Trend Analysis
Separating durable structural shifts from noise, with leading indicators you can actually monitor.
Most trend decks are vibes with stock photography. We treat trends as hypotheses about structural change and test them: time-series decomposition to separate signal from seasonality and noise, diffusion analysis to locate a trend on its adoption curve, and leading indicators, patents, hiring, funding flows, regulatory dockets, to date it.
Who this is for
Strategy and investment teams deciding when (not just whether) a shift deserves capital.
How the work is done
Formulate the hypothesis
Each candidate trend restated as a falsifiable claim about behavior or economics, what would be true, by when, visible in what data, because a trend you cannot falsify is a slogan.
Test against the series
Time-series decomposition (trend, seasonal, irregular components) and structural-break tests on the relevant series distinguish genuine inflections from cyclical recovery and reporting artifacts.
Locate on the curve
Diffusion analysis places the shift on its S-curve using adoption data and analogs; cross-impact analysis maps how the trend interacts with adjacent forces, since second-order effects are where the money usually is.
Build the watchlist
A scenario cone (plausible trajectories, not one prediction) with dated signposts and thresholds: the specific observations that should upgrade or downgrade your conviction, reviewed on a set cadence.
How the Trend Analysis engagement runs
We begin with the decision, use the evidence that can genuinely change it, and make the reasoning reviewable from first input to final handover.
What we need to begin
- The trends the client believes are operating, stated plainly enough to be proven wrong.
- Time series long enough to decompose: ideally five years or more at monthly or quarterly frequency.
- The decision horizon, since a two-year and a ten-year view call for different methods.
- Candidate leading indicators the client already watches, and why they trust them.
If an input is unavailable, we state the gap, its effect on confidence, and the agreed workaround. It is never quietly ignored.
Your four-phase engagement map
-
Phase 1
Formulate the hypothesis
Restate each candidate trend as a falsifiable claim: what would be true, by when, visible in what data.
-
Phase 2
Test against the series
Time-series decomposition and structural-break tests distinguish genuine inflections from cyclical recovery and artifacts.
-
Phase 3
Locate on the curve
Diffusion analysis places the shift on its S-curve; cross-impact analysis maps interactions with adjacent forces.
-
Phase 4
Build the watchlist
A scenario cone with dated signposts and thresholds that upgrade or downgrade conviction on a set cadence.
Methods and models we draw on
- Falsifiable trend hypotheses
- Time-series decomposition
- Structural-break testing
- Diffusion & S-curve placement
- Cross-impact analysis
- Leading-indicator design
- Scenario cones with signposts
Methods are chosen for the problem, not the brochure, expect a subset of these, applied properly, plus whatever the evidence demands.
The decision this enables
Conviction (or scepticism) with a timestamp, and the indicator list that updates it automatically.