Generative AI for Strategic Plan Development
Jesse Ponnock · 2025
Read Generative AI for Strategic Plan Development by Jesse Ponnock (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-07405.txt.
Introduction
Jesse Ponnock writes in 2025 about using generative AI to help draft strategic plans for large public bodies. Such plans are costly to write, yet law and rules demand them. Ponnock sets out a modular model that assigns each part of plan drafting to a step where AI can help.
One module finds themes in piles of reports. Two topic-finding methods were trained on many reports from a federal watchdog office. Both found themes matching every vision element of a real plan. One method did best, with over half its matched topics scoring medium or strong. Next steps test the rest of the model in real use.
Catalog record
| CREATOR | Jesse Ponnock |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-07405
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Generative AI for Strategic Plan Development Jesse Ponnock Abstract Abstract Given recent breakthroughs in Generative Artificial Intelligence (GAI) and Large Language Models (LLMs), more and more professional services are being augmented through Artificial Intelligence (AI), which once seemed impossible to automate. This paper presents a modular model for leveraging GAI in developing strategic plans for large scale government organizations and evaluates leading machine learning techniques in their application towards one of the identified modules. Specifically, the performance of BERTopic and Non-negative Matrix Factorization (NMF) are evaluated in their ability to use topic modeling to generate themes representative of Vision Elements within a strategic plan. To accomplish this, BERTopic and NMF models are trained using a large volume of reports from the Government Accountability Office (GAO). The generated topics from each model are then scored for similarity against the Vision Elements of a published strategic plan and the results are compared. Our results show that these techniques are capable of generating themes similar to 100% of the elements being evaluated against. Further, we conclude that BERTopic performs best in this application with more than half of its correlated topics achieving a “medium” or “strong” correlation. A capability of GAI-enabled strategic plan development impacts a multi-billion dollar industry and assists the federal government in overcoming regulatory requirements which are crucial to the public good. Further work will focus on the operationalization of the concept proven in this study as well as viability of the remaining modules in the proposed model for GAI-generated strategic plans. 1 Introduction 1.1 Background The purpose of this paper is to analyze the ability of GAI to develop strategic plans for large scale government organizations. The proposed framework decomposes the various modular tasks involved in developing a strategic plan and evaluates the suitability of GAI to address these needs. Further, industry-leading GAI techniques are used to model outputs for a domain-specific use case and their performances are evaluated to identify a leading solution. Strategic planning is defined as “a process in which an organization’s leaders define their vision for the future and identify their organizations goals and objectives” [1]. When executed effectively, these goals and objectives will dictate the projects and activities that the organization undertakes over the lifecycle of the plan. Because of this relationship between visioning and business planning, a strategic plan will deeply influence the direction of an organization and can be solely responsible for the failure or success of achieving its mission. In government, a strategic plan is a necessary tool in demonstrating transparency in direction, accountability in elected officials, and good stewardship of public funds. For this reason, in 1993 congress passed the Government Performance and Results Act (GPRA) requiring all major agencies to develop and maintain a strategic plan. In private industry, strategic planning can be crucial in ensuring growth and profitability. Notably, 71% of fast-growing companies have strategic plans, business plans, or similar long-range planning tools in place [2] and those which do not consistently report weaker financial results than their peers [3]. Strategic planning is a multi-billion dollar industry led primarily by large consulting firms due to the expertise required in developing a strategy. It is estimated that the global strategy consulting market will reach $111.4 billion by 2031 [4]. The method for developing a strategic plan can differ from case to case however a typical implementation will involve a small group of professionals working for several months, sometimes a year or more, at a project cost at times accruing to millions of dollars. The steps taken by these individuals include a series of discussions and analysis to extract key themes around an organization’s envisioned future and deriving a series of goals and objectives from these findings. It is the hypothesis of this research that with recent breakthroughs in fields such as natural language processing, sentiment analysis, topic modeling, and GAI, a system can be designed and developed to automate the generation of a strategic plan with minimal manual human involvement, significantly reducing the cost and time required to produce. To begin exploring the potential for a GAI-generated strategic plan, we must first understand the components required for such a product. The below diagram depicts the components of a strategic plan, a brief description of each, and an associated cognitive task behind the creation of each. Figure 1: Cognitive Model for Strategic Plan Development The model above proposes five key components in developing a strategic plan: Vision Elements, Goals, Measurements, Strategies, and Activities. It indicates alternative groupings for these components to describe their relationship to complementary artifacts which are common in the field of strategy consulting. For example, Vision Elements are sometimes compiled into a standalone document to describe an organization’s Vision or future state. Similarly, a series of strategies and activities will often be compiled in an effort to develop a Roadmap. For the purposes of this study, we will view all of these components in aggregate as the strategic plan. In the ovals that correspond to these 5 components to the right of the diagram, we describe the proposed primary cognitive task associated with developing each of these components. For example, development of the Vision Elements, or future state, for an organization typically involves stakeholder interviews with one or more senior executives and a review of large collections of documentation from and about the organization in an effort to identify future needs and potential modes of operation. This task aims to collect large swaths of information which through analysis is then summarized into key themes which will represent the targeted future state of the organization. For example, if several sources describe impacts of aging technologies, an analyst may identify IT modernization as a theme for a Vision Element as part of the broader Vision. It should be noted that the contents of a strategic plan can differ by implementation and is heavily driven by the design choices of its implementor as well as the specific purpose the plan is looking to serve. It is generally agreed upon however that at its core a strategic plan will include a representation of a future state (referred to above as the Vision), goals and corresponding measures, and some depiction of actions or recommendations that will actualize the strategic direction (referred to above as Strategies and Activities). The described model above focuses on these core elements for the purposes of this study but others may exist with differing terminology and scope within the areas described. The proposed primary cognitive tasks provide a structure to quantify the largely qualitative problem of strategic plan development. It follows that if we can break down the various tasks in this process into their equivalent human cognitive requirements, and then solve for each of those tasks using machine learning, we will have solved for the holistic larger process. Viewing the results of this exercise tells us several things about this exploration. First, it will not be a singular machine learning model that will solve this problem holistically. Instead, several different models will be required, addressing different tasks within the larger problem, and some form of model orchestration will be necessary to manage the models working in sequence. This is meaningful but unsurprising, given this is often the