When Demand Started Moving Faster Than Supply
For decades, demand planning was built on a comforting assumption: demand moves slower than supply. We forecast, we plan, we make inventory available, and the system absorbs small errors through buffers. That assumption quietly collapsed with the rise of D2C and quick commerce.
Today, demand doesn’t wait for your planning cycle. It forms in real time, triggered by a viral reel, a micro-promotion, a weather spike, or a delivery promise shrinking from two days to ten minutes. In India, platforms promising 10–30 minute delivery have redefined not just consumer expectations, but manufacturer response expectations as well. When a quick-commerce player asks an FMCG brand, “If we deliver in 8 minutes, why do you need 2 days to replenish?”, it’s not rhetoric—it’s a structural challenge to the old operating model.
This is where demand sensing moves from theory to necessity.
Demand sensing is not about predicting the future better. It is about detecting what is changing right now—early enough to still do something about it. In D2C and quick commerce, that window is measured in hours and days, not weeks.
Why Demand Sensing Matters More in D2C & Quick Commerce
D2C and quick commerce amplify three realities that traditional forecasting wasn’t designed to handle.
1. Hyper-Volatility
Demand today is event-driven and contextual. A flash sale can triple volumes in an afternoon. A sudden rainstorm can double noodle sales in one city and kill ice-cream demand in another. A cricket match, a festival, or even a meme can swing category demand overnight. Traditional weekly or monthly forecasts simply cannot react fast enough.
2. Near-Zero Lead Times
When consumers receive groceries in 10–15 minutes, retailers expect manufacturers to replenish in hours. Monthly or even weekly replenishment cycles become liabilities. The planning horizon collapses, and the margin for error disappears.
3. Granular, Fragmented Demand
Demand is no longer aggregated through a few distributors. It is fragmented across:
Thousands of SKUs
Multiple channels (GT, MT, D2C, marketplaces, quick commerce)
Hundreds of micro-locations (dark stores, pin codes)
What sells in one Bangalore dark store on Monday morning may not sell in a Mumbai suburb on Friday night. This granularity demands granular sensing, not just granular data.
The old signal chain—consumer buys → retailer orders → manufacturer reacts—is too slow. Demand sensing short-circuits this chain by capturing demand at the point of consumption and translating it into supply action almost immediately.
Demand Sensing ≠ Forecasting ≠ S&OP (How They Fit Together)
One of the biggest mistakes organizations make is treating demand sensing as a replacement for forecasting or S&OP. It is neither.
Forecasting: The Structural View
Forecasting answers a long-term question:
“What demand pattern should we provision the system for?”
It relies on history, aggregation, seasonality, trends, and promotions. It is indispensable for capacity planning, long term procurement, and financial projections. Forecasting provides stability and direction.
S&OP: The Alignment Engine
S&OP (or IBP) is not a model—it is a management discipline. It aligns commercial ambition, supply capability, and financial constraints over a medium-term horizon. It forces trade-offs between growth, service, profit, and cash.
Demand Sensing: The Near-Term Reality Check
Demand sensing answers a different question:
“What has changed inside the replenishment lead time, and can we still act?”
It operates:
At higher frequency (daily or near-real-time)
At finer granularity
With an execution bias
If forecasting is your map, S&OP your route plan, then demand sensing is your windshield. You don’t redesign the journey every time you see traffic—but you absolutely slow down, change lanes, or take a detour when conditions change.
When used correctly, demand sensing refines execution without destabilizing the long-term plan.
The Reality Check: Lead Time Is the Bottleneck
Here is the most important—and most ignored—truth about demand sensing:
If you cannot act on a demand signal within your replenishment lead time, sensing it is useless.
Many demand sensing initiatives fail not because of poor algorithms, but because of misaligned response capability. Organizations sense demand changes 30 days out while needing 45 days to respond. The result is frustration, not value.
Demand sensing must be designed backwards from response capability:
Production flexibility
Inventory positioning
Logistics speed
If your production lead times are long, demand sensing still has value—but primarily in allocation and prioritization, not in creating new supply. For example, sensing a surge allows you to redirect limited inventory to the highest-value channels rather than reacting after stockouts occur.
Quick commerce has forced many companies to confront this reality and aggressively compress internal lead times—sometimes from days to hours. Without that execution muscle, demand sensing becomes nothing more than an elegant dashboard.
Signals That Actually Work (And Signals That Mislead)
Demand sensing lives and dies by its signals. The art lies not in collecting more data, but in choosing signals with a clear cause-effect relationship to demand.
Signals That Work
Actual sales / POS data: The strongest signal—real consumption.
Inventory depletion at last mile: Indicates true demand vs stockouts.
Web traffic + conversion metrics: Especially “add-to-cart” for D2C.
Promotions and price changes (yours and competitors’).
Weather data for weather-sensitive categories.
Event calendars (festivals, IPL matches, holidays).
Advertising intensity.
Signals That Mislead
Raw social media buzz without sales correlation.
One-day spikes caused by data errors.
Monthly internal forecasts treated as live signals.
Overreaction to competitor stockouts or promos.
Vanity metrics like page views without conversion context.
The principle is simple: hard demand signals first, soft signals second. Every signal must earn its place by demonstrably improving short-term prediction accuracy or bias.
Granularity: How Deep Should You Go?
Granularity is both a weapon and a trap.
The rule is not “go as granular as possible,” but:
Go as granular as demand behaves differently and you can respond differently.
SKU-Fulfilment Centre-Channel is a strong starting point for many FMCG and D2C organizations. Going deeper (SKU-store-day) only makes sense where:
Volumes are high enough to be statistically meaningful
Fulfilment decisions are actually made at that level
Too much granularity without sufficient signal creates false precision. Mature organizations adopt a hybrid approach—coarse planning where patterns are stable, fine sensing where volatility and value justify it.
From Signal → Decision → Execution: Closing the Loop
Demand sensing delivers value only when it forms a closed loop.
Sense – Continuously ingest sales, inventory, and contextual signals.
Interpret – Decipher evolving demand patterns, especially demand shifts.
Decide – Generate recommendations or automated actions for smooth flow of goods.
Execute – Resize inventory buffers, trigger replenishment, reallocation, production, or promotion changes.
Learn – Compare outcomes to expectations and refine models to improve prediction accuracy and buffer sizing.
This loop often runs daily, sometimes multiple times a day during peak events.
Examples of Practical If-Then Rules
If sales exceed forecast by >15% for 3 days → expedite replenishment, look for missing demand driver.
If inventory = 0 and sales = 0 → treat as lost demand, not demand drop.
If promo underperforms in first 48 hours → scale back supply.
If substitution rate spikes → prioritize replenishment to that node.
If demand > supply → allocate to highest-margin or strategic channels.
Automation handles the routine; humans manage the exceptions.
A Practical Operating Model
Technology alone does not make demand sensing work. Operating discipline does.
Key Roles
Demand sensing analyst / S&OE manager
Supply and replenishment planner
Sales & marketing interface
Data/IT support
Executive sponsor
Cadence
Daily control-tower huddles (15–30 mins)
Weekly S&OE review
Monthly S&OP integration
Governance
Clear escalation thresholds
Override discipline with reason codes
One version of truth
CFO-relevant KPIs tied to accountability
When done right, meetings shift from debating numbers to making decisions.
KPIs That Prove Value (CFO-Proof)
Demand sensing must show business impact, not just analytics sophistication.
Key metrics include:
Forecast accuracy at lead time
Forecast bias
Fill rate / OTIF
Lost sales reduction
Product freshness at point of sale
Inventory turns / days of supply
Waste and expiry reduction
Expedite cost reduction
Promo ROI
Planner productivity
Customer satisfaction metrics
These connect demand sensing directly to revenue, margin, and working capital
Closing Reflection: What Demand Sensing Really Reveals
Demand sensing does not magically fix supply chains.
It reveals something more uncomfortable:
Whether your organization is capable of responding at the speed the market demands.
When paired with execution agility, demand sensing becomes a competitive weapon—lifting service levels while reducing waste and inventory. Without that agility, it simply tells you, in real time, what you are failing to deliver.
In D2C and quick commerce, the winners will not be those who predict best—but those who sense early and respond decisively.
