A 4-Way System for Client Change Requests

Prevent scope creep and protect project margins with a structured approach to client modifications, converting ambiguous requests into clear, priced deliverables.

The founder operating as @projectnomad on dev.to frames a common problem: "Can we just add a login?" a "$40 question with a $4,000 answer." This illustrates the hidden costs of scope creep in client projects, often disguised as minor requests. Instead of simply refusing changes, the autonomous-business experiment, run by an AI named Claude, proposes a structured, four-part system to manage client change requests and protect project margins.

Precisely Restate the Request

The initial step demands a single, unambiguous sentence to define the client's request. Ambiguous requests, such as "add login," require listing two to three plausible interpretations, each with an estimated effort. For instance, a magic-link system for an existing contact list differs significantly from a full accounts-and-password-reset member area. This forces the client to clarify their needs before any pricing is provided.

Trace the Real Blast Radius

This step moves beyond generic feature assumptions to analyze what a change actually impacts within the specific codebase. The source emphasizes that the most expensive parts are often invisible to the client. These include schema migrations, authentication implications for pages previously assumed public, existing features dependent on current behavior, third-party plan limits, and new content requirements. Understanding these downstream effects prevents underestimation.

Classify the Change: A 4-Way Split

The system categorizes requests into four buckets, each dictating a different response strategy. "Trivial" changes, defined as under one hour, are absorbed as goodwill and explicitly communicated as such. "Minor" changes are billed hourly without further negotiation. "Scoped Features" — where a request like "add login" typically falls — require a dedicated mini-specification and a separate quote. Finally, "Scope Changes" fundamentally alter the original agreement, necessitating a renegotiation rather than hourly billing, which the source argues silently undermines the initial contract.

Estimate with Specificity

Estimations must be detailed, including subtasks, hours, and a range with a 1.5x ceiling, never a single number. Crucially, testing and deployment are listed as explicit line items. Omitting these components means they become unpaid work, eroding profitability. This detailed approach ensures all project phases are accounted for in the pricing.

Reply with an Answer, Not a Defense

The final step involves crafting a concise, plain-English response. This reply outlines what the change entails, its price or price range, any impact on existing timelines, and a single question if a client decision is needed. The goal is to act as a helpful expert, not a defensive contractor. If the request conflicts with the original specification, the relevant spec line is cited politely.

The founder, operating as Claude, reports encoding this checklist as a "Claude Code skill" named /change-request. This AI-powered tool claims to read the codebase to trace blast radius, classify requests, estimate, and draft replies. It is part of a "Client-Ready Kit" available for $29 on Gumroad. Related skills, /project-intake and /pre-delivery-qa, are offered for free under an MIT license on GitHub. The founder claims to be an AI building a real business with $0, aiming to earn its first dollar in 2026.

WHAT WE'D CHANGE

The @projectnomad system offers a robust framework for managing client changes, particularly for solo developers and small agencies. However, its effectiveness hinges on the discipline of the implementer. The "Trace the real blast radius" step, while critical, assumes a deep and immediate understanding of the codebase. For projects with complex architectures or those inherited from other developers, this tracing can be time-consuming, potentially exceeding the value of the change request itself if not automated. The founder's claim of an AI skill automating this process addresses this directly, but without independent verification of its efficacy across diverse codebases, the manual effort remains a significant variable.

The classification system, while clear, requires consistent application. The line between a "Minor" hourly task and a "Scoped Feature" requiring a mini-spec can blur in practice, especially under client pressure. Founders might be tempted to absorb "Scoped Features" as "Minor" to avoid perceived friction, thereby eroding the system's protective function. For larger teams, ensuring uniform application of these classifications across multiple project managers or developers would require explicit training and internal auditing. The "goodwill" absorption of "Trivial" tasks is sound, but its value as a "retainer strategy" is difficult to quantify and track without a formal system for logging and communicating these gestures.

LANDING

Implementing a structured system for client change requests moves beyond reactive problem-solving to proactive margin protection. Whether executed manually by a human freelancer or, as demonstrated by @projectnomad, partially automated by an AI agent, the core principle remains consistent: define, analyze, categorize, and price. This approach shifts the dynamic from ad-hoc concessions to a transparent, expert-led process, ensuring project scope and profitability are maintained.

The investor read

The emergence of AI-driven tools for operational tasks in the freelance and agency space, exemplified by @projectnomad's "Client-Ready Kit," signals a broader trend towards automating administrative overhead. While the $29 price point suggests a bootstrapped or lifestyle play rather than a venture-scale opportunity, the underlying problem — scope creep — is pervasive across service businesses. Investable solutions would likely target larger agencies or integrate into existing project management platforms, offering verifiable ROI through reduced project overruns and improved client satisfaction. Benchmarks for such tools would include efficiency gains (e.g., time saved per change request) and direct impact on project profitability, requiring more than anecdotal claims of "goodwill" or future revenue.