Sena forecasts demand from what consumers across a market purchased, read by segment, format and price tier. Every figure opens to the purchases underneath it.
A demand forecast sets what gets made, held and committed for the period after it. Three patterns bend it before the arithmetic starts.
Own shipment history is the series the model runs on. It records what a company managed to sell, which is demand filtered through whatever was in stock and on shelf at the time.
A single figure covers a market where formats, price tiers, and segments move in different directions. The parts cancel inside the total, and the total looks stable.
The model recognizes a change once the change is already in the history. By the time the series shows the turn, the decision it should have informed was taken two cycles back.
Sena is the decision AI with access to real-world data. It counts what consumers across a market purchased and when, including the purchases made from every other player, so a projection rests on the category and reaches past a single company's shipments.
Sena counts category purchases across the market, so the series behind a projection describes what buyers did and the supply they met.
Sena reports movement per segment, format, and price tier, so the parts of a market move independently in the projection.
Sena captures purchase per market on request, so a change in direction reaches a team inside the cycle it began in.
A demand forecast holds up on two things: whether the series underneath it measures the category or one company's shipments, and whether the parts of the market were projected apart before they were added together.
Demand forecasting projects how much of a category buyers will purchase in a future period for a named market and a named segment. It describes the market, and it holds regardless of which company ends up serving it.
The two terms travel together and answer opposite halves of the same problem.
| The question | What it produces | |
|---|---|---|
| Demand forecasting | What will the market buy? | A projection of category demand, per segment and period. |
| Demand planning | What should be held against that? | Stock, capacity, and replenishment decisions. |
Forecasting is the input, and planning is the response. A planning process running on a forecast built from shipments optimises the supply of a demand figure that supply already shaped.
Four approaches cover most of what runs in practice, and each carries a known limit.
Project the past series forward, with seasonality and trend fitted. Fast and stable, and blind to anything the history has yet to contain.
Relate demand to drivers such as price, weather, or income. Explanatory, and only as good as the driver data behind them.
Ask the people closest to the market. Catches turns early and carries whichever bias the room brought.
Count what buyers purchase across the category as they purchase it, then project each segment apart. Reads the market ahead of the shipment record.
Most teams run the first and supplement with the third. The fourth is the one that changes what the other three are built on.
Sena builds the series from four collected inputs, applied per market.
| Input | What it answers |
|---|---|
| Receipts | What buyers purchased across the category and when, so the series counts demand met by every player. |
| Store captures | Price and availability in real outlets, which separates a demand fall from a supply gap. |
| Geo-verified photos | Facing and pack read off the image, so an empty shelf is recorded as an empty shelf. |
| Stated preference | What buyers say they will purchase next, which reaches the part of the period the history has yet to cover. |
The team's own numbers form a separate row. Shipment history and planning models feed in through 250+ integrations, so own volume can be set beside the category moving around it.
Accuracy is usually reported as one percentage, and four things decide it.
Category purchase describes demand; own shipments describe supply meeting demand.
A figure accurate at the market level can be wrong at every segment inside it, and segments are what production runs against.
Precision falls with distance, and a projection quoted to one decimal twelve months out invites a decision the series is too coarse to support.
A forecast checked against what buyers went on to purchase improves. One checked against its own revision does the opposite.
The projection is used in three places: a production or buying commitment, a launch sizing, and a mid-period review after a series of moves.
Real people share what they buy and prefer, under explicit consent. Sena captures it directly at the source, so every figure traces back to where it came from whenever a number comes under question.
Consented purchase supplies the series, and Sena reports each segment with the purchases behind its direction.
Sena records what buyers met in the outlet, so a fall caused by an empty shelf reads differently from a fall in wanting.
Sena projects formats, tiers, and segments separately, so the parts of a market keep their directions.
Any point in the series can be opened to the weeks of purchase that produced it.
A projection carries weight only as far as the series underneath it, and most series measure supply meeting demand. Sena builds every figure from real-world signal captured when the question needs it, from consumers purchasing across the category through to the dated capture a forecast review is checked against.
Counts what buyers purchased across the category and when, including purchases made from every other player, so the series describes the market.
Reads price and availability in real outlets, which is what separates a fall in demand from a gap in supply.
Signal arrives from the consumer network under explicit consent. What buyers expect to purchase next is stated by those buyers.
Shipment history joins over 250+ integrations, so the company's own record can be read against the category series.
Every point in the series holds its market and its capture date, so a projection reopens on the evidence it was built from.
Historic series and shipment exports become one base, so every period the team has measured sits together.
Four owners read the same series, and each needs it at a different level before a commitment is signed.
| Team | What they own and what they need |
|---|---|
| Demand planning | The response. Needs a forecast built on category purchase, per segment. |
| Supply and production | The commitment. Needs the level to match what production runs against. |
| Category and range | The mix. Needs formats and tiers projected apart, with direction per segment. |
| Planning and finance | The number. Needs a horizon stated honestly and gaps opened to evidence. |
The same dated series, built on whatever the category leaves behind.
Category purchase per format and price tier, with availability recorded beside it.
Household take-up across providers and what buyers state they will move to.
Subscriber movement between tariffs and operators, period over period.
Category purchase by specification and price band, with replacement timing stated.
One mechanism, applied per market and per category. Each step is documented, which is what keeps the projection defensible.
Explicit consent, markets named
Real contributors share what they buy and pay under explicit consent, in the markets named.
Availability recorded alongside
Sena builds a dated series of category purchase, with availability recorded alongside it.
The horizon stated
Each segment, format and tier carries its own direction forward, with the horizon stated.
Most demand forecasts run on the one series a company holds in full, which is its own shipment history. Sena measures the category that history sits inside.
| Capability area | Typical setup | Sena |
|---|---|---|
| The underlying series | Own shipments and orders. | Category purchase across every player in the market. |
| Demand against supply | Combined in one number. | Separated, with availability recorded in the outlet. |
| Level | Market total, split downward by rules. | Each segment, format and tier carrying its own series. |
| Speed of a turn | Visible once it enters the history. | Visible in the cycle it began in, from dated captures. |
| New buyers | Inferred from a lift in the total. | Counted, so growth from new buyers separates from growth in frequency. |
| A miss | Explained against the previous revision. | Opened to the purchases that happened instead. |
| What comes back | A number for the period. | A projection per segment, with the series and dates behind it. |
Three situations where a category-built series changes the commitment.
Project each format and tier apart, so a production run is sized against the segment it serves.
Read availability beside purchase, so a drop caused by an empty shelf is answered with supply, and a drop in wanting is answered differently.
Open the period to what buyers purchased instead, so the review improves the next projection.
A projection describes the market. The four decisions below are what a team does with it.
Every growth decision Sena answers, in one place
The ceiling a series runs under
Sizes what the category could return, which bounds any projection built inside it.
Explore market potential analysis 02How much of it comes here
Divides category demand across the players holding it, which turns a category projection into a company one.
Explore market share analysis 03Acting on what the series shows
Names the lever that moves own volume inside a category already measured.
Explore market penetration strategy 04The plan a projection sizes
Sets every growth move against the demand a forecast says will be there.
Explore business growth strategyThe walkthrough builds a dated series for one category in one market, projects each segment apart, and opens the purchases behind any point in the series while the team watches.
Name the category and the market, and we will build the series.