Data analytics without a data science team

Data is the raw material of the digital economy. Yet while large corporations have long employed specialized data science teams, small and midsize companies face a seemingly insurmountable hurdle: How can data-driven decisions be made when neither the budget nor the personnel for a dedicated analytics department exist? The answer is encouraging—and more realistic than many business owners suspect. According to the AI Study 2025 by maximal.digital, 86 percent of midsize companies recognize the relevance of artificial intelligence and data-based methods, yet only 23 percent have successfully implemented concrete projects. This gap is not a technological barrier but a question of the right strategy, the right tools, and a pragmatic entry point. This article shows you how to anchor data analytics in the Mittelstand (privately held midsize companies)—without having to build a single data science team.

Why many midsize companies are falling behind in data analytics

Germany’s Mittelstand, comprising over 99 percent of all companies, forms the backbone of the economy. Yet when it comes to systematic data utilization, it lags significantly behind. A current IDC study confirms that only one in four midsize companies uses automated data analysis to develop new business ideas. A mere 17 percent of surveyed companies fully exploit the potential of their existing data by their own assessment.

The reasons are multifaceted and concern not only technology but the entire corporate structure.

Resource shortages and the skills gap

The most obvious bottleneck is personnel. While a large enterprise can employ five, ten, or more data scientists, midsize companies often lack even a single person with in-depth analytics expertise. The Deloitte study on AI in the Mittelstand identifies competency shortages at 65 percent as the biggest hurdle, followed by implementation barriers at 52 percent and data problems at likewise 52 percent. Data scientists are among the most sought-after professionals on the job market. Their salaries far exceed the budget of many midsize companies.

But the skills shortage is only half the story. Often what is missing is not the ability to conduct analyses but the fundamental understanding of which questions can and should be answered with data. So-called data literacy—the ability to read, interpret, and critically question data—is underdeveloped in many business departments.

Data silos and quality problems

The second structural problem lies in the data itself. The AI Study 2025 delivers alarming figures: 76 percent of midsize companies struggle with inadequate data quality and data silos, and 83 percent have not formulated a “comprehensive data strategy.” Sales, accounting, marketing, and production work with different systems and formats. Excel spreadsheets, CRM systems, ERP databases, and manually captured information exist side by side without being linked.

A study by S&P Global in collaboration with AWS shows the European perspective: Only 18 percent of European midsize companies describe themselves as strongly data-driven. 54 percent see themselves as predominantly data-driven, and 28 percent classify themselves as not very data-driven. Compared to North America, the AI maturity of European midsize companies is three times lower—23 percent versus 69 percent in an advanced phase.

Cultural barriers: from gut feeling to data-based decisions

The third and often underestimated hurdle is cultural in nature. In many midsize companies, the experience-based knowledge of management dominates. Decisions are made based on years of industry experience and personal assessment—the so-called gut feeling. This experience-based knowledge is valuable, but it has limits: It does not scale, it is not transferable, and it fails when markets change faster than personal experience can keep up.

According to bofest consult, 67 percent of companies report employee reservations toward data-based approaches. The concern of being replaced or monitored through numbers is real and must be actively addressed. Data-driven decision-making should not replace experience-based knowledge but complement and validate it.

Self-service analytics: the key for midsize companies

The good news: The market for self-service analytics is growing rapidly, making professional data analysis accessible even without a data science team. According to Research and Markets, the global market for self-service analytics will grow from $6.17 billion in 2025 to $16.61 billion by 2031—an annual growth of nearly 18 percent. This boom directly benefits midsize companies, as the new tools are specifically designed for business users without programming skills.

What self-service analytics concretely means

Self-service analytics describes platforms and tools with which business departments can independently analyze and visualize data—without dependence on the IT department or external specialists. These platforms connect directly to databases, spreadsheets, cloud systems, or SaaS applications, clean and standardize datasets in the background, and present results through dashboards, visual filters, and natural language search functions.

The decisive paradigm shift: Early BI systems required specialized analysts who created reports and passed them on to business departments. Self-service analytics reverses this model. The business department asks its questions itself, explores data interactively, and gains insights in real time. The IT department provides infrastructure and governance, but the analysis itself is in the hands of the employees who know the business best.

Leading self-service BI tools compared

Choosing the right tool is one of the most important strategic decisions in the analytics domain for midsize companies. The following table compares the most relevant platforms by the criteria that matter most for the Mittelstand.

• Criterion: Entry cost per user per month | Microsoft Power BI: approx. $14 (Pro license) | Tableau: approx. $75 (Creator license) | Zoho Analytics: from €24 (Basic) | Google Looker Studio: free (basic version)
• Criterion: Learning curve | Microsoft Power BI: Low (Excel-like) | Tableau: Medium to high | Zoho Analytics: Low | Google Looker Studio: Low to medium
• Criterion: AI integration | Microsoft Power BI: Copilot, natural language queries | Tableau: Einstein AI, Ask Data | Zoho Analytics: Zia AI assistant | Google Looker Studio: Gemini AI (in development)
• Criterion: Best suited for | Microsoft Power BI: Microsoft ecosystem, Excel-savvy teams | Tableau: Data visualization, analytical depth | Zoho Analytics: Small to medium enterprises, all-in-one | Google Looker Studio: Google ecosystem, marketing analytics
• Criterion: Data source connectivity | Microsoft Power BI: Over 150 native connectors | Tableau: Over 100 native connectors | Zoho Analytics: Over 250 native connectors | Google Looker Studio: Google services, limited external
• Criterion: Mobile usage | Microsoft Power BI: Full-featured app | Tableau: Full-featured app | Zoho Analytics: Full-featured app | Google Looker Studio: Limited
• Criterion: Governance and security | Microsoft Power BI: Enterprise level | Tableau: Enterprise level | Zoho Analytics: Geared to midsize companies | Google Looker Studio: Basic

For Germany’s Mittelstand, a clear picture emerges: Companies already working with Microsoft 365 find the most natural entry point in Power BI. The Pro license at $14 per user per month is economically attractive, the interface follows familiar Excel patterns, and integration with Teams, SharePoint, and Azure works out of the box. A study by BetterBuys underscores the benefit: Companies that use BI tools make their decisions five times faster than companies without such systems.

Zoho Analytics deserves special mention for midsize companies. The platform was recognized in the Gartner Magic Quadrant 2025 for Analytics and Business Intelligence for the fourth consecutive year. 91 percent of surveyed users recommend the solution, and 79 percent report high satisfaction. The integrated AI assistant Zia enables natural language queries—a function that significantly simplifies access to data analysis for employees without a technical background.

AI-powered analytics: the next evolutionary step

The integration of artificial intelligence and machine learning is fundamentally transforming the self-service analytics market. Natural language processing and generative AI enable business users to automate complex queries and data preparations. According to market analyses, automated report generation is already 60 percent faster than manual methods and drastically reduces the effort required for analyses.

The next step is called Agentic BI: Systems that not only analyze data but independently generate decision proposals and derive recommendations for action. For midsize companies, this means: The gap between a missing data science team and professional data analysis is increasingly being closed by AI-powered tools.

Practical guide: six steps to a data-driven decision culture

Getting started with data analytics does not have to be expensive or complex. The following guide shows a proven path for midsize companies that begins pragmatically and scales step by step.

Step 1: inventory and prioritize data sources

Before you select a tool, you need to know what data is actually available. Conduct an inventory of all data sources: ERP system, CRM, accounting software, Excel spreadsheets, email inboxes, web analytics tools, and manual records. For each source, list what data resides there, who maintains it, how current it is, and in what format it exists.

Then prioritize by business relevance. Start with the data that has the most direct impact on your most important business decisions—typically revenue and sales data, customer data, and financial metrics.

Step 2: establish data quality

Without clean data, all analyses are worthless. The AI Study 2025 confirms: 76 percent of midsize companies struggle with inadequate data quality. Clean up duplicates, close data gaps, standardize formats, and define binding naming conventions. This step is not glamorous, but it is the decisive success factor. A pragmatic starting point: Bring the completeness of your master customer data to at least 95 percent.

Step 3: introduce a self-service BI tool

Choose a tool that fits your existing IT landscape and the competencies of your employees. For most midsize companies in the Microsoft ecosystem, Power BI is the logical entry point. Start with the free desktop version for a single analyst or power user and expand to Pro licenses as needed. Begin with a clearly defined use case—such as a sales dashboard or a revenue overview by region and product group.

Step 4: implement a pilot project with quick ROI

Choose a use case that delivers visible results within four to six weeks. Suitable entry projects include:

• Sales analysis: Which products, customers, and regions generate the highest contribution margin?
• Inventory optimization: Which items have too high or too low a turnover rate?
• Customer analysis: Which customer segments are most profitable, and where is churn threatening?
• Cash flow forecast: How will liquidity develop over the next three to six months?

Step 5: build data literacy in the team

Train your employees in using the chosen tool and—even more importantly—in critical thinking with data. The Data Literacy and Data Science for midsize companies project, in which nine universities from Baden-Wuerttemberg have been collaborating with SICOS BW since 2019, demonstrates a proven approach: teaching fundamentals in mathematics, computer science, and business competence for data-based decision-making. For everyday business, internal workshops and peer learning formats, in which power users share their knowledge with colleagues, are often sufficient.

Step 6: scale and automate

When the pilot project is successful, expand step by step. Integrate additional data sources, build additional dashboards, and automate recurring analyses. Use cloud-based solutions to keep infrastructure costs low. According to a survey, 76 percent of respondents expect that big-data-as-a-service applications will improve the quality of data analyses, 64 percent anticipate savings in infrastructure and personnel costs, and 67 percent expect an acceleration of data-driven decision-making at C-level.

Industry example: metal processing company increases margin by 12 percent

A midsize metal processor in southern Germany with 85 employees and annual revenue of approximately €14 million illustrates what is possible with a pragmatic analytics approach.

Starting situation

The company managed its order, production, and financial data in three separate systems. Calculations were created in Excel, sales management was based on monthly reports from the ERP system, and production planning worked with its own spreadsheets. Decisions about bid prices, batch sizes, and material orders were largely based on the management’s experience.

Measures

Instead of hiring a data science team, the company introduced Power BI and connected it to the existing ERP system and financial accounting. An external consultant accompanied the implementation over three months. In parallel, two employees from sales and controlling were trained as internal power users. The first use case was a contribution margin dashboard that evaluated orders by customer, product, and machine group.

Results after twelve months

The data analysis revealed that 23 percent of orders were unprofitable under a realistic full-cost calculation—a fact previously concealed by lump-sum calculations. Through targeted price adjustments, the elimination of unprofitable product variants, and data-driven material ordering, the company was able to increase its net margin by 12 percent. The investment in Power BI licenses, external consulting, and training amounted to approximately €28,000—against annual margin increases of over €180,000.

The key takeaway: No data science team was needed. What was needed was a clear goal, the right tool, and two motivated employees with basic analytics competence.

Frequently asked questions

What does getting started with data analytics cost for a midsize company?

The entry costs are significantly lower than many business owners expect. Power BI Desktop is free to use, Pro licenses cost $14 per user per month. Zoho Analytics starts at €24 monthly, and Google Looker Studio is free in its basic version. A realistic total investment for a first analytics project—including data cleaning, tool introduction, and training—is between €10,000 and €35,000 for a midsize company with 20 to 100 employees. Against this stands a typical ROI that amortizes this investment within six to twelve months.

Do we still need IT skills on the team?

You do not need programming skills or data scientists. What you need is one or two employees willing to learn a self-service BI tool. The learning curve for Power BI is gentle for Excel-savvy users—basic dashboards can be created within a few days. External support can be useful for the initial setup of data connections and governance, but ongoing operations then rest in the hands of your business departments.

How do we handle data protection and GDPR?

Data protection is a legitimate concern but not a barrier. All leading BI platforms offer GDPR-compliant hosting options within the EU. Power BI can be operated through Azure data centers in Germany, and Zoho Analytics offers European data centers. The key is to define a clear data governance policy: Who may access which data? How are personal data anonymized or pseudonymized? Which retention periods apply? The AI Study 2025 shows that nearly all midsize companies rate data protection and AI security as critical, but only 24 percent have established a comprehensive governance framework.

How long does it take from start to the first actionable insights?

With a focused approach—a clearly defined use case, an already available data source, and a motivated power user—you can productively use initial dashboards and analyses within two to four weeks. A complete analytics pilot project with data cleaning, tool introduction, and training typically takes three to four months. The most common mistake is planning too long and starting too late. It is better to begin with a simple dashboard and iterate than to try to design a perfect system.

Which data should we analyze first?

Start with the data that has the greatest effect on your business results. For most midsize companies, these are revenue and sales data (Which products and customers are most profitable?), financial data (Where do costs arise, and how is cash flow developing?), and customer data (Which segments are growing, and where is churn threatening?). Avoid trying to analyze all data simultaneously. Focus beats breadth—especially at the beginning.

References

The following sources form the basis of this article and were published or evaluated in the period from 2025 to February 2026:

• maximal.digital—KI-Studie 2025: KI im Mittelstand und KMU [in German] (2025): Study results on AI relevance, data quality, and adoption rates in German midsize companies—https://maximal.digital/studie-ki-im-mittelstand-und-kmu-2025-einblicke-und-impulse-aus-der-ki-studie-2025
• Research and Markets—Self-service Analytics Market Size, Share and Forecast to 2031 (2025): Market analysis of the global self-service analytics market with growth forecasts—https://www.researchandmarkets.com/report/self-service-analytics
• Deloitte—Künstliche Intelligenz im Mittelstand [in German] (2025): Study on barriers and opportunities of AI adoption in mid-sized companies—https://www2.deloitte.com/de/de/pages/mittelstand/contents/kuenstliche-intelligenz-im-mittelstand.html
• bofest consult—Datengetriebene Entscheidungsfindung in KMUs: Die 3 größten Herausforderungen [in German] (2025): Practical report on cultural and organizational hurdles in the introduction of data-driven processes—https://bofestconsult.com/daten-entscheidungen-kmus/
• American Express Business Insights—Data Analytics for SMEs (2025): Practice-oriented guide to data analysis strategies for small and medium-sized enterprises—https://www.americanexpress.com/de-de/business/trends-and-insights/articles/data-analytics-kmus/
• SICOS BW—Data Analytics: Wissensaufbau für KMU leicht gemacht [in German] (2025): Report on the Data Literacy and Data Science for midsize companies project—https://www.sicos-bw.de/blog/data-analytics-wissensaufbau-fuer-kmu-leicht-gemacht/