← Back to archive
Issue #005May 9, 2026

Demand Planning & S&OP in 2025: AI Takes the Forecast

Demand planning is moving beyond monthly statistical forecasting into AI-driven sensing, simulation, and probabilistic decision-making. This issue breaks down the market, vendor moves, and the questions buyers should force into every demo.

SC Radar NewsletterIssue #005May 9, 2026

Opening Hook

McKinsey's benchmark still lands hardest: AI-powered forecasting can cut supply chain forecast errors by 20% to 50% and reduce product unavailability by as much as 65%. That single stat explains why demand planning budgets are still getting signed in a cautious software market. The category is no longer selling better curve-fitting. It is selling faster reaction time: sensing demand shifts earlier and turning S&OP from a monthly meeting into a live operating system.

Market Snapshot

Mordor Intelligence pegs the demand planning solutions market at $4.7 billion in 2025, while Grand View Research sized it at $4.8 billion in 2024. The drivers are familiar to every supply chain leader: volatile demand, pressure to release working capital from inventory, and the cost of getting the forecast wrong when promotions, channel mix, and lead times shift faster than planners can remodel them manually.

That changes the buying criteria. Teams are no longer just asking for consensus workflows and baseline forecasting. They are asking whether the platform can ingest POS, orders, weather, pricing, and supplier signals in near real time, model multiple scenarios, and connect the plan directly to inventory, supply, and commercial decisions.

Vendor Moves

  • o9 Solutions: o9 keeps pushing the category toward decision-centric planning. Its March 2026 APEX launch framed the platform around agile, adaptive, and autonomous planning and execution, backed by more than 130 go-lives in 2025. o9's pitch is no longer "better planning software." It is a continuously learning control layer across demand, supply, and commercial planning.
  • Kinaxis: Kinaxis has moved decisively from orchestration language into agentic planning. Maestro Agents went live in October 2025, and Maestro Agent Studio followed in April 2026, giving planners a no-code way to build AI agents inside workflows. For manufacturers already running Kinaxis for scenario planning, this is a bid to own the next layer of planner productivity.
  • Blue Yonder: Blue Yonder is tightening the connection between forecast, supply response, and execution. Its planning roadmap centers on unified demand and supply planning plus AI agents inside day-to-day workflows. The strategic point: Blue Yonder wants demand planning to be less of a forecasting module and more of an operating nerve center tied to inventory, fulfillment, and labor.
  • SAP IBP: SAP is steadily making IBP smarter rather than flashier. The 2508 release added AI-assisted forecast explanations in planning workflows, and later releases pushed forecast-results analysis further into general availability. For SAP-heavy enterprises, the message is clear: the platform is evolving from a statistical engine into a more explainable, AI-assisted planning environment without forcing a stack change.
  • Anaplan: Anaplan made the clearest inorganic bet in the category by acquiring Syrup Tech in September 2025. Syrup brought AI-native retail forecasting that uses granular sales and external signals, while Anaplan contributes scenario planning and cross-functional modeling. The move sharpens Anaplan's story with retailers and other inventory-intensive businesses that want connected planning plus more precise demand sensing.

The AI Angle

The AI opportunity in demand planning is not one feature. It is three capabilities.

First, ML-driven demand sensing helps planners incorporate fast-moving signals that old monthly models miss: retailer sell-through, short-term orders, weather swings, pricing moves, social activity, and channel shifts.

Second, real-time data integration matters as much as the model itself. A forecast that updates daily but waits on stale ERP extracts is not intelligent. The vendors gaining ground are the ones connecting planning to upstream commercial data and downstream supply constraints in the same workflow.

Third, probabilistic forecasting is becoming a better operating language than a single-number forecast. Supply chains do not need one "right" number; they need confidence bands, scenario ranges, and explicit trade-offs so commercial, supply, and finance teams can decide faster under uncertainty.

Buyer's Angle

Supply chain teams evaluating demand planning and S&OP platforms in 2025 should press vendors on five questions:

  1. Where does forecast lift actually come from? Ask how much improvement comes from ML, what signals are used, and how lift is measured versus your current baseline.
  2. How quickly can the system absorb new demand signals? Weekly batch updates are not real demand sensing.
  3. Can planners understand and override the model? Black-box accuracy claims do not help if teams cannot explain a forecast in executive S&OP.
  4. How tightly does the forecast connect to supply, inventory, and finance plans? Better prediction without downstream action is just prettier analytics.
  5. What governance exists for AI-generated recommendations? You need audit trails, human-in-the-loop controls, and clear thresholds for automation.

Stat of the Week

$4.69 billion. That is Mordor Intelligence's 2025 estimate for the global demand planning solutions market, a reminder that this is now a core enterprise planning battleground, not a niche add-on.

CTA

If you need the vendor map, not just the headline, the SC Radar Premium Report is $149 and gives you competitive positioning, category scoring, and buyer guidance. If you want continuous coverage across planning, execution, and supply chain AI, the SC Radar Monthly Subscription is $99/mo. See both options at SC Radar pricing.