The book trade’s on-demand printing revolution has always had an awkward final step. For twenty years, digital printing allowed a publisher to manufacture single copies economically; what it did not allow was for that capability to live natively inside the publishing house’s own nervous system. Ordering a print-on-demand copy meant leaving the ERP — the enterprise system that holds inventory, orders, royalties and metadata — and entering a separate vendor universe of portals, file transfers and manual reconciliation. That gap is now closing, and its closure is quietly reorganising how publishers think about what a print run even is.
The integration arriving this year is deeper than a convenience. Modern print-on-demand connectors plug directly into publisher ERPs, treating manufacturing as an ordinary business transaction: a sale or a reorder triggers a print order automatically, at the right facility, with the right files, with costs flowing back into title-level P&Ls without human intervention. Several mid-sized publishers have gone further, implementing dynamic logic that treats the ERP itself as the inventory — the system holds no stock for long-tail titles, knows the demand curve, and dispatches production in quantities the data supports rather than the spreadsheet guessed at eighteen months earlier.
The operational consequences compound. Reprint decisions, once monthly meetings with anxious print buyers, become continuous background processes. The distinction between “in print” and “available” — a source of quiet commercial damage for decades, as titles technically alive were practically unorderable — dissolves: every title with a digital file is, for practical purposes, a live title. Returns, the industry’s structural ulcer, shrink accordingly, because inventory is manufactured against demand rather than demand being hoped against inventory.
The economics deserve honesty. Single-copy manufacturing remains more expensive per unit than offset at volume, and the ERP integration does not change that physics; it changes the comparison being made. The relevant comparison is no longer POD unit cost versus offset unit cost, but POD total cost — no warehousing, no returns processing, no remaindering, no capital tied up in speculative stock — versus the true fully-loaded cost of the conventional chain. For titles selling in the low hundreds annually, that arithmetic has quietly inverted for several publishers, who now default new midlist acquisitions to POD-first without ceremony.
Resistance persists, and some of it is principled. Production departments built on the craft of print management see their expertise marginalised by a process that treats manufacturing as a data event. The POD vendor landscape, meanwhile, remains concentrated, and publishers integrating deeply with single vendors are asking the classic dependency questions — pricing power, service continuity, file custody — that the trade has been burned on before. Quality control on short digital runs, though much improved, still occasionally surprises publishers who specified offset habits for digital realities.
But the direction of the plumbing is unmistakable. Publishing’s front office — acquisitions, marketing, rights — digitalised years ago. The back office held out longer, protected by the physical romance of the warehouse. ERP-native on-demand production is the warehouse’s quiet surrender to the same logic: capital, space and risk are too expensive to spend on speculation when the data to spend them on demand is finally, natively, at hand. The print run as an act of faith is ending. The print run as an act of arithmetic has begun.