La Marie Beauty — Service Page Variants Scope Dispute

Overview

An unresolved scope conflict emerged during the 2025-11-11 sync call over whether redesigning 81 service pages to support product variants constitutes new scope. Asymmetric and La Marie Beauty hold opposing positions, and the dispute was escalated to decision-makers on both sides without resolution on the call.

This is a useful reference case for how variant complexity can be obscured during initial scoping when a CSV is used as the handoff artifact rather than the live system of record.

The Conflict

Asymmetric's Position: This is a scope increase

• The original estimate was based on a CSV file provided by La Marie Beauty, in which each row represented a flat, individual service item — variants were not represented.
• The work now required includes not just adding variant selectors, but full service page redesigns: new content, galleries, before/after images, and layout changes.
• Estimated effort: 40+ hours (assuming ~30 minutes per page across 81 pages, likely faster with templating but still significant).
• Melissa (Asymmetric PM) acknowledged that Asymmetric had already begun absorbing some of this work before the scope question was formally raised.

"It's just the full product page redesigns for all of them." — Melissa Cusumano

La Marie Beauty's Position: Variants are not new scope

• Square's booking flow already represented services as variants before the project began (e.g., "Vampire Facial" with options for "Face," "Face + Neck," "Face + Neck + Décolleté").
• This structure was visible in Square's UI and surfaced in JSON API responses Chris shared during earlier Zapier debugging sessions (base service + array of variant objects).
• Kimberly (La Marie Beauty PM) argued that the programmatic concept of variants was always present in the source of truth — Square — and that the CSV was an imperfect representation, not the authoritative spec.

"The source of truth in this case, which has been Square throughout, that data was already represented as variants existing." — Kimberly Gehrmann

Key Nuances

Variants vs. Enhancements

Not all 81 service pages necessarily require the same treatment. Kimberly and Roxana (La Marie Beauty's service expert) need to sync separately to distinguish:

• True variants — meaningfully different service options with distinct pricing, duration, or scope (e.g., face only vs. face + neck)
• Enhancements / add-ons — optional upsells that don't constitute a separate service variant

This distinction matters because it may reduce the actual number of pages requiring full redesign treatment.

Content Volume Escalation

Even if variant selectors themselves are considered in-scope, the volume of new content (galleries, before/after photos, expanded descriptions) attached to the redesigned pages is where Asymmetric sees the clearest scope expansion. The original service pages were minimal; the new template is substantially richer.

The CSV Problem

The root cause of the misalignment is that the CSV — not Square — was treated as the scoping artifact. When Roxana compiled the CSV, she was also actively updating content, meaning the CSV reflected a point-in-time editorial view rather than the structural reality of Square's data model. This is a process gap worth addressing in future projects: scope estimates should reference the live system of record, not a derived export.

Status at Time of Call

Action Items

• Melissa → Discuss scope conflict and cost options with Mark (Asymmetric leadership)
• Kimberly → Discuss with Lisa (La Marie Beauty) and align on their position
• Kimberly + Roxana → Sync 1:1 to categorize each service: true variant vs. enhancement, and flag which pages require multiple photos/galleries vs. which are simpler
• Both sides → Reconvene with findings to negotiate a middle path (e.g., partial scope absorption, phased delivery, or change order)

Generalizable Lessons

1. CSV handoffs obscure data structure. When a client's system of record has relational or hierarchical data (variants, arrays, nested objects), a flat CSV will flatten that structure and create false simplicity in estimates. Always validate against the live system before scoping.

2. Scope disputes are harder to resolve when work has already started. Asymmetric had begun absorbing variant work before the conflict was formally raised. Early flag → early decision.

3. Content volume is often the real scope driver, not the feature itself. Adding a variant selector UI is small. Populating 81 pages with galleries, copy, and before/after images is not. Separate the two in estimates.

4. Distinguish the source of truth from the working document. The CSV was a working document. Square was the source of truth. These should not be conflated during scoping.

Related

• ^[1]
• ^[2]
• ^[3]