Implication vs Observation
An observation is what the data says; an implication is what it means for the decision - and interviewers reward the second, not the first.
What it is
An observation is a neutral statement of what you see in the data: 'revenue fell 12% last year'. An implication is what that fact means for the problem at hand: 'so the decline is steep enough that a cost-cutting response alone will not restore profit - we need to address the revenue driver directly'. The difference is the move from noticing to reasoning. A chart or a data point on its own is inert; its value comes entirely from the 'so what' you attach to it. Strong case performance is a chain of implications, not a recitation of observations - every fact you surface should be immediately connected to what it means for the recommendation.
When to use it
The distinction applies every time you are handed data, an exhibit, or a fact - which in a case is constantly. Whenever you read a chart or a number aloud, the discipline is to follow it with its implication rather than moving on. It is especially critical when an interviewer slides an exhibit across and asks 'what do you see?' - the weak candidate describes, the strong one interprets. There is no real exception: even confirming a fact benefits from stating why it matters. Do not narrate data without a 'so what'; observations without implications signal that you can read but not reason.
How to apply it
Every time you state a fact, immediately ask 'so what does that mean for our question?' and say the answer in the same breath. Build the habit of a two-part sentence: the observation, then the implication - 'X is true, which means Y for the decision'. When reading an exhibit, resist listing everything on it; pick the one or two data points that move the answer and interpret those. Chain implications toward the recommendation: each 'so what' should narrow the problem or point at an action. If you find yourself describing without concluding, stop and force the implication before continuing. The observation earns its place only by leading somewhere.
Common mistakes
The classic error is data narration - reading an exhibit line by line ('sales are 100 here, 120 here, the blue bar is bigger') without ever saying what it implies, which fills time but adds no reasoning. Another is drawing an implication the observation does not support - over-reaching from a single data point. People also observe the obvious and miss the decision-relevant point on the same chart, or list five observations when one carries the insight. And they separate observation from implication in time, noting facts now and interpreting 'later', so the connection never gets made.
Example
An interviewer presents a chart showing a retailer's online sales growing fast while total sales are flat. A pure-observation response: 'Online is growing and total sales are flat.' True, and useless. The implication response: 'Online is growing fast while total is flat, which means online growth is cannibalising the stores rather than adding new demand - so the digital push is shifting the channel mix, not expanding the business, and the real question becomes whether online carries enough margin to justify the store decline it is causing.' Same chart, but the second reading extracts the decision - the observation was only the doorway to it.
Practice it
Practice on data-heavy and exhibit-interpretation cases by pairing every fact you state with its 'so what', selecting only the decision-relevant points on a chart, and chaining implications toward the recommendation. The skill to build is the reflex to interpret rather than describe - never leaving an observation stranded without the meaning it points to.