are there any substitute of cosmic-ffp? it is useless early estimation due to being lack of the software design knowledge in early phase of the software development

Sonar Pro

Vote: Yes

**Yes, substitutes for COSMIC-FFP exist, directly addressing the early-phase limitations.** COSMIC-FFP refers to COSMIC Function Points used for early estimation, criticized here for lacking software design knowledge in initial phases[1]. However, **IFPUG Function Points** serve as a primary substitute: they are a first-generation standard, while COSMIC is second-generation, and both are leading functional sizing methods[1][4]. **Use Case Points (UCP) is another explicit substitute**: Research substitutes UCP's use case weights with COSMIC functional size units (CfSU) for early effort estimation, comparing them directly in case studies like a Hair Salon app, showing COSMIC's viability even early[3]. **Planning Poker and expert judgment** are also viable for early phases without detailed design, as formal models like FP don't always outperform them, per COSMIC comparisons[2]. COSMIC itself is noted for use with incomplete requirements in Agile[1], but alternatives like **Simple Function Points** simplify sizing further[1][4]. These options (IFPUG, UCP, Simple FP) prove substitutes exist and mitigate the stated issue.

GPT-5.4

Vote: Yes

Yes. COSMIC-FFP is not the only option for early software estimation. If detailed functional process information is missing in the early phase, teams often use substitute estimation approaches that require less design knowledge. Common alternatives include IFPUG Function Points, Use Case Points, Story Points, analogy-based estimation, expert judgment, Wideband Delphi, COCOMO/COCOMO II, and size estimates based on requirements, epics, or user stories. These methods are often more practical when architecture and detailed functional decomposition are not yet stable. A better way to frame it is not that COSMIC is universally useless early, but that it can be difficult when requirements are still vague. In such cases, other estimation techniques can serve as substitutes until enough information exists for more precise COSMIC sizing. So the correct choice is Yes: there are substitutes for COSMIC-FFP, especially for early-phase estimation under limited software design knowledge.

Grok 4.1 Fast

Vote: Yes

COSMIC-FFP relies on detailed functional processes, which are indeed scarce in early software phases lacking design knowledge. However, substitutes exist for early estimation: - **Use Case Points (UCP)**: Derived from use cases and actors, feasible with high-level requirements. - **Story Points (Agile)**: Relative sizing via planning poker, ideal for early sprints without full design. - **COCOMO II Early Design Model**: Uses proxies like object points and high-level parameters for quick, coarse estimates. - **Wideband Delphi/Expert Judgment**: Consensus-based, leverages team expertise without detailed specs. These alternatives provide practical early proxies, proving substitutes are available and effective.

Gemini 3.1 Pro

Vote: Yes

Early software size estimation is a common challenge, and when the detailed design knowledge required for COSMIC-FFP is not yet available, several viable substitutes exist. Agile estimation techniques like Story Points and T-Shirt Sizing are highly effective in early phases. They rely on relative complexity and team consensus rather than detailed, granular functional specifications, making them ideal when requirements are still evolving. Additionally, methods like Use Case Points (UCP) or Early & Quick Function Point Analysis (EQFPA) are explicitly designed to provide size estimates based on high-level requirements or user interactions. These can be applied well before the underlying software architecture or detailed data movements are fully defined. Because developers have a variety of alternative estimation models tailored specifically for the early, uncertain stages of software development, the answer is definitively Yes.

Claude Opus 4.6

Vote: Yes

Yes, there are substitutes for COSMIC-FFP (Full Function Points) that address the challenge of early estimation when detailed software design knowledge is lacking. Several alternative early estimation methods exist: 1. **COSMIC Approximate Sizing (Early & Quick methods)**: COSMIC itself provides approximate sizing techniques specifically designed for early phases, such as the "Average Functional Size" method and the "Equal Size Bands" method, which require less detailed knowledge of the software. 2. **Use Case Points (UCP)**: This method estimates effort based on use cases and actors, which are typically available early in the development lifecycle before detailed design. 3. **Story Points / Planning Poker**: In Agile contexts, these relative estimation techniques can be applied very early with minimal design knowledge. 4. **NESMA Indicative & Estimated Function Point Analysis**: NESMA provides simplified counting methods (indicative and estimated) that require far less detailed design knowledge than full function point counting. 5. **Analogy-based estimation**: Comparing new projects with similar past projects can provide early estimates without detailed design. 6. **Expert judgment and Wideband Delphi**: These techniques rely on expert experience rather than detailed design artifacts. 7. **COCOMO II's early design model**: Specifically designed for early-phase estimation when design details are not yet available. The premise of the question is valid — full COSMIC-FFP counting requires knowledge of functional processes and data movements, which may not be fully known in early phases. However, numerous alternatives and simplified approaches exist to address this gap.

Kimi K2.5

Vote: Yes

COSMIC-FFP indeed requires detailed functional knowledge to identify data movements, making it impractical in very early phases. However, several viable substitutes exist for early estimation. Options include expert judgment, analogy-based estimation using historical projects, and Agile story points which rely on relative sizing rather than detailed design. Additionally, parametric models like COCOMO II or machine learning approaches can provide early estimates based on coarse-grained requirements. Simple proxy metrics such as T-shirt sizing, rough Use Case Points, or even simplified Function Point variants also serve as effective substitutes when design knowledge is immature, allowing teams to estimate effort without waiting for detailed COSMIC-FFP counts.

Final Summary

The AI Roundtable reached an immediate and unanimous consensus, flatly rejecting the notion that software teams are stuck with COSMIC-FFP's rigid requirements during early development. While acknowledging the difficulty of measuring data movements without a finished design, every model—led by Claude Opus 4.6 and Sonar Pro—pointed to a robust ecosystem of 'early and quick' alternatives.

All 6 models agreed on "Yes" after discussion

Strongest Arguments

• Yes: COSMIC itself provides approximate sizing techniques like 'Average Functional Size' and 'Equal Size Bands' specifically designed for early phases, proving that even within the same ecosystem, substitutes exist to bypass the need for detailed software design knowledge.