What's wrong with being a gold digger

Data for Growth - These 10 Tips Will Make You a Gold Digger!

This is how you can turn data into gold! Learn from this article with 10 tips how to use data successfully to optimize internal processes and customer communication. With the free infographic from datenGOLD (downloadable at the end of the article) you always have all the steps in front of your eyes and become a gold digger!

Data is the gold of the 21st century.

It is nothing new that companies like Facebook, Spotify or Amazon are also so successful because they use the terabytes of user data to optimize internal processes and customer communication. According to a 2013 report by the Aberdeen Group, companies that based their business decisions on data had an average annual increase in sales of 27%, while other companies only had 7% growth.

But why don't all online companies do it this way? Why does a large part of the digital mountain of data simply remain unused for important decision-making processes? And above all: How can you change that?

Data belongs in the corporate culture

A major reason for the untapped potential is that many companies lack a data-driven decision-making culture. Often decisions are still made on the basis of experience, gut feeling or the "HIPPO" principle (highest paid person’s opinion). Many companies are not even aware of how they can improve their processes through data-driven decisions. Some do not even see the added value in it, since they have grown successfully for years without extensive analyzes.

No wonder, because they do not know the value of possible wrong decisions and cannot calculate how high the growth would have been if, instead of the great wealth of experience, one had used the large mountain of data for company decisions. But when you suddenly realize that the competition is threatening to overtake you, you ask yourself what should be done differently.

But it is the case that a corporate culture develops and consolidates over the years. Company values ​​must be internalized and lived by being deeply anchored in the way employees think and work. This is the only way to change corporate processes in the long term.

That sounds complicated and after years of work. It is. But the good news is that you can start right away. If your own actions are geared towards a data-driven way of thinking and you act as a role model, you contribute to a step-by-step rethinking in your company.

The following 10 building blocks for a data-driven corporate culture show you how this works.

1. Know the goal of your website!

"What do we actually want to achieve with our website?"

Basically, this is a simple question, but you get different answers depending on who you ask in the company:

  • Answer 1: “Sure thing. Selling, that is, sales. What else?"
  • Answer 2: “Sales, yes, ok. But we should rather look at the numbers after returns. "
  • Answer 3: "It's actually more about pushing offline sales."

If it is not clear what the actual goal is, it is also unclear which KPI your own decisions should be based on. There can be several corporate goals. However, these should be put into a prioritized target matrix and shared across the company.

The next step is to have a consistent idea of ​​how exactly these goals are measured.

An example:

The company's main goal is to increase customer loyalty. But what exactly is it? This goal can be measured in different ways: by the number of repeat buyers, the churn rate or the long-term customer value. It must be clear what is meant by this corporate objective and how the achievement of objectives can be quantified with data.

The corporate goals are defined for individual departments in the form of sub-goals, because different departments usually also pursue different goals. However, all target figures contribute to the company-wide goal. These sub-goals can be measured by data and serve as an evaluation criterion for the individual teams for their own work.

2. Don't let your data get dusty, use it!

According to Forrestor, 74% of companies want to be “data driven”, but only 29% manage to make real deductions from the vast amount of data collected (smart data).

The mere collection of data is not the problem for most companies in the online industry. On the contrary: All possible actions of the user on and outside the website are recorded and written to a wide variety of databases. After a while, the result is a confusing mountain of data. What then really makes it a challenge is to ask the right questions and then to draw the right insights and conclusions from the vast amount of data. More on this in this article. We at konversionsKRAFT also speak of datenGOLD in this context.

"Actionable Insights" is the key word here, almost degenerated into a meaningless buzzword. On the other hand, you should adhere to a clear process of data analysis in order to be able to answer the relevant questions correctly:

You can find out how this process works in detail in the infographic on the datenGOLD process. These can be downloaded for free at the end of the article. ↓

3. A battle is won with allies!

It is often discussed whether a data-driven corporate culture is something that should develop top-down or bottom-up. Ultimately, it is a combination of both sides.

How difficult it becomes if one of the two sides does not go along with it is shown, for example, when a data analyst meets the so-called hippo, a manager who always relies on his "feeling" (the sacred feeling of goodness) when making important decisions, and who is often long-term Experience familiar. There are even cases of this species that almost refuse to take a look at the elaborate analyzes that have been specially created and which actually say exactly the opposite of the well-paid belly.

As a data-driven analyst, the first thing to do is to keep calm and come up with a strategy on how to get support anyway. Because this is important to have a lasting influence on company processes:

  • Goals: Put yourself in the shoes of the manager. What are his goals? What's his strategy? How can you support him in achieving the goals with the help of data analysis?
  • Potential analysis: With a business case you can quantify the additional value of data-driven optimizations for the company in EUR or show how high the loss is if you make mistakes with your gut decisions.
  • Benchmarking: Show how the competition works successfully with the database and what effects this has on company growth. What happens when you are constantly lagging behind the competition? Then your own company will be left behind in the long term by the competition and maybe even pushed out of the market.

4. Ensure direct results!

Of course you can start with an 18-month project to first connect all data sources, set up the tracking again and acquire, test and implement all the necessary tools.

In the same way, instead of simple analyzes, you can start building a neural network directly, which is then developed in 12 months and brings results.

Neither is a good idea.

Initial data analyzes should achieve ROI without lengthy, expensive projects. If these are successful, the support and the budget for larger projects come almost by themselves, simply because the trust is there.

The data analytics team is not a laboratory in which research is carried out for years until you get outside with the world domination formula. It must always remain practical, business-relevant and goal-oriented.

Often it is small things that require little effort, but are extremely helpful for certain teams. For example, a sensible report on the use of the filters in the online shop, a clean analysis of the shopping cart values ​​adjusted for the extremely high orders from the B2 area or the analysis of A / B test results for certain customer segments (e.g. new customers).

5. Don't play with wrong data!

Data-driven decisions are only really useful if the data is correct. I'm sure you'll immediately agree how frustrating it is to have to make important decisions based on data that you suspect is wrong. When tracking user behavior on the website in particular, incorrect implementations often appear, which, for example, lead to not all page views, conversions or clicks being recorded correctly.

Before starting the analysis, a plausibility check and data cleansing should always be carried out. This also applies to the subject of data integration and the linking of different data sources, e.g. data from the CRM system with the on-site data.

If an A / B testing tool is used, it is advisable to link this to the web analytics system. This not only enables deeper analyzes of the test results for individual segments, but also checks the data consistency.

Due to different tracking methods and cookie definitions, there will always be slight deviations between the two systems. It is important, however, that the trend is correct, i.e. that both systems show the same direction of results. More about it here.

Often there are different measurements of one and the same KPI in the web analytics tool. While there are no errors in the data here, these different calculations mean quite different things. The following picture shows how the conversion rate can be measured as an example:

Depending on the goal of the analysis, you would choose a different KPI: If you want to find out whether an optimization concept leads to users ordering at all, you would use the unique conversions / visitors. If it is more about retaining users and also recording follow-up orders, the second definition of the conversion rate would be more appropriate.

In web analytics systems there are often several conversion rate definitions that can be used for analysis. It is important that the different meanings are clear and accessible.

6. A little bragging is allowed!

Just as important as getting successful data projects off the ground quickly is to increase the visibility of the topic in the company.

This can be done, for example, in the form of regular team or department events in which colleagues give an update on their projects. A little presentation about how the last A / B test significantly increased the conversion rate and how even new insights into customer behavior were obtained with the help of segmentation analyzes, that's the real dataGOLD.

Some companies also have an internal communication system (intranet) or a blog that can be used specifically. Let the other teams see how a simple analysis can create value for the company. This not only increases the ambition to want to show such a success. The others are also made to want to incorporate more data into their decisions on their own.

7. Reach your goal with teamwork!

Most of the large online companies have clustered data-related skills in a team. There is a good reason for this form of organization: If you want to professionalize the topic within the company, you also need a team that takes care of all matters relating to the topic centrally.

Within the team there are different roles and responsibilities that are clearly defined. The process of data collection, analysis and generation of insights is coordinated and targeted.

This team is a central point of contact for other departments that need support in the area of ​​data analysis. The project manager ensures that the cooperation with the other departments is well organized and acts as an interface between the two worlds of business and science, which must be combined here.

In addition, the experts ensure that the topic is sufficiently visible to the outside world and that the knowledge is passed on to other departments (keyword enabling). In the best case, this team is also involved when it comes to the role data should play in medium to long-term corporate planning and how important strategic goals can be achieved through data analysis and the right technical infrastructure.

8. Fail fast and often - but learn!

Sure, especially when you try something new, something goes wrong. Or maybe twice. If you start the first data analytics project and it does not bring the output that you hoped for, the first hype is often over quickly.

To prevent this, you usually start with smaller projects that have a very high probability of being really successful. Nevertheless, failures can never be completely avoided. The only question is how to deal with it then. Nobody likes to go to their boss to tell him that the last A / B test did not produce a significantly positive effect or even a downlift.

In an article, Gabriel Beck explains how you can still draw findings from the test, which can be used for marketing-relevant decisions. And that's exactly the point. A messed up project remains a messed up project. But with the right knowledge of why it actually went like this, important next steps can often be derived that can ultimately lead to measurable corporate success.

9. Data for everyone!

In order to be able to fall back on data for as many company decisions as possible, unrestricted access to the data sources is required. However, this does not mean that all teams are given access to the raw data in the form of huge log files.

Rather, it is important to provide good tools that make it possible to work comfortably and purposefully with data. In the long term, bad tools mean that less and less data is included in the decision-making process. A good technological data infrastructure and tools that cover individual needs are therefore an important building block in order to really be able to live a “data driven culture”.

10. Knowledge is power

Data access and tools have no added value if they are not used properly. Basic knowledge of statistics, mathematics and Excel or R is important for valid analyzes and the derivation of the right measures. You really don't have to be a trained statistician to include data in your decision-making processes.

Most companies have experts in these areas (see 7th Data Analytics Team) whom one can turn to when things get complicated. Nevertheless, it is helpful to teach yourself the most important skills in order to do as much as possible yourself. The integration of data into your own work process becomes more and more natural and simple over time.


A data-driven corporate culture can only be developed if there are values ​​that are consistently lived by the employees and conveyed to the outside world. This requires that you recognize the added value of data for business decisions and try as often as possible to integrate data into your own process in order to be able to make better decisions.

All 10 datenGOLD tips at a glance

1. Know the goal of your website!

# Quantify the goals of the company and your department with the right KPIs.

# Measures should be specifically checked to see whether they have the potential to increase these KPIs and contribute to the company's success.

2. Don't let your data get dusty, use it!

Don't watch the mountain of data getting bigger and more confusing, but use the right data to answer important questions.

3. A battle is won with allies!

Be persistent and don't give up trying to convince the skeptics of the benefits of a data-driven approach.

Even if this is often tedious, it will help you in the long term to get support for the really big data topics.

4. Ensure direct results!

Don't make it too complicated at the beginning but focus on the low-hanging fruits. This quickly creates the first data ROI and, above all, trust.

5. Don't play with wrong data!

#A good tracking documentation in which the calculation of the web analytics KPIs is recorded creates clarity.

# This should always be kept up to date and made accessible to all teams that work with the data.

6. A little bragging is allowed!

# Be the data analytics evangelist. Speak as much as you can about the topic and create awareness and interest on all levels.

# Take this opportunity and be a role model for your colleagues to show how data can be used profitably.

7. Reach your goal with teamwork!

# For smaller companies, too, it makes sense to bring employees with an analytical background together in the organizational structure.

# This not only promotes the exchange between experts but also shows the relevance of this topic in the company.

8. Fail fast and often - but learn!

# Don't let failure discourage you, learn how things will go better next time.

# A "lessons learned" appointment leads to weak points being identified and optimized.

9. Data for everyone!
Conduct a needs analysis when choosing a tool. Find out which teams are working with the tools and what is important for them in everyday use.

# Before making a decision, there are test phases to find out whether these requirements are met in the practical test.

10. Knowledge is power!
# Spread the knowledge of the data analytics experts in the company, e.g. through internal trainings, workshops and close cooperation between the analytics team and other departments.

The datenGOLD infographic for free download

Get your way to datenGOLD for free as a download and become a gold digger with data-driven decisions!

You can download the infographic here for free: