Trend analysis is the method for turning repeated observations into a direction, and its output is only as good as the series underneath it. Sena supplies that series from recorded purchase, captured on the same basis each cycle in each market.
Trend analysis is a well-established method, and the method is rarely the problem. Fitting a line to a set of points is arithmetic. The difficulty sits one step earlier, in whether the points were gathered on a consistent basis, whether they cover the markets the conclusion will be applied to, and whether the gaps between them are short enough for the movement being measured.
A series assembled from several sources changes definition partway along. A category measured one way for six periods and another way for the next six produces a line with a genuine-looking bend in it, and the bend is a change in method. Any analysis run over that series inherits the artefact and reports it as a market movement.
Where observations are spaced widely, movement between them is invisible. A quarterly series shows a change that started and reversed inside one quarter as a flat line, and there is a strong tendency to treat the flat line as evidence of stability. The interval sets the smallest movement the analysis is capable of resolving, and that limit rarely appears alongside the output.
Series get built where the data is easiest and applied where the decision is. A line fitted to the home market and used to plan for eleven assumes the eleven move together, which is the assumption the analysis exists to test. Extrapolating in that direction turns a measurement into a guess and presents it with the authority of a chart.
Sena captures purchase, shelf and stated preference on the same terms each cycle across 190+ countries. A series built on that holds its definition from the first point to the last, which is what makes the arithmetic afterwards worth doing.
The same capture method every period, so a bend in the line is a market event.
A line per country, rather than one line fitted to the easiest market.
Capture frequent enough that a movement inside a period stays visible.
Trend analysis takes repeated observations of the same measure and establishes whether it is moving, in which direction, and how fast. In a commercial setting it is applied to demand: what consumers are buying, in what quantity, and how that changes. This page treats that application, and the series it is run over comes from recorded purchase in each market.
| Page | What it holds |
|---|---|
| Trend analysis | How the series behind all three gets built and read. This page. |
| Market signal intelligence | The earliest reads, before a shift has a name. |
| Early trend detection | The onset, named before a series can be fitted. |
| Consumer trend intelligence | What is shifting in consumer choice, named and sized. |
| Sense | What it refers to |
|---|---|
| Statistical | Fitting a line to any series of numbers. |
| Financial | Chart reading on a traded instrument. |
| Project | Tracking how milestone dates move. |
| Market | How consumer demand changes over time. This page. |
This page holds the fourth. The first is the underlying arithmetic and applies everywhere. The second and third are separate disciplines that borrow the phrase, and a reader arriving from either is better served by a page that names the boundary than by one that ignores it.
Market trend analysis applies the method to a category rather than a company: what the whole market is doing, against what one firm’s share of it is doing. The two get conflated constantly, and they answer different questions. A company can gain share in a category that is contracting, which reads as success internally and as a warning externally. Running both series together is what separates the two readings.
The consumer variant runs the series on choice: which option is gaining against which, and among whom. It needs a finer measure than category totals, since a shift between two formats inside a flat category leaves the total unchanged and moves everything that matters underneath it. That is the movement a purchase-level series records and an aggregate one absorbs.
| Property | What it fixes |
|---|---|
| Same definition | The measure means one thing throughout. |
| Same method | Points gathered the same way each period. |
| Adequate frequency | Intervals shorter than the movement. |
| Stated coverage | Which markets each point represents. |
The first two make the line honest, and the last two make it usable. A series that changes method halfway produces a bend the analysis will faithfully report. A series covering one market and labelled globally produces a conclusion applied to ten places it left unmeasured. Both errors survive any amount of statistical care applied afterwards.
Every series has a resolution limit set by how often it was captured, and the limit is a property of the collection rather than the analysis. A movement completing inside one interval is invisible at any level of statistical sophistication. That matters most in fast-moving consumer categories, where a format can gain and lose position inside two quarters. Shortening the interval is the only fix, and it is a collection decision made before the analysis begins.
A direction is a description, and the reason behind it sits with the consumers who changed their purchases.
Extrapolation assumes the conditions that produced the movement persist, which is an assumption rather than a measurement.
The analysis treats the series as given, so a definitional change partway along enters the output as a market event.
The search results for this term run to definitions, spreadsheet method pages, financial commentary, and project-management explainers. That says the phrase is shared and the commercial demand sense is a minority of the results. A reader looking for how to read a market over time arrives among pages explaining a formula, which is the arithmetic they already have.
| Output | What it settles | Where it goes wrong |
|---|---|---|
| Direction | Which way demand moved | Read off a shifting basis |
| Rate | How fast it is changing | Flattened by wide intervals |
| Per market | Where the line differs | Fitted to one country |
| Composition | What moved inside the total | Hidden by an aggregate |
The first two describe the movement, and the last two say whether the description applies here. A category flat in total and shifting hard between two formats underneath produces a horizontal line and an urgent commercial situation. Reading composition alongside direction is what keeps a stable-looking chart from concealing the thing worth acting on.
| Series | What the line is built from |
|---|---|
| Receipts | Units taken per period, per market, per format. |
| Store captures | What was available to buy in each period. |
| Geo-verified photos | Price and claim at each point on the series. |
| Stated preference | What consumers said at the time, alongside what they did. |
Reporting systems, the category history, and existing series keep the owner they have. 250+ integrations set them beside the purchase series, so an internal line and a market line sit on one row.
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.
Each period’s purchases arrive on the same basis as the last, so a series holds its definition from end to end.
Figures in this exchange are illustrative.
A separate line for each country, rather than one line applied to all.
Composition change under a category that looks stable.
Which markets are running ahead of which, measured in cycles.
A reporting system can extend a series it already holds. Whether the earlier points were gathered the same way, it can say only if somebody recorded that, since a method change leaves zero mark on the numbers themselves. This page is built from captures taken on one method, market by market, period after period.
An assembled series inherits every definition it was built from. Every point behind this page comes from the same capture routine, run in the same markets on the same terms, which is what makes a bend in the line a market event.
A recorded purchase supplies one point per period on a basis that holds across all of them.
Shelf photography records what was available at each point, which is what makes a change interpretable.
Signal arrives from the consumer network under explicit consent. How long a consumer has bought a format is reported by the buyer.
Reporting systems, category history and existing series attach over 250+ integrations.
Every point on a line opens to the outlets and the weeks underneath it.
Points accumulate rather than being re-collected, which is what gives the series its consistency.
Own the category view. Need composition under the total.
Own the demand series. Need one basis across periods.
Own the forecast. Need a market line beside the internal one.
Own the local series. Need their own country’s line.
Format composition tracked under a flat category total.
Occasion and pack mix followed cycle after cycle.
Movement between own-label and branded lines over time.
Adoption curves built from consumer take-up.
One series, extended each cycle from the same capture base, beside the reporting already in place. Every point keeps the captures behind it, so a bend in the line can be opened and explained.
Each period’s purchases and captures are added to the series on the method used for the ones before it.
Direction and rate are read per market, so the eleven countries produce eleven lines rather than one average.
Composition is examined under the total, which is where a flat category usually turns out to be moving.
Most trend analysis is sound arithmetic applied to a series that went unaudited. Sena supplies the series.
| Capability area | Typical setup | Sena |
|---|---|---|
| Series basis | Assembled from several sources. | One capture method throughout. |
| Definition drift | Enters as a bend in the line. | Held constant by construction. |
| Interval | Quarterly or annual. | Short enough to see the movement. |
| Coverage | One market, applied widely. | A line per country. |
| Composition | Hidden inside the total. | Read under the aggregate. |
| Reason | Supplied by interpretation. | Stated by the consumers who moved. |
| Evidence in a review | A chart and a source note. | Photographs, dated and placed. |
Three series questions an assembled dataset answers poorly.
Check whether a line the team already runs holds one definition end to end.
See global consumer intelligence →Read what moved between formats while the category total stayed level.
See competitive shelf intelligence →Compare country lines to see which is a year in front of the others.
See market entry and expansion intelligence →Every innovation decision Sena answers, in one place
Naming what the line shows
Takes a measured direction and turns it into a named, sized shift.
Explore consumer trend intelligence → IN9When it moved enough
Rules on whether a line has moved far enough to act on yet.
Explore early trend detection → IN5Too few points to fit
Carries the reads at the stage where a series holds too little to analyse.
Explore market signal intelligence → IN6The same method, on companies
Runs the series across a field of competitors rather than a single category.
Explore innovation benchmarking →The session assembles a category series from captures across eight cycles, fits a line per market, opens the composition under the total, and marks the markets running ahead of the others.
Bring the category and the periods to run it over.