The core mistake in facial aesthetics is treating gender as an add-on rather than the structural foundation. Most procedures begin with an idealized template—high cheekbones, narrow jaw, soft lips—then attempt to layer masculinity or femininity afterward. This sequence guarantees friction. Gender characteristics are not decorative; they are the load-bearing architecture of the face. Any intervention that ignores this order produces the familiar result: a face that looks 「done」 rather than coherent.
Begin with the minimum viable skeleton. The male craniofacial frame is wider at the gonion and supraorbital ridge, longer in the vertical maxillary dimension, and flatter in the forehead-to-nose angle. The female frame is narrower at the bigonial width, shorter in the upper facial third, and more curved at the forehead-glabella transition. These are not preferences; they are measurable vectors. A 6 mm increase in bigonial width shifts perceived gender more reliably than any soft-tissue filler. Therefore the first decision in any plan is whether to preserve, reduce, or augment these bony landmarks. Everything else is secondary.
Soft tissue follows bone, never the reverse. Subcutaneous fat distribution differs systematically: males carry more in the submental and lateral cheek compartments, females more in the medial midface and temporal hollows. When an injector places volume without first establishing the underlying skeletal envelope, the fat sits in the wrong plane and creates the 「puffy yet angular」 paradox seen in over-treated male faces. The minimal correction is to map compartment dominance before any needle enters the skin. If the skeleton already reads male, additional medial cheek volume is usually redundant and risks feminization. If the skeleton is ambiguous, the smallest effective volume is placed only where the compartment map shows deficiency relative to the target gender vector.
Proportions function as the verification layer. The golden ratio is frequently cited yet rarely gendered. A more precise check is the ratio of upper to lower facial height combined with intergonial to interzygomatic width. In male faces the target intergonial width is 70–75 % of interzygomatic width; in female faces the same ratio drops to 65–68 %. These numbers are not aesthetic ideals; they are statistical thresholds at which observers reliably assign gender. Crossing the threshold without intending to produces the 「feature confusion」 the classification description warns against. The minimal viable check is therefore a single ratio measurement taken on standardized photographs before any treatment plan is finalized.
Expression patterns close the loop. Dynamic lines are not merely age markers; they are gender signals. Male forehead lines tend to be deeper and more horizontal because of larger frontalis and corrugator muscles anchored to a heavier brow ridge. Female lines are finer and more oblique. When neuromodulators are dosed identically across genders, the male face loses its characteristic resting tension while the female face loses its characteristic micro-movement. The minimal effective dose is therefore calibrated to preserve the amplitude of gender-typical motion rather than to erase all lines.
The sequence above—skeleton first, compartment mapping second, ratio verification third, dynamic preservation fourth—constitutes the minimum viable product for gender-coherent outcomes. Any additional step (extra filler, aggressive bone shaving, simultaneous multiple modalities) must demonstrate incremental improvement against this baseline. In practice, most cases stabilize after the first three steps. The fourth step is rarely required once the static architecture is aligned.
Over-feminization occurs when the second or third step is skipped. A common pattern is heavy medial cheek volumization on a male skeleton whose bigonial width already exceeds the female threshold. The result is a face that reads neither male nor female but 「treated.」 The same logic applies in reverse: aggressive jawline definition on a female skeleton whose intergonial width is already narrow produces a face that appears masculinized even when soft tissue remains untouched. Both errors trace back to the same root: treating gender as a variable to be adjusted rather than the constraint within which adjustment occurs.
The practical implication is restraint at the planning stage. A single, correctly chosen bony or volumetric move that respects the gender vector outperforms three moves that fight it. This is not conservatism for its own sake; it is recognition that facial gender is a low-dimensional system. Once the primary vectors are aligned, further variables add noise faster than signal. The aesthetic result that survives this filter looks neither masculine nor feminine in isolation; it looks like the patient, only more so.
