Compositional and structural differences among dietary fibres (DFs) impact their utilization by the gut microbiota. Additionally, DFs have gained increasing attention for their crucial, yet less well-understood, role in facilitating the colonic delivery of bioactive compounds. Previously, we showed that apple cell wall analogues (aACWs) can bind and gradually release apple-derived polyphenols during faecal fermentation, transiently enhancing acetic acid production. In some donors, retained polyphenols influenced as many or more microbial families than the aACWs themselves. However, whether these effects extend to individual polysaccharides or unstructured mixtures, and how fibre architecture shapes polyphenol-mediated microbial metabolism, remains unclear. This study aimed to determine how a physiologically relevant polyphenol extract modulates the prebiotic potential of dietary fibres and to distinguish the roles of fibre chemistry and supramolecular organization in governing these effects. To this end, we used a well-defined, modifiable aACW produced via bacterial synthesis of cellulose with concurrent incorporation of pectin and xyloglucan, reproducing the hierarchical polysaccharide architecture of plant cell walls. In addition, we employed an unstructured blend (U-aACW) of the same components, as well as the individual polysaccharides (cellulose, pectin, xyloglucan, Figure 1). Each substrate was tested alone or with an apple pomace polyphenol extract (APP) at an intestinally relevant concentration in 24-h faecal batch fermentations. Alongside short-chain fatty acids (SCFA: acetate, propionate, butyrate) and microbiota composition (16S rRNA sequencing), underexplored branched SCFA (BSCFA: isovalerate, isobutyrate) were also quantified. Results showed that pectin and xyloglucan generated the highest SCFA concentrations (67.8 and 64.3 mM), followed by the three-component composites (aACW and U-aACW; ~47 mM), and cellulose the lowest (31 mM). BSCFA, associated with proteolytic metabolism, showed the opposite trend and was highest with cellulose (2.06 mM). Addition of APP consistently shifted metabolism toward carbohydrate-derived SCFA and away from BSCFA. After 24 h, aACW and cellulose exhibited the strongest APP-induced SCFA enhancement (log₂ fold change ≈ +0.25), whereas U-aACW, pectin, and xyloglucan showed smaller increases (+0.12 to +0.14 log₂ FC). Conversely, APP addition led to transient reductions in BSCFA that persisted modestly in structured systems (aACW and cellulose; −0.1 to −0.13 log₂ FC). These results indicate that polyphenols most effectively enhance beneficial fermentation when the substrate is structured and slowly degradable. Taxa β-diversity and hierarchical clustering revealed four substrate-driven groups (control/cellulose, aACW/U-aACW, xyloglucan, and pectin), indicating that fibre chemistry predominantly determines community composition, with minimal influence from APP. Correlation analyses linked a broad set of taxa with metabolic outcomes—for example, SCFA production correlated with Blautia , Parabacteroides, and Faecalibacterium(typically linked to eubiotic gut states) whereas BSCFA correlated with Negativibacillus and Peptoclostridium (associated to dysbiosis). APP selectively modulated these taxa, enhancing SCFA-associated and suppressing BSCFA-associated groups, particularly within structured fibres, consistent with the observed metabolic shifts. Overall, polyphenol addition caused only minor global changes in community structure but consistent, targeted modulation of taxa central to SCFA/BSCFA metabolism. These focused responses, most evident in structured and slowly fermentable matrices, amplified beneficial metabolic outputs.