The authors are right to emphasize that Vendozoa probably include both filter feeders on plankton and those deriving nutrition directly from the underlying substratum on which they lie or move and to argue that hard and soft surfaces would have provided habitats offering partially contrasting selective forces favouring distinct body forms. However, my paper already argued that some Vendozoa were bifacial filter-feeding fronds and others likely to have been horizontal dwellers on soft surfaces that may have fed phagocytically on substrate microrganisms by ventral non-choanocyte cells and dorsally on plankton by choanocytes. A better interpretation of Fractofusus than that of Dufour & McIlroy [1] might be that it was just such a dorsal collar-cell and ventral substrate feeder; if so it was a presponge, not a pre-placozoan. If Ediacaran organisms of that dual feeding mode existed, the dichotomy between plankton feeders and substrate feeders was less sharp than they imply. Their comment raises seven issues: (i) conceptually, how does the ‘pre-placozoan grade of organization’ really differ from a presponge. (ii) Can we reliably infer from fossils the actual feeding mode and different cell types of Vendozoa? (iii) What are the phylogenetic relationships between Vendozoa, sponges and Placozoa; and the inferred phenotypes prior to each branch point? (iv) What is the relative timing of vendozoan, sponge and bilaterian origins? (v) Does substrate feeding offer a transition to the first animal as plausible as the choanoflagellate to presponge path? (vi) What selective force was crucial for the origin of the nervous system? (vii) What is required for good explanations of major evolutionary transitions? McIlroy and Dufour introduced the term ‘pre-placozoan grade’ to apply to the earliest rangeomorph fossils [2], but did not define or clearly explain it. They now say pre-placozoans ‘differ from presponges by their lack of …