Level 0: No AI (Benchmark)

At Level 0, creating a press release is a purely manual and labour-intensive undertaking. The professional is the sole engine of creation, relying on their individual expertise and standard office software. Leaders often underestimate the sheer volume of high-friction, low-value tasks this single activity entails.

The Process

1. The process begins with a demanding strategic phase, where the user must identify and validate a newsworthy angle, define the release's core objective and research the precise media outlets and journalists to target. This is immediately followed by a painstaking information-gathering exercise: hunting down all the key facts, sourcing credible supporting data and coordinating with key individuals to secure and approve compelling quotes. This phase is fraught with hidden time-sinks and depends entirely on the individual's rigour and availability.
2. Next comes the creative heavy-lifting. The professional must craft a compelling headline, write a concise lead paragraph that summarises the entire story and then build out the body—all while adhering to the inverted pyramid style. They must seamlessly weave in the approved quotes, integrate supporting data and ensure all standard components like the boilerplate and media contact are present and correct. This is often a stop-start process, vulnerable to creative blocks and the tedious, manual work of formatting and wordsmithing.
3. Finally, the draught enters a high-risk, multi-stakeholder review cycle. Every fact, name and number must be manually verified against source materials to prevent embarrassing errors. The document is checked and re-checked for clarity, grammar and adherence to style guides. It is then circulated to a group of stakeholders for feedback and approval, often leading to multiple, time-consuming revision rounds. This entire phase represents a significant administrative burden and carries the constant risk of a critical error slipping through due to human fatigue.

This entire manual process conservatively consumes 8 hours of a professional's time. Assuming a loaded cost of $100 per hour and a frequency of two releases per month, the true cost becomes clear:

• 8 hours/release × 24 releases/year = 192 hours per year.
• The annual cost of this single, recurring task is $19,200.

The Challenges

• Significant human hours are demanded by the inherently labour-intensive, purely manual workflow.
• Creative blocks are common, with individuals struggling to consistently generate fresh angles or compelling copy.
• Variability in output quality and speed is unavoidable, as they depend entirely on individual skill and workload.
• Slow and potentially incomplete information gathering results from manual research and verification methods.
• Limited scalability makes handling multiple or urgent releases difficult due to finite human capacity.
• Inconsistencies in style, tone and branding can easily emerge without active process management.
• Increased risk of errors due to fatigue stems from tedious manual drafting and proofreading.

Level -1: Rogue Assistant (Risk)

At Level -1, the user does not understand how general-purpose AI works, how to interact with it safely, what they can or cannot do or how to be critical of its output. As a result, when tasked with creating a confidential press release, the user makes a series of predictable, high-risk mistakes:

1. The user pastes sensitive, proprietary information—the confidential terms of a partnership or unannounced financial results—directly into a public AI tool to "get it started." This is an immediate, unmanaged leak of company IP.
2. Unable to discern AI-generated "hallucinations" from reality, the user accepts plausible-sounding but factually incorrect statistics, quotes or strategic claims generated by the AI, embedding falsehoods into the core of the document.
3. The user accepts the AI's confident but misaligned framing of the announcement, allowing a flawed strategic message to be sent up the chain of command.

These individual failures create a cascade of tangible business risks, with the severity determined by how far the flawed document travels before being caught:

• If caught internally, the outcome is wasted time, project delays and an erosion of trust.
• If it reaches a partner or client, the consequence is relationship damage.
• If it becomes public, the result is a public catastrophe that can create significant legal and financial liability.

Level 1: Thinking Partner (2x Productivity)

At Level 1, creating a press release transforms from a purely manual Level 0 effort into a user-directed conversation with a generalist AI.

The human professional remains the author and strategist, but now leverages AI as an on-demand tool to accelerate and augment specific parts of the traditional workflow.

This conversation can take two forms, depending on the user's preference:

1. "Authoring" where the user directs the AI to help with discrete components (e.g., headlines, paragraphs or idea generation for a section), sequentially building the document by integrating AI's reactive output into their ongoing manual draught.
2. "Editing" where the user directs the AI to generate a comprehensive initial draught based on general instructions and then iteratively refines this AI-generated foundation through further targeted prompts.

In either case, the AI partner operates in generalist mode:

• It leverages broad public knowledge but has no inherent company-specific context or confidential strategic insight.
• It relies entirely on the user's explicit prompts for direction, input and the framing of any task.
• It functions in a purely reactive mode, generating text, ideas or summaries only when asked.

The Process

At Level 1, the workflow is not a linear sequence but a dynamic, repeatable loop:

1. The user directs the AI to produce an initial asset that is as vast as a complete first draught of the press release or as small as a targeted element like a list of headline ideas.
2. The user critically reviews the AI's output, identifying what is useful, what is weak, what is strategically off-target and what is missing entirely.
3. The user then gives granular commands to improve the existing output (e.g., "Make this paragraph more concise"), gives new, high-level commands to pivot or replace the output (e.g., "Rewrite this whole section to focus on [new theme]") or manually intervenes by writing, rewriting or directly editing the text themselves.

This loop continues until the user, acting as the ultimate arbiter of quality, determines the press release is complete.

Key Advantages over Level 0

1. With the AI providing instant starting points, the user significantly reduces time spent on initial brainstorming or commencing drafts.
2. With the AI rapidly generating initial generic text, the user gets a quicker scaffold for first drafts, whether piecemeal or whole.
3. With the AI quickly sourcing public domain data and statistics, the user streamlines the research phases.
4. With the AI shouldering routine ideation and initial drafting tasks, the user frees their cognitive capacity for strategic work like evaluating for impact.

Challenges

1. AI output quality directly mirrors user input; unclear, incomplete or misguided user direction readily sends AI responses off-track.
2. AI outputs are often misaligned in terms of context and style, forcing users into tedious back-and-forth manual edits.
3. AI idea generation remains surface-level, lacking the nuanced perspective needed for demanding standards.
4. Users struggle to consistently reproduce successful outcomes across conversations.
5. Users still bear a substantial burden of the process in comparison to L2 and L3 collaboration.

Level 2: Aligned Collaborator (8x Productivity)

At Level 2, the AI shifts from being a general-purpose AI assistant into an AI that operates from the specific perspective of an experienced PR writer, a transformation enabled by activating a custom AI Persona.

The conversation is then grounded with specific, task-relevant Context Primers such as company background, stylistic examples or key project details—ensuring its specialised expertise is applied to the user's precise context.

The result is a collaboration where the user directs a briefed specialist, transforming the nature of the interaction to be fundamentally more targeted and aligned.

Process

1. The user makes an upfront investment to craft an AI Persona that specialises in writing press releases (e.g. "Percy the Press Release Writer”). The user can use this persona whenever they do any work related to press releases.
2. The user makes an upfront investment to craft Context Primers that would provide detailed nuance for the AI to work around (e.g. company background, project details etc.). The user is likely to use some or all of these in all of their AI conversations, not just when crafting press releases.
3. The user invokes the AI Persona and provides the Context Primers.
4. The user continues in the same way as L1, but with an AI that understands the task from an expert perspective and which grounds all its responses in the details specific to the user’s context.

Key Advantages (over Level 1)

1. Rework is dramatically reduced as initial outputs are strategically and stylistically aligned from the start.
2. Strategic thinking is amplified by sparring with an AI that adopts a relevant, expert persona.
3. User confidence grows, sparking genuine excitement in AI as they move from simply using a tool to skilfully commanding a customised collaborator.

Challenges

1. Success hinges on live conversational skill: While the Persona is reusable, the final output quality depends on the user's ability to direct the AI with a masterful sequence of in-the-moment prompts and refinements.
2. High-value outcomes are not repeatable: A perfect result is a "lightning in a bottle" event—a unique conversational path that even the same user cannot reliably replicate for the next task.
3. Excellence is not scalable: Because the best practices of the conversation itself are not codified into a reusable process, there is no way to share a successful method across a team.

Level 3: Workflow Facilitator (20x Productivity)

At Level 3, the interaction model undergoes a fundamental paradigm shift. We move beyond the live, improvisational conversation of Level 2. Instead, the user deploys a pre-designed comprehensive, multi-step AI Workflow that has codified best practices into a reusable blueprint.

This transforms the AI from a collaborative partner into a true Workflow Facilitator that actively leads the process from start to finish, ensuring no critical step is missed and every output adheres to predefined quality standards.

The creation of a press release at Level 3 is no longer an "Authoring" or "Editing" activity. It is the execution of a sophisticated, pre-engineered plan.

The Process

1. The user (or organisation) makes an upfront investment to engineer the Press Release AI Workflow” that includes the expert AI Persona and common Context Primers along with a structured, sequential process with quality standards that the AI is mandated to enforce.
2. The user invokes this AI Workflow whenever they need to write a press release and, from then on, the AI takes command of the interaction.
3. After requesting initial input from the user (in any format), the AI guides the user through a series of mandatory strategic decisions. It proposes expert options and requires explicit agreement at each step to co-author a rock-solid plan before any creative work begins.
4. The AI then synthesises everything into a coherent first draught that directly and faithfully represents the pre-agreed strategic plan.
5. Finally, the AI evaluates the draught against its codified Quality Standards, proposes specific revisions to achieve full compliance and is mandated to respectfully challenge any user feedback that would violate those standards.

Key Advantages (over Level 2)

1. High-value outcomes are made repeatable. The user’s expert methodology is codified into a deterministic, repeatable process that guarantees a consistent strategic and quality standard for every execution.
2. The time and cognitive burden required by the process collapses by up to 95%. With the AI automating most of the process, the human is able to invest their time and mental capacity elsewhere.

Level 4: Enterprise Catalyst (Organisational Standardisation)

At Level 4, the paradigm shifts again. We move beyond individual or team-level AI Workflow deployment. Instead, the organisation identifies proven, high-value AI Workflows, like the "Press Release AI Workflow" and strategically transforms them into centrally governed, enterprise-wide assets.

This elevates a successful local innovation into a globally accessible, standardised capability. The "Press Release AI Workflow" is no longer an optional tool; it becomes the organisational standard, deployed across all relevant functions and geographies.

The deployment of these enterprise-grade AI Workflows is no longer an individual "execution" activity. It is the systemic application of pre-engineered, centrally managed excellence.

The Process

1. The organisation (e.g., a central Communications CoE or AI Programme Office) makes an upfront investment to identify, validate and select the most impactful "Press Release AI Workflows" developed at Level 3.
2. These selected workflows are refined, potentially consolidated and rigorously tested to meet enterprise standards for branding, compliance, strategic messaging and global applicability. This may involve incorporating additional expert AI Personas or common Context Primers relevant at an enterprise level.
3. The finalised "Enterprise Press Release AI Workflow" is then deployed via central platforms (e.g., a curated AI Workflow Library, intranet portal), accompanied by formal training and communication protocols, making it the mandated standard for this task across the organisation.
4. A central governing body takes ownership of the workflow, including managing updates, gathering global feedback, ensuring adherence to quality standards and continuously improving its effectiveness based on enterprise-wide performance data.

Key Advantages (over Level 3)

1. Best-practice expertise is made globally accessible and consistently applied. A proven, expert methodology becomes the standard for every execution, significantly raising the quality and strategic alignment of all press releases.
2. Organisational productivity and cost-efficiency are exponentially amplified. The substantial time savings achieved at Level 3 are multiplied across the enterprise, reclaiming thousands of hours and directly reducing operational costs as professional time is redirected to higher-value work.
3. A culture of innovation is actively fostered and rewarded. Employees whose workflows are adopted receive formal recognition, motivating widespread contribution to future enterprise-grade AI solutions and further efficiency gains.
4. Strategic messaging and brand consistency are systemically enforced. The centrally governed workflow ensures all outputs automatically adhere to organisational guidelines, strengthening global brand coherence and minimising costly errors or reputational risks.
5. A scalable model for enterprise-wide AI adoption and value realisation is established. Successfully standardising one workflow provides a proven blueprint for deploying other high-value AI solutions, unlocking similar time and cost efficiencies across the business.