{{title}}
{{textarea}}
{{task}}
{{select}}
{{team}}
{{input}}
{{solution}}
{{result}}
ICDS Insights Consumer Behaviour
The Same Media Peak, Two Products
A media peak changes search for a face cream and for a snack in different ways, and only in one of them does the split of opinion matter.
VM
Vadim Matyushkin Behavioral Scientist & Sociologist · ICDS
A brand gets a surge of mentions. Search interest in the brand rises. It is easy to assume the same mechanism works for a chocolate bar, a face cream and a bottle of water, and that a more positive conversation always helps. The ICDS data separate these products into groups that respond to the same kind of event in different ways.
Balance of Expression predicts the search response only in categories where buying carries risk. For a snack or a drink, the presence of a peak is enough, and the balance of positive and negative statements changes nothing measurable.
66FMCG brand peaks10 categories · 26 countries
ρ = +0.75Cautious: balance → searchAt the peak; +0.69 a month later (p < 0.01)
ρ = −0.16Spontaneous: no effectBalance structure predicts nothing durable
ρ = +0.10Effect is the brand’sA peak lifts the brand, not the category (p = 0.41)
Three Types of Purchase, Three Shapes of Response
Spontaneous categories (snacks, beverages, food products) respond at once and lose the effect within a month. Routine categories (alcoholic beverages, household chemicals, part of dairy, baby care) retain the effect weakly. Cautious categories (cosmetics, premium confectionery, part of dairy) keep gaining search after the peak has passed.
| Category type | Brands | Search change, t+1 − t0 | Brands with t+1 > t0 | Balance vs. search at t0 |
|---|---|---|---|---|
| Spontaneous | 26 | −22.8 points | 46% | ρ = −0.16, n.s. |
| Routine | 22 | +2.3 points | 55% | ρ = +0.10, n.s. |
| Cautious | 18 | +9.4 points | 72% | ρ = +0.75, p < 0.001 |
The Category Sets the Type, the Split Sets Its Strength
In Cautious categories the Balance of Expression correlates with search both at the peak month (ρ = +0.75) and a month later (ρ = +0.69, p < 0.01). Search keeps growing after the peak whatever the balance. The category decides that the response is delayed and durable. The balance decides how strong it is. In Spontaneous categories no balance structure produced a durable effect.
We read this through the Elaboration Likelihood Model. A buyer facing a costly or socially visible choice reads brand information carefully, and a field in which positive voices clearly dominate while a minority objection remains looks like a case that has been weighed. A buyer of a snack responds to the signal and moves on.
The Effect Belongs to the Brand
Brand and category search indices were almost uncorrelated (ρ = +0.10, p = 0.41). A peak did not lift the whole category. It produced a search impulse for the brand that had the peak. In Cautious categories brands also gained share more often than they lost it against their main competitor (28% against 11%), and this tendency followed the Balance of Expression (ρ = +0.59, p = 0.010).
ρ = +0.10Brand vs. category search — a peak lifts the brand, not the category (p = 0.41)
28% vs 11%Cautious brands gained share against the main competitor more often than they lost it
ρ = +0.59Share gain followed the Balance of Expression (p = 0.010)
What the Data Moved
The original assignment put confectionery among Spontaneous categories. The behavioral profile did not fit: search for Ferrero Rocher, Ben & Jerry’s and Lindt kept growing by 28 index points from the peak month to the next. These are premium and gift-oriented products, and a gift says something about the giver’s taste and status. Confectionery was moved to Cautious. Food products moved the other way, to Spontaneous, because their search fell by 38.6 points after the peak. Perceived risk turned out to belong to the brand’s positioning inside a category, not to the category’s name.
What Follows for Practice
1
Check the category type first.
Raising positive tone gives a measurable search gain only where the purchase is cautious. Elsewhere the first task is to have a peak.
2
In cautious categories, aim for clear dominance, not unanimity.
Article 6 reads a 75/25 split as the useful range: clear dominance with a visible minority view. A field with no objections at all gives a buyer less to weigh.
3
Read search after the peak.
In Cautious categories the effect continues into the following month, so a measurement at the peak month alone understates it.
What Is Established, and What Is Not
Established
The demarcation between Cautious categories and the other two, on 66 peaks. Hypotheses H2 and H3 from Article 5 are confirmed. Hypothesis H1, that very strong positive dominance is optimal for Spontaneous categories, was not confirmed.
Established as a signal
The Cautious group has 18 observations. That is enough for a first analysis and not enough for reliable testing of interactions. The corpus needs to grow before definitive publication.
Method limits
Search data is an aggregate index without the split into informational and transactional queries. The design is observational: brands with loyal audiences may produce more positive peaks to begin with. Google Trends undercounts in markets where search runs through other channels.
Methodology note. Balance of Expression is the ratio of positive to negative evaluative mentions within the expressive part of a peak. Categories were checked by three diagnostics: change in the normalized search index from t0 to t+1, Spearman correlation between tonality and search response, and the share of brands with t+1 above t0. Normalized index = brand index minus category index. Institute of Communication and Data Science. Working draft for internal discussion.
VM
About the author
Behavioral Scientist & Sociologist · ICDS
Vadim Matyushkin is a psychologist and sociologist with close to twenty years of research into digital behavior, trust, and information dynamics, and a researcher behind the Institute of Communication and Data Science (ICDS). His work has supported organizations including Coca-Cola, PepsiCo, Mars, Danone, Nestlé, and Bayer in moving from self-reported survey data toward direct behavioral evidence.
ICDS — Institute of Communication and Data Science is an independent research institute focused on understanding how trust, behavior, and reputation are formed in digital environments. ICDS operates as an intellectual institution — not an agency, not a platform, and not an educational provider.