Location Intelligence: the new way of looking at maps

Sergio Vázquez

23 October 2016

It’s held that the world’s first known map is a small clay record sheet from Babylon that is no more than 12 centimeters long. Created around 500 A.D., we can currently see it at the British Museum in London. The truth is that it is possible that the Babylonians were already making maps since 2300 B.C. with the intention of measuring distances of their territories for tax collection.

The world’s first-known map

A map is a representation of reality that gives us the ability to obtain information and apply intelligence to the information represented in their respective locations.

The first case of successful location intelligence happened in 1854 when England’s most violent cholera outbreak struck with 700 people dying within a week of each other in London’s Soho neighborhood.

John Snow, the English physician, considered to be the founding father of modern epidemiology, had a passion for maps and used them to support many of his articles and exhibits. His practice was very close to the heart of the epidemic, and with this knowledge of the vicinity and the victims, he decided to buy a map and record the location of all the deaths during the month of September. The result was illuminating: the greatest concentration of deaths occurred around Broad Street.

John Snow had carried out an excellent exercise in making the correlation between frequency and concentration, but it didn’t just stop there; he also geo-referenced all the neighborhood’s wells and could observe that the majority of the victims were near the well on Broad Street.

Thanks to his representation, they successfully located the cause of the epidemic and avoided its spread. They gained intelligence from geolocated information. Coincidence? There’s a term for this: location intelligence.

snow cholera map

John Snow’s original map where he marked the location of all cholera victims and all wells with an X.

Location Intelligence

Location intelligence doesn’t only involve analyzing the information on the map all by itself, but it also includes analyzing specific data to identify relationships and gain critical insights to solve specific problems.

More than 80% of the data contains a location component that directly impacts the decisions that can be made about them. Any company has locations for their customers, stores, offices, devices, fleets of vehicles, and even their competition.

Thanks to the introduction of Big Data, we can nurture our spatial searches with a plethora of external information. Information we can derive from public or private sources, and even we have the ability to add it on to our geo-referenced information in real-time.

  • Picture this: if we add external information sources to the location of our clients in a geographical area like census figures, demographic data, statistics, or other public information, we’ll be able to get an unprecedented greater degree of detail. This way we’ll be able to answer questions like: How old are they? What’s their socioeconomic level? In what areas are they primarily location? All these answers will count on a location component, and we will be able to respond to these queries through location intelligence efficiently.

This information about spatial relationships produced from these pieces of information grants us a greater understanding of behaviors, influences, or trends presented but could be hidden within a labyrinth of tables and figures.

Using technology from Geographic Information Systems (GIS), we will be able to understand and obtain value about location intelligence. With a GIS, we will be able to map geographic elements derived from our company’s pieces of information to find patterns and relationships that would otherwise be impossible. Besides that, we will be able to add on information from a plethora of external sources. All of this comes together in a blend of various data layers from which we derive substantially valuable insights.

And, what if what we represented on a map were neither internal nor external sources containing statistical and demographic data, but instead conversations happening across digital media platforms?
On a mentions map, we can represent in what areas people are talking about a particular topic, product, brand, hashtag, event, etc. For example, we will successfully be able to visualize those conversations that have an adverse impact on my company. We will also be able to detect in what areas there is the greatest volume of conversation around my brand, and therefore, conduct marketing campaigns about them or even controlling the reach a campaign has had and see the consequences across digital media platforms.

Digital Universe

Next, we will see how we can represent conversations that happened on different digital platforms about five soccer teams from Spain’s La Liga.

To do so, we have chosen the tool that I consider to be the best for geo-located data analysis and visualization: CARTO (previously CartoDB).
To start, we extracted mentions on social media, blogs, forums, and the official platforms from the five teams using Brandwatch. The teams analyzed are Real Madrid, Fútbol Club Barcelona, Atlético de Madrid, Real Sociedad, and Alavés.

The data was filtered for mentions only in Spanish and English in August 2016.
The month of August coincides with half of the preseason and the beginning of the 2016/2017 season, including the first two match days of the season on August 20/21 and August 27/28.

The teams were chosen based on results from Brandwatch’s reputation report about Spanish soccer teams that looked into the following metrics: social visibility, general visibility, overall sentiment, growth in reach, engagement, and content.

Now we already have the basis to start our example (and with what pieces of information).

More than 1 million mentions about the five clubs, making the importing and filtering of this data, a job taking many hours, not to mention we only had 10% of the total existing mentions for Real Madrid and FC Barcelona.

The next step is to determine the location of the person or media format that made a comment. If we lack coordinates (longitude and latitude), we can use his or her city, country, or municipality and with GoogleMaps’s API apply a geo-referencing process that gives us the physical coordinates from one of these pieces of information.
With a perfectly created dataset (mention + coordinates) we only needed to put it all into CARTO and visualize an interactive map that will show the real conversation volume for each soccer club through the month of August of 2016.

Go ahead! You can play with the map, zoom in, activate/deactivate the temporary bar, etc.

As you can see on the map, the mentions for every team get represented using a different color.

At any point, with help from the timescale that appears in the left-hand margin, we can stop the visualization on a specific day, or even navigate the map and get closer to our city or town. Exciting, right?

We can extract fascinating information for every day, and to exemplify just how we are going to highlight a few specific days and analyze what we see on the map with what happened in the soccer world:

August 3rd
Friendly: Leicester City – F.C. Barcelona (2-4)


We see lots of chatter about FC Barcelona practically every in the world. But let us continue our analysis:

August 4th
Friendly: Real Madrid – Bayern Múnich (1-0)


The map changes color because the next day all eyes are on Real Madrid.

August 9th
European Supercup Final: Real Madrid – Sevilla (3-2)

9 agosto

We can tell that it is a final and not a friendly, lots of chatter about Real Madrid is noteworthy worldwide and significantly greater than there was on August 4th.

August 21st
La Liga: Real Sociedad – Real Madrid (0-3)

La Liga: Atlético de Madrid – Alavés (1-1)

In this case, we can observe the interaction among several teams, since this was the first day of the season and a lot of teams were playing that day (with Barcelona- Betis (6-2) included), this is perfectly reasonable.
What the results show is that soccer knows no geographical barriers and the conversations about these five football teams get extended to all ends of the earth.
We can gain very clear insights from this analysis, that we probably already knew, like:

  • The greatest volume of conversation occurred on the same day that a particular team played, and diluted as days went by.
  • The greatest and most globally-extended conversation volume occurred on the days of the most important games.
  • On those days when several teams played, there is an observed interaction between all of them without one club clearly dominating above the rest.

We have seen how we can make highly valuable maps through information coming from a social media monitoring tool which helped us not only rely on already-geo-located information, but we could also gain a supposedly “flat” type of information and then represent it later on a map. By only letting out year creative soul and think about what kind of information you can get to solve a problem that you would not have been able to solve using traditional techniques.

We at Good Rebels have been using this type of technology for quite some time and have been increasingly applying it more in different projects that see the need for the introduction of new kinds of analysis, which we will surely keep showing you on the blog.

And this is only the beginning of location intelligence software. The appearance of new techniques, software, and hardware, will make us live this type of analysis in another dimension, whereby gaining, even more, information, greater interaction, more access to data, and all of that will undoubtedly result in decreased processing time.