Home›Solutions›Innovation›Market signal intelligence
Innovation

Reads before a name

Sena reads market signals from what consumers bought, which is the point where an indication becomes a commitment.

  • 190+ countries
  • 5M+ consumer network
  • 250+ integrations
The problem

Where early reads mislead

The earliest available indications are the loudest ones, and loudness is a poor guide to commitment. Social posts, search spikes, and press coverage all move before purchase does, which is what makes them attractive. They also move for reasons that have close to zero connection to what people will buy, and a monitoring stack tuned to volume treats both cases as the same event.

01

Attention over intention

A category can generate discussion for a week and sell exactly as it did the week before. Talking about a product and buying it are different acts performed by overlapping but distinct populations, and the gap between them is largest for the categories where discussion is cheapest. An early read built on attention measures the discourse and reports it as the market.

What shows up
  • Discussion counted as demand
  • Talkers and buyers differ
02

The vocal minority effect

Posting is a minority act and an unrepresentative one. The consumers most likely to publish an opinion about a category are the least typical members of it, skewing younger, more engaged, and more polarized. A read assembled from published opinion inherits that skew at every stage and reports it as a market movement.

What shows up
  • Publishers skew unrepresentative
  • Skew inherited by the read
03

Signals with zero geography

Online indications arrive attached to a platform rather than a place. A spike can originate almost entirely in one country and read as global, or reflect a diaspora conversation with little bearing on the market it appears to describe. A build team that acts on it commits engineering time against a movement whose location stayed unestablished.

What shows up
  • Platform replaces place
  • Location left unestablished
What Sena does for the tell

Early reads from purchase

Sena reports what consumers across 190+ countries bought this cycle and last, at a frequency that makes small movements visible. A tell read that way is already a commitment, since somebody paid for it.

Movement before the name

Caught while still small

Small shifts in what is taken are caught while they are still small.

Placed in a country

A market, not a platform

Every read attached to the market it happened in, rather than a platform.

Purchase over posting

Paid for, not published

Indications drawn from what consumers paid for, rather than what they published.

The definition

What a market signal is

A market signal is an early, ambiguous indication that consumer demand is moving. It sits before a trend has a name and before its size is settled. Market signal intelligence is the work of catching those indications, placing them, and judging which are real. The version here is read off purchase records, one country at a time.

Where this sits

Four pages read demand, and this one holds the earliest stage

PageThe stage it holds
Market signal intelligenceThe reads before any of the rest applies. This page.
Early trend detectionDates the crossing point into a trend.
Consumer trend intelligenceNames what is shifting in consumer choice, and sizes it.
Trend analysisThe method that turns repeated reads into a line.
Market signals from social media

Strong on vocabulary, weak on scale

Social listening is the most common source of early reads, and it is genuinely useful for one thing: catching vocabulary. A new word for a format usually appears in posts before it appears in a category report. What social sources measure poorly is scale, since posting volume tracks engagement rather than purchase, and placement, since a platform metric describes an audience rather than a market. Pairing the vocabulary from posts with the volume from receipts uses each source for the thing it does well.

A signal against a trend

A trend is a signal that resolved

A trend is a signal that resolved. Treating the first as the second funds builds against movements that were noise, and treating the second as the first delays a decision until the advantage has gone.

What a tell should carry

Four elements, and an online source leaves out the fourth

ElementWhat it establishes
The movementWhat appears to be changing.
The marketWhich country it was read in.
The dateWhen the capture was taken.
Purchase behind itWhether anyone paid for the change.
The loudness problem

Reading purchase inverts the ranking

Monitoring stacks surface the largest measurable movement, and the largest measurable movement online is the one that generated the most engagement. Engagement correlates with novelty and platform incentive. It correlates weakly with whether consumers changed what they buy. Reading purchase inverts the ranking, since a small movement in units taken is a stronger indication than a large movement in mentions.

Where an online read ends

Three things a monitoring stack cannot settle

01

Whether anyone acted

Discussion establishes interest, and interest converts to purchase at rates that vary enormously by category.

02

Where it happened

A platform-level spike carries an audience and needs a country attached before a build team can use it.

03

How representative it is

The people who publish about a category are a subset of the people who buy it and a skewed subset.

Two senses share the term

A securities sense and a consumer-demand sense

The search results for this term run to explainers, trading commentary, and platform marketing. That says the phrase is contested, carrying a securities sense and a consumer-demand sense in the same results.

What the tell read returns

Four outputs, and where each one goes wrong

OutputWhat it settlesWhere it goes wrong
Movement sizeHow much of the category shiftedMeasured in mentions
CountryWhere the read holdsReported per platform
DirectionWhich way it is goingFixed by a single snapshot
PersistenceWhether it survived a cycleJudged on the first spike

The first two make a read usable, and the last two say whether to keep watching. A movement that appears once and fades is a fluctuation. The same movement present across three cycles in two markets is a candidate trend. Separating them is what keeps a watch list from becoming a record of everything that ever moved.

Four reads per tell

What the tell carries

CaptureWhat the tell carries
ReceiptsThe first purchases of a format that was rare last cycle.
Store capturesA product appearing on shelves it was absent from.
Geo-verified photosWhere and when the change was first visible.
Stated preferenceWhat consumers say they expect to buy next.

Listening stacks, alerting rules, and the watch list keep the owner they have. 250+ integrations set them beside the purchase read, so an online mention and its unit evidence sit on one row.

Direct from real consumers

Shared under explicit consent

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.

Zero-party data Provenance on every figure 190+ countries PII protected
Sena for market signals

Ask Sena the tell

The purchase and the market it happened in arrive on one record, so an early read is placed before it is judged.

Figures in this exchange are illustrative.

01

Movements backed by units

Which early reads have purchase behind them, separated from those that have talk.

02

The country it started in

Where each movement was first captured, dated to the cycle.

03

Held or faded

Whether a read survived into the following capture window.

How Sena reaches the answer

What the tell read uses

A listening stack can report that mentions of a format rose. Whether anybody bought one, it can say only by inference, since the outlets where the buying happens publish zero posts.

The evidence base

The till, not the timeline

A mention records an expression. Every read behind this page starts from a purchase made in a real outlet, in a named country, on a dated cycle short enough that a small movement still registers.

190+ countries · every read placed per market
01

Consumer activity

A recorded purchase confirms a change was acted on, which is what separates a tell from talk.

02

Computer vision

Shelf photography catches a format arriving in outlets it was absent from last cycle.

03

Zero-party data

Signal arrives from the consumer network under explicit consent. What a household expects to buy next comes from the household.

04

Connect the alerting stack

Listening rules, alert thresholds, and the watch list attach over 250+ integrations.

05

Trace every answer

Each tell reports the stores, the dates, and the shoppers it was picked up in.

06

From files to databases

Each cycle is read on the same basis, which is what makes persistence judgeable.

Who reads it

Who watches the tell

Consumer planning

Own the watch list. Need purchase behind each entry.

Category strategy

Own the early view. Need the country a read came from.

Product development

Own the concept queue. Need a read before it is common.

Regional marketing

Own the local picture. Need their market read directly.

By industry

Early reads across industries

01

CPG and retail

A format reaching shelves it was absent from last cycle.

02

Beverages

First purchases of an occasion that was rare a quarter ago.

03

Pharmacy and health

Movement toward a delivery format ahead of any coverage.

04

Financial services

Early demand read from consumer card and payment records.

The mechanism

Capture to tell in three steps

One capture routine runs on the categories on watch each cycle, beside the listening stack already in place. Every read keeps its date and its country, which is what lets it be judged later.

STEP 01

Capture

Purchases and shelf photographs are taken on a frequency short enough that a small movement registers while it is still small.

STEP 02

Separate

Movements with units behind them are divided from movements with mentions behind them, so the watch list orders on evidence.

STEP 03

Raise

A read that persists across cycles is raised for a detection decision, and one that fades is recorded and left.

What changes

Heard and verified

Most market signal work measures what is being said and leaves whether anyone acted open. Sena reads the acting.

Capability areaTypical setupSena
BasisPosts, searches, and mentions.Purchases in real outlets.
What it measuresAttention.Units taken.
PopulationConsumers who publish.Consumers who buy.
PlacementA platform.A named country.
PersistenceJudged on one spike.Compared across cycles.
VocabularyStrong and early.Paired with a unit figure.
Evidence in a reviewA mention chart.Photographs, dated and placed.
Use cases

Where the tell decides

Three early questions a listening stack answers poorly.

See it on one category

Test one watch list live

The walkthrough takes a monitoring list the team already keeps, checks each entry for purchase evidence across the captured markets, places the ones that hold, and shows which faded between the last two cycles.

What a walkthrough covers

  1. 01Entries carrying unit evidence
  2. 02The country each read came from
  3. 03Movement across two capture cycles
  4. 04Reads that faded, and when

Book a demo

Bring the watch list and the categories on it.

FAQ

Market signal questions

01 What are market signals?
Early, ambiguous indications that consumer demand is moving. They sit before a change has a name and before its size is settled, which is what separates them from trends. A signal may resolve into a trend or fade entirely, and the value of catching one lies in the time between it appearing and it becoming common knowledge.
02 What is market signal intelligence?
The work of catching early indications, placing them in a market, dating them, and judging which are real. Conventionally the indications come from social posts, search volume, and press mentions. They can also come from purchase, where the change has already been paid for by somebody and the read carries a unit figure from the start.
03 How do market signals from social media work?
Listening stacks count mentions, sentiment, and reach across platforms, surfacing categories where discussion is rising. They are strong on vocabulary, since a new word for a format shows up in posts early. They are weak on scale and placement, because posting volume tracks engagement rather than purchase, and a platform describes an audience rather than a country.
04 What is the difference between a signal and a trend?
A signal is a read that might resolve either way. A trend is a signal that resolved. The distinction is operational: acting on a signal as though it were a trend funds builds against movements that turn out to be fluctuations, and waiting for certainty means acting once the advantage has gone to whoever moved earlier.
05 Why is purchase a stronger signal than discussion?
Because paying is a commitment and posting is an expression. The populations overlap and differ, and the people most likely to publish about a category are among its least typical buyers. A small movement in units taken carries more information about where a market is going than a large movement in mentions of it.
06 How is a market signal placed geographically?
By reading it where it happens. A purchase occurs in an outlet in a country, so the read arrives with a location attached. Online indications arrive attached to a platform, and converting that into a market requires assumptions about who the audience was.
07 What stays uncertain even with purchase evidence?
The outcome. A signal is ambiguous by definition, and reading it from purchase improves the evidence while leaving the ambiguity in place. Some reads that carry units still fade. What the purchase basis buys is a better-ordered watch list, so scarce attention goes to the movements most likely to hold.
08 When does a market signal become a trend?
When it persists across capture cycles and reaches a size too large for the category to absorb as noise. That threshold is what early trend detection exists to set and apply. Until it is crossed, the read stays on a watch list, and the decision to cross it is a judgment made against evidence rather than a formula.
09 How often should signals be captured?
More often than trends, since the whole purpose is catching a movement while it is small. Monthly capture suits most consumer categories, because the interval sets the smallest change that can register. Capture spaced too widely converts an early read into a late one, at which point the page next door is the right one to be reading.