mitigation:
R&D Investment: Allocate sufficient resources to R&D to maintain the pace of innovation while also preparing contingency plans for any developmental roadblocks.
Diversified User Acquisition: Employ a variety of marketing channels and strategies to mitigate the risk of user acquisition cost overruns.
Continuous Technological Upgrades: Regularly update the mobile application and chatbot interface to stay ahead of technological advancements and competitor offerings.
Robust Security Protocols: Invest in top-tier security measures and regular audits to protect user data, ensuring compliance with all relevant data protection regulations.
Scalable Customer Support: Develop a scalable customer support system that can grow with the user base while maintaining a high quality of service.
By addressing these risks proactively, the service can better position itself to capitalize on its innovative approach while safeguarding against potential pitfalls.
Overall, Wardley Maps should be seen as a significant tool for anyone involved in strategic decision-making, regardless of the scope of that responsibility. They can help to organize an awareness of the proposed organization or business proposal, can provide an invaluable tool to business owners, investors, and engineers, and can serve as a foundation for mitigating churn in the deployment of this technology, even to the extent of becoming an integral part of the DevOps automation landscape.
Kurt Cagle is the Editor of The Ontologist, and is a practicing ontologist, solutions architect, and AI afficianado. He lives in Bellevue, Washington with his wife and cats.
For more posts and resources on semantic modeling and ontology development, subscribe to The Ontologist:
Sign up for free Ontology Office Hours at Calendly.
Subscribe to The Ontologist
Principles of Metadata Architecture, Graph Theory and AI