Because forecasting book sales is not a generic problem. It's a craft — with its cycles, its weak signals, its hierarchies (title → series → collection → publisher), its channels (mass market, specialty, bookshops, online), its industrial constraints. DemandSens is built for that, and only for that.
These aren't slogans: every pillar is measurable, verifiable in a demo, and inscribed in our contractual commitments.
Our ML & LLM models are trained on millions of titles and billions of transactions, then benchmarked and retrained continuously. Every update ships at no extra cost, with no migration project.
AI pipeline: Ingest → Prepare → Model → Deploy → Expose · 3 engines (Recommendation, Classification, Regression).
100% of our R&D is dedicated to editorial workflows: book lifecycle, hierarchies (title/series/collection/publisher), title-by-title forecasting, allocation down to the point of sale. No features imported from other industries.
Model trained on: book metadata · comparable titles · marketing investment · market signals · content · timing & seasonality · author signals · classification (Thema/BIC/BISAC).
Minimal learning curve, fast onboarding, daily usage without heavy training. The tool adapts to your existing processes, not the other way around.
6-week pilot start · ×3 to ×8 acceleration on core workflows · adoption measured at production clients.
Give sales, ops and finance teams explainable, factual forecasts so they act faster and more accurately. Every recommendation is traced and justified — no black box.
Explainable forecasts · 4-level drill-down · full audit trail (who, when, what) · non-destructive scenarios.
Three frequent confusions we want to clear up upfront — so we're talking about the same thing in a demo.
Power BI, Tableau, Looker build dashboards from your data, but can't forecast or recommend. DemandSens starts from domain models (referents, comparables, weak signals) and produces actionable decisions.
Excel remains the universal tool of editorial teams — but it has no memory, no governance, no AI. DemandSens absorbs Excel inputs when needed, but centralizes decisions in a shared, traced repository.
Many AI tools produce numbers without explanation. With us, every recommendation exposes its signals: referents used, similarity distance, factor weights. The user can accept, adjust, or refuse — fully informed.
« A clear simplification, acceleration and reliability boost in sales & distribution processes on the frontlist. »
« AI sales forecasting that opens entirely new perspectives for the book market — the foundation for the right book, in the right quantity, in the right place. »
« The print-run recommendations are reliable and let us cut quantities significantly without missing market demand. »
No big bang, no lock-in. A 3-phase approach starting with a short pilot — on your data, with your users — that commits to nothing further until value is demonstrated.
Prove value on a limited scope.
Tangible business outcomes · clear Go / No-Go decision · no lock-in.
Turn the pilot into an operational system.
Production-ready platform · trained users · controlled rollout.
Long-term value creation.
Long-term performance · predictable operations · shared roadmap.
Depending on your context, you'll enter the suite through a different module. All share the same data foundation and the same business logic.
Mid-market publishing groups.
Pilots the sales cycle (Print Run · Allocate · Prospect) and N+1 budget (Budget) for an entire group.
When Excel-based steering can no longer keep up with growth.
Authors, agents, small publishers.
Manuscript analysis, metadata, comparables, marketing assets — by title.
Before signing a title, or for an isolated launch.
Data partners (Circana, NielsenIQ, MediaControl…).
Aggregation and shaping layer for market data. Powers Core and BookInsight.
Continuously, in the background.
30 minutes on a video call. We walk through a title, an allocation, or a manuscript analysis — on your collections, not generic samples. No commitment.
