RankAnalyst Lab

From RankAnalyst Lab
Jump to navigation Jump to search
RankAnalyst Lab.png
Category Big Data Dependency Exploration, SEO, User Analysis, Website-Optimization
Development Steven Broschart, Rainer Monschein
Year of development 2018 - today
Latest version 0.9
Operating System platform-independent

What is RankAnalyst Lab?

RankAnalyst Lab (short: RAL) is a Dependency Exploration Tool (Big Data, Search). It was developed for the visualization and analysis of optimization-specific correlations. This makes vertical analysis (structural analysis of a domain) and horizontal analysis (cross-domain comparisons) possible. By using parallel coordinates, multidimensional dependencies (e.g. the interaction of different KPIs as ranking factors) can also be investigated. In addition, it has many functions for regular reporting and presentation of research results.

The system can handle interaction and placement data from any data source. These must be merged into a common CSV file (comma-separated) via KNIME and a Workflow prepared for this purpose.

RAL is currently not available as free software, but is used as a specialized tool within the scope of consulting services. However, projects that have been created can be activated publicly, so that they can also be used without an account.

Difference to classic SEO tools

Compared to classical SEO tools, RAL differs in the following points:

Data processing

RAL keeps all data in the RAM of the computer. The records are only retrieved when the interface is completely reloaded. This allows you to work very quickly and perform complex requests that would not be possible with a server-based architecture. Any data source can be used and linked. This data is uploaded to the system as a CSV file and can be used immediately.

Unlike other systems, the system does not process live data from its own inventory.

Keyword vs. context

The system enables the mapping of contexts. Contexts are semantically connected rankings that are defined using one or more key terms. If a context is defined e.g. via "sports shoes|running shoes", RAL takes into account all rankings that include "sports shoes" or "running shoes" in their wording. This architecture also allows contextual movements to be tracked over time. In addition, the performance of domains whose product range only overlaps in parts can also be compared in this way.

Parallel Coordinates

A visually essential component of the system consists of the interactive streamline diagram. This shows the development of filtered rankings and contexts over time. More parallel axes can be docked to this display in order to examine selected parameters, such as bounce rate or loading time, for their effect on ranking performance.

Onescreen-Interface & Filtering Options

RAL is designed as an onescreen interface. The system supports complex filtering, which is managed via different input areas and mutually influences each other. With a multiscreen interface, the overview would no longer be guaranteed in an adequate form.

KPI Orientation

Most SEO tools present their key performance indicators on an economic background or for the quantitative evaluation of a development. Of course, comparable indicators can also be found in RAL, but the focus is on root cause analysis. The primary objective is to uncover the underlying market and rating mechanisms. The knowledge gained in this way can then be used to initiate suitable measures.

Swarm Analysis

Search engines use specific rules when evaluating websites. By using different algorithm components and the resulting interactions, each website seems to be evaluated according to very individual parameters. With RAL, the application of individual control patterns can be discovered and analysed. In this way it becomes clear which criteria are important for strategic planning; which approach "works" and which does not. We get a clear picture of how the search engine "understands" the website to be optimized. RAL thus allows a very targeted approach to the optimisation of a website.

An essential component is the swarm analysis. It examines the movements of related search phrases over time. We define this bond by one or more keywords that appear in each search phrase of such a group of search phrases. We also speak of a swarm in such a group. In a broader sense, a swarm describes a context (topic), which we can track over time and compare with competing domains.

The swarm analysis takes place over different ranges. The most noticeable area is probably the interactive streamline diagram, which makes up a large part of the interface.


The raIndex (RankAnalyst-Index) is available for a quick and easy evaluation of whole or selected trends. In its definition, particular emphasis was placed on a comprehensible and transparent calculation. The raIndex is a product of the average ranking and the total number of identyfied rankings:

  • raIndex = (101-ØRank) * #Rankings

Since a good ranking must lead to a higher raIndex, we reverse the scale by subtracting 101. There is no weighting according to search result page, demand or competition. An incomprehensible impairment is thus impossible; an evaluation of the actual state, but also of the quantitative development (in interaction with the filtering possibilities of RAL also qualitative) is easily possible.

More Information