Integrating Internal Company Geodata in GIS

What is data integration in relation to geographic information systems?

Andreas Marth: In order for companies to be able to see data such as location data, sales data and customer data on a digital map, so they can connect them with each other and thus analyze them, this data must first be integrated into a geographic information system (GIS) such as WIGeoWeb.

What is the advantage of integrating data into a GIS?

Michael Steigemann: By integrating data with a spatial reference into a GIS, the data points can be related to each other even though they have nothing to do with each other. This means that I can combine a wide variety of data and display it on a digital map so that it makes sense. By comparing the location data and the purchasing power in the catchment area of my locations, I can see at which locations I am not yet exploiting my potential.

Andreas Marth: For example, if a company makes five times as much turnover at one location as at another, you can conclude that this location performs much better. However, if the locations and turnover are placed in spatial relation to one another on a digital map, you can see that ten times as many people live in the catchment area of the location with five times more turnover. Then it suddenly becomes clear that the first impression was deceptive. In fact, you may even find that the location with less turnover is performing better than the other one because, although its potential is not as great, it is being better exploited.

When should a company start thinking about data integration?

Michael Steigemann: Anyone who uses a geomarketing tool like WIGeoWeb or WIGeoLocation will quickly reach the point where they want to integrate their own data. Without data integration, internal company data cannot be displayed and analyzed in a GIS and thus on a digital map in order to draw conclusions for the operational business. A comparison: On well-known digital maps, such as Google Maps or OpenStreetMap, you normally only see the world as Google or OSM see it, but not with your own data, such as customer addresses and turnovers.

Andreas Marth: Another example is sales territory planning. It happens very often that customers have lists of ZIP Codes that they assign to sales representatives. Only when you integrate this data into a GIS can you see on a digital map that the assignments are completely mixed up or that there are even white spots that have not been assigned at all.

Which data can be integrated into a GIS?

Andreas Marth: All geodata, i.e. data that has a spatial reference, can be integrated into a GIS. That is a lot of data, because around 80 percent of all data has a spatial reference. The spatial reference is not always recognizable at first glance. For example, a list may only contain branch numbers and sales figures next to them, but no addresses. However, this data also has a spatial reference, because every branch has a location. Thus, the turnover can be put into a spatial relationship with the respective branch location.

How is company data integrated into a GIS?

Andreas Marth: The first step is to sit down with the companies and discuss what they expect to achieve by integrating their data. Then we review their data together. This is where we usually find that companies have much more data that can be integrated than they thought. In most cases, we do the technical data integration for the customer.

Michael Steigemann: There are also customers who integrate their data themselves. However, data integration is not just a technical issue, it is also a question of know-how. You need to know which questions you can answer with which data in order to produce meaningful statements and useful results. Our USP is the consulting expertise we have gained from decades of experience in a wide range of industries.

How should the data be structured to ensure smooth data integration?

Michael Steigemann: Basically, we can integrate every common data format. Often it is data from databases or web services. It is important that the data is complete and as error-free as possible, because we do not do data cleansing. The principle “shit in, shit out” applies here. The significance of the analysis results depends largely on how good the input data is.

Andreas Marth: The quality of the input data increases the quality of every analysis you perform with it. The more structured the data is, the faster errors can be identified. It is also necessary that the data is geocoded, but we usually take care of this step for our customers.

What is geocoding in data integration?

Andreas Marth: Without geocoding, data integration is impossible. Typically, companies have address lists. In order for us to integrate the addresses into a GIS and display them on a digital map, the addresses must be provided with a geocode, i.e. an X and a Y coordinate. We have worldwide address databases that know the geocodes of almost every address. We usually do the geocoding, but it is also possible for companies to perform this step themselves using our geocoding software JoinAddress.

Michael Steigemann: Here too, it is important that the addresses do not contain any errors. The less precise the address database or address list is, the more error-prone and difficult the geocoding is and the more manual editing that is necessary.

Can data integration also be automated?

Michael Steigemann: Yes, data integration can be fully automated. In order to integrate your data, we need current data. However, customer data, location data, sales data and many other data are constantly changing. We therefore advise companies to set up their systems in such a way that the data is sent to us automatically on a regular basis. Generally, this is carried out via various exchange interfaces that we make available to our customers or that we receive from them.

When should data be aggregated or disaggregated?

Andreas Marth: Aggregating means combining smaller data points into a larger unit. We do this frequently, for example when sales are aggregated from the ZIP Code level to the state level. Although the sales figures in the original files are only available at the ZIP Code level, the data aggregation in the GIS shows how much turnover was made in which state.

Michael Steigemann: Disaggregation of data is the reverse step. Often, companies do not have data at the customer level, for example, but at the sales territory level. However, for their further planning, they need the data at the customer level. Then, for example, we can disaggregate sales of plows using zoning data down to the customer level by including how much land in that sales territory is used for agriculture. Of course, we still do not know which customers exactly bought the plows from this retailer, but we can calculate a probability that is helpful for further planning.

Can global geodata be integrated or are there geographical restrictions?

Andreas Marth: In principle, there are no geographical restrictions on our geoinformation systems.

Michael Steigemann: However, the quality of the data available from different regions of the world may vary. For example, in some regions of the world there is no reliable income data.

What creates the added value of data integration in GIS?

Andreas Marth: The special added value of data integration is created by blending and combining data. For example, to find out whether a company is exploiting its potential in its catchment area, we enrich its customer list with the number of residents in that area. On the map, we see the company is in a district with 1,200 inhabitants that fit its target group, however, only 400 of them are already customers.

Can the data also be automatically imported back into my systems?

Michael Steigemann: Yes, the data can be imported back from a GIS into your company's system in the exact form your company needs.