How data analytics startups make money

Current start-up dates

Start-ups are firmly connected with attributes such as flexibility, personnel efficiency and flat hierarchies and are not only part of the business world, but also often serve as role models for this reason. It is not for nothing that they are regarded as the modern counterpart to global corporations, because they enter the market “fully digitized”. Targeted data analysis therefore opens up new opportunities, especially for start-ups - be it in performance marketing or in customer support operations. The everyday business world has long been a data world.

It is not for nothing that, according to a Bitkom survey, three out of four companies with more than 20 employees see the use of big data as one of the key technologies that determine competition, with six out of ten companies already using this technology or planning or discussing its use. Start-ups are also well advised to conduct data analysis professionally right from the start. This article shows how successful data analysis can be achieved in start-ups.

Data Analytics: Insights through data

Data is created from day 1 and also gives start-ups every opportunity to optimize. Young companies should therefore have data analysis in mind right from the start. Specifically, this means: Make sure that you save all data in such a way that they can be found afterwards. Because you do not know today which analysis questions may arise tomorrow. The data scientist David Kriesel, who has achieved national fame with his data analysis for Spiegel-Online, expresses this briefly and succinctly: "Raw data is awesome".

In order for data to be collected, structured and analyzed in a targeted manner, it is necessary that every employee understands why which data has to be collected and under which aspects it has to be evaluated. A data-driven corporate culture is definitely state of the art in the 21st century. Therefore, also defines owners who are responsible for storing the collected data in a systematic and structured manner so that nothing is lost to posterity. When saving the data, the following applies: the more, the better. For the later evaluation, the motto is: Quality is half the battle. Big data has to become smart data here.

Data-driven corporate culture: three decisive factors

1. Clear vision

The concept of the DNA of a company has established itself among start-ups: The success of a young company depends heavily on the fact that each individual employee knows and understands which goal a company is pursuing in the market, what the company's characteristics are, which ones Corporate culture is required and promoted. The founder (s) and managing directors of a young company must also formulate a clear vision separately, but especially with regard to data analysis, and then communicate this clearly and comprehensibly to all employees.

2. Holistic planning

In the second step, the analytical challenges are derived from the formulation of objectives. What is my challenge and how do I want to solve it analytically? One question could be, for example: Which marketing message do I send to which (potential) customers in order to increase sales? The specific analytical challenges in turn result in the methods and tools that should be used to solve the analytical tasks.

But young companies shouldn't ignore planning either. In order to establish a successful analysis culture in the company, a target / actual analysis with a clear formulation of the success criteria is useful. The experts from The Information Lab, for example, provide valuable support. The inventory is carried out using a comprehensive catalog of questions with internal and external assessments of the use and analysis of data, which is followed by a formulation of objectives. Based on this, concrete measures are defined that enable the company to draw new conclusions from existing data more quickly and efficiently. This also includes the decision about the software required: Is Excel sufficient in the long term or is it not more productive to invest in comprehensive and specialized software at an early stage? Should only employees with in-depth programming knowledge be able to use the software or should a variant be chosen that does not require any programming knowledge and is therefore accessible to a much larger group of users?

A big advantage of the so-called self-service software solutions: They make it possible that relevant data can be systematically and comprehensively analyzed by employees using drag-and-drop, even outside the IT department - or even completely without it (!) can. In young companies in particular, the separation between IT and other departments is disappearing because, on the one hand, employees are expected to have more IT knowledge and they want to be more actively involved. Employees in the individual departments (e.g. Sales Management, Human Resources, Logistics, etc.) can become so-called Citizen Data Scientists within a very short time with the right offers such as Alteryx, Tableau, Qlik or Power BI and analyze and visualize data independently. The company also benefits from the specialist knowledge of its employees because it is incorporated into the analysis of the data obtained.

3. Transparency and Security

For companies, the findings from the data analysis create tangible added value. The data used should be as accurate as possible and their security guaranteed, otherwise there is a risk for companies, for example, that data will be viewed. Standardized mechanisms allow incoming data to be stored and categorized centrally. In this way, data can easily be integrated for analysis purposes and updated regularly if necessary.
Software solutions that also offer data catalogs that also make the metadata searchable are also helpful for young companies.

Practical data analysis in three steps

Collecting, structuring, analyzing, evaluating and visualizing data and finally summarizing it - this is how practical data analysis works.

1. Collect and structure

Data is everywhere, every click of a customer on a website can be proven. Such internal and external data are inserted into existing databases and software. If different data sources are used, it is important to link them to a uniform format and evaluate them in an integrated manner. The data is also checked for input errors, gaps, etc. in this step.

2. Analyze, evaluate and visualize

The following analysis can be approached from two directions: In unsupervised learning, algorithms try to recognize patterns in the existing data and draw conclusions from them. In supervised learning, on the other hand, an attempt is made to formulate a hypothesis that makes predictions that are as accurate as possible. During the analyzes, data can therefore be checked for hypotheses - such as B. "Will sales of my product increase if my company spends more money on social media campaigns?" A formula is sought that explains the data and is valid to the extent that it can also explain new data (validation sample).

Start-ups that want to approach analyzes can first start with A / B test procedures. This method is used in marketing, for example, to determine how well customers respond to certain patterns or designs and which ones have more interaction. To do this, you create two different emails, for example, which have the same content, but are provided with different texts, designs or call-to-action buttons. The analysis of the different reactions to these emails provides information on productive optimization potential.

Incidentally, the visual preparation of the data is very well suited to illustrate trends, patterns and relationships of large amounts of data. The data is made visible as a "story", can be processed quickly by the brain and gives rise to new ideas.

3. Summarize

In the last step, the knowledge gained from the data is at best summarized and communicated in writing and graphically. Again, the presentation must be clear and understandable so that the recipients can get the best possible benefit from the presented results. It is then up to the decision-makers to put the knowledge gained through data analysis into practice and to lead their company to success with data-based decisions.

In addition, it is also about embedding analytical models directly in business processes so that analyzes can be carried out in real time and deliver targeted results. In particular, start-ups that, compared to established companies, start with technologies on the green field can integrate them into the success of their company right from the start.


The technological developments of the last few years have produced very user-friendly tools with which even programmers can now analyze data quickly and easily. At the same time, larger and more complex amounts of data can be reliably examined and repetitive tasks can be automated. For companies that use data analysis in a targeted and structured manner, this means a decisive advantage that can be decisive for the war in the face of tougher global competition. Young companies in particular are therefore well advised to invest in targeted data analysis right from the start.

The author Tom Becker is General Manager Central & Eastern Europe at Alteryx, one of the leading providers of data analysis software. Becker is responsible for the company's market positioning, product localization and building up the community in the DACH region