Demand intelligence
Visirank® Demand
Search intelligence, delivered since 2015.
We have been reading demand signals since 2015 — long before anyone thought to sell it as a category.
Visirank Demand is the ARvis.it service that measures market demand through search volumes. For every topic we rebuild twenty-four months of historical series across hundreds of queries and aggregate them into semantic clusters. The result tells you what a market will be looking for, months ahead of the season — which is when assortment, promotions and campaigns are actually decided.
The track record
Five years of monthly reports, on a contract.
This is not a method assembled around a demo. It ran as a continuous service, on a fixed monthly deadline, for a client who used it to make decisions.
The service was called sector studies, then Sector Reports, then Digital Trends. Visirank Demand is the fourth generation, and the first with a brand on it.
The method
From search demand trends to the cycle of a category.
The work does not start from a keyword. It starts from a commercial topic — a category, a department, a moment of consumption — and from the demand that surrounds it.
- The queriesHundreds of queries per topic: all the different ways people ask for the same thing, including the phrasings nobody inside the company would use.
- The historical seriesTwenty-four months of monthly volumes for each query. Two years is the minimum that lets you tell a cycle from an episode.
- The clusterQueries are grouped into semantic clusters: sets that move together because they describe the same need. The cluster, not the keyword, is the unit of demand.
One keyword on its own is noise. A cluster with two years behind it is a cycle.
The proof
Spanish ham, read in August.
In a study on cured meats delivered for the Christmas season, one family of queries behaved unlike all the others. These are the monthly volumes we measured.
| Query | Oct | Nov | Dec |
|---|---|---|---|
| prosciutto spagnolo | 880 | 6,600 | 4,400 |
| prosciutto serrano | 210 | 390 | 880 |
| prosciutto patanegra | 320 | 480 | 720 |
The point is not seasonality. It is scouting.
A stable cycle that repeats every year is useful, but it is not a discovery — everyone in the category already knows it. Something else was happening here. The Christmas peak for Spanish ham was 1,600 in 2017 and 6,600 the following year. Not a cycle: a trend being born.
That jump only becomes visible in November, by which time the season has started and the assortment is locked. Reading it in August, inside a cluster, means arriving in December having decided rather than having reacted.
The counter-proof
Seasonality: three behaviors, one instrument.
A method that finds seasonality everywhere is not measuring, it is confirming. In the same study, three closely related families of queries behave in three different ways.
| Query | Behavior | Volumes |
|---|---|---|
| prosciutto crudo | flat all year, no peak | 4,400–6,600 |
| prosciutto e melone | inverted, summer seasonality | 3,600 in June · 70 in November |
| prosciutto spagnolo | Christmas peak, and growing | 880 in October · 6,600 in November |
Italian cured ham is baseline demand: high, constant, indifferent to the calendar. Ham and melon is an entire summer compressed into two months, and by November it barely exists. Spanish ham is a Christmas product, and on top of that it is growing.
When a method can recognize what is not seasonal, what it says about seasonality becomes credible.
What changed
What used to take a month now takes a day.
Building the clusters, aligning the series, isolating the anomalies: for years that was days of spreadsheet work. It is now a fraction of the time, which changes two things — more topics can be covered, and the data can be revisited when a question changes instead of once a month.
What has not changed is the part that matters. Every category throws off dozens of swings: holidays, weather, passing fads, statistical noise. Telling the one that is becoming a trend from the one that will fade is a judgment, and it is built on knowing the client's market. That is what Aigmented® means at ARvis: the instrument accelerates, the decision stays human.
The data was never the hard part. The decision was.
Who it is for
Anyone who has to decide before the season starts.
- Retail and groceryAssortment planning, promotional calendar, the themes that go in the flyer.
- Branded manufacturersCategory demand read upstream of sell-in, as an argument to take to the trade.
- E-commerceCategory seasonality and demand signals applied to range, campaigns and stock.
- Publishing and content planningEditorial plans built on when a topic is searched for, not on when it suits us to publish it.
Frequently asked questions
Frequently asked questions about Visirank Demand
- How long have you been doing this?
- Since 2015. The service ran monthly for Iper La Grande i, one of Italy's major grocery retailers, from October 2015 through 2020, and the current methodology comes out of those five years. Visirank Demand is the fourth generation of that service and the first one to carry a brand.
- What is search intelligence, exactly?
- It is the reading of public search volumes as a measure of demand. For each commercial topic we rebuild twenty-four months of monthly volumes across hundreds of queries and aggregate them into semantic clusters. The output is not a keyword list: it is how a category behaves over time, and what that implies for the season ahead.
- How is this different from Google Trends or a keyword research tool?
- A tool gives you volumes. The service gives you the clusters, the historical series and the reading — which swing is turning into a trend and which one will fade. That judgment is the work, and it arrives months before the season rather than after it.
- How far ahead do the findings arrive?
- Three months by contract, often more. The study written for the 2019 Christmas season is dated August 2, 2019: enough time to revisit assortment, promotions and communication before the season begins.
- What data is it based on?
- Public search volumes, rebuilt into monthly series over twenty-four months for hundreds of queries and aggregated by semantic cluster. No personal data, and none of the client's sales data: the method measures what a market is looking for, not what it has already bought.
Once demand becomes a contact with a name and a company behind it, the other line measures it: explore Visirank Lead →