Back to articles list
Database designers work with data structures and data flows, while data analysts extract insights from huge amounts of data. Both careers have similarities and differences, and pros and cons. Read along to decide which one is the best fit for you.
Do you like data, but find it hard to decide how you prefer to work with it? You have (at least) two paths to choose from: be the one who designs the structures to store and process the information, or be the one who analyzes large volumes of data in order to extract the truths hidden in them. That's the first of the differences between a database designer's job and a data analyst’s.
There are many points in common to both roles – to such an extent that, if you choose one of the two careers and eventually discover that it is not the one you really prefer, you’ll be able to use much of the knowledge you’ve acquired to switch to the other.
To understand the differences between a database designer and a data analyst, we’ll look at what’s involved in becoming a database designer or a data analyst. We’ll look at the differences and commonalities, the pros and cons, when comparing database designer vs. data analyst roles.
The Database Designer
A database designer defines the optimal structures for storing and processing information. We could assume that, once a database is created, the designer has nothing else to do.
Not quite.
Database designers do much more than just drawing schemas (e.g. data modeling). They are also responsible for implementing their designs, transforming them into operational databases and maintaining the documentation that will enable developers and others to understand and use the data architecture.
It is also database modelers’ responsibility to ensure the efficiency of the databases created from their designs. Once a database design has become a working database, the designer must be available to see that the database maintains consistent information and that it performs well with all the applications that use it.
Becoming a Database Designer
The usual way to become a database designer is to obtain a degree in computer science, then specialize in databases. This allows you to develop a broad mastery of:
- Database theory: e.g. the relational model, normalization, conceptual/logical/physical schemas, etc.
- Data definition and manipulation languages: mainly SQL, but also Python and R.
- Database optimization: trace analysis, index creation.
- Data warehousing: types of schemas (snowflake, star, galaxy), dimension and fact tables, types of dimensions, etc.
- ETL processes: or how to populate databases and data warehouses.
- Big Data and non-relational databases.
You will also need years of experience to learn the things that you won’t find in books. To get an idea of what this entails, read about the lessons I’ve learned from years of data modeling.
You Need More Than Technical Skills
In addition to the technical skills that a database designer must master, there are a series of “soft skills” (as opposed to purely technical “hard skills”) that involve communication, interpersonal relationship management, and business acumen.
A database designer must be able to interpret the requirements of a project’s users, sponsors, and stakeholders – and these people rarely master technical language. So, it’s vital for a database designer to have a constructive dialogue with all parties involved in a project. This is not only important for interpreting requirements, but also for documenting and conveying design decisions in a way that anyone can understand their justifications.
Business acumen is also important for database designers; it lets them align their design decisions with the company or project’s business objectives. Database designers must also be team players, as their work must be coordinated with that of everyone on the software development team.
Database Design as Part of the Software Development Process
It is very common for database design to be a stage in the software development life cycle (SDLC). In software development teams, database designers work with the results of requirements analysis and generate database schemas, which help developers define the interactions between applications and data.
A typical database design process includes reading software product requirements, identifying entities and relationships based on those requirements, generating data models containing those entities and relationships, and finally converting those models into operational databases. For these tasks, the designer uses design tools (such as Vertabelo) that facilitate the creation of entity-relationship diagrams (ERDs). Such tools often also generate scripts to transform ERDs into databases and derive diagrams from existing databases (i.e. reverse engineer the database). There are many benefits that justify the use of ER diagrams in database design; see this article on database design with Vertabelo to learn what you can do with this tool.
Once the development is completed and the databases move to the production stage, the database designer must be available to make changes to the models. These could be due to changes in requirements, optimization needs, or refactoring to improve the quality of the software product.
Before going to a job interview for a database designer position, read about the top 19 database modeling interview questions.
The Data Analyst
If you ever have a big pile of chaff and need someone to come in and separate it from the wheat, find a data analyst – he or she will be happy to do it. (In the previous sentence, replace "chaff" with "data" and "wheat" with "insights".)
A data analyst is someone who uses SQL, databases, statistics, data visualization programs, and more to gather data, prepare it, analyze it, and draw insights from it. They also develop reports, presentations, and visualizations (i.e. charts and graphs) to help people understand the information and its context. They play an important role in many organizations; without them, key facts would stay hidden in the data.
Becoming a Data Analyst
There are universities that award specific degrees in data science, which would be ideal for applying for a data analyst job. But if you read about how to become a data analyst, you’ll find that a degree is not essential; many data analyst jobs require specific knowledge (such as applied mathematics or statistics) and some computer science skills (database design, programming, or working with data analysis tools).
The Data Analysis Process
To obtain insights from mountains of data, data analysts apply a series of steps that we call the data analysis process.
An important difference between a database designer and a data analyst is that a data analyst is not usually part of a software development process. Data analysis is a process in itself. It typically breaks down into the following stages:
- Identify information needs and data sources.
- Collect data.
- Clean the collected data.
- Model the data.
- Analyze the information.
- Interpret the data and present the results.
Let’s see what each of these steps looks like.
1. Identify Information Needs and Data Sources
At this stage, analysts must clearly establish the results expected from their work. This is similar to the requirements engineering stage of the software development process, only the definitions required are narrower. Basically, data analysts only need to know what information they will work with and what questions they need to answer in their analysis.
For real estate information, as an example, this might be:
- Data sources:
- Surveys conducted with sellers and potential buyers.
- Files resulting from a periodic web-scraping process of all public real estate sales sites in the United States.
- Questions to be answered:
- Which demographic groups are firmest in their intentions to purchase real estate?
- Which regions are experiencing the greatest increase in property values?
2. Data Collection
The data that an analyst must work with does not usually come on a platter. Indeed, one of the skills that makes them most valuable is knowing how to collect the data – and how to automate that collection as much as possible.
At the beginning of the collection process, quite often the data does not even exist in a digital or physical medium. This is the case with information from surveys, interviews, or polls. This original, raw information is extracted directly from the sources; it’s known as primary information. Data analysts are expected to know how to develop surveys, interviews or polls to obtain useful and valuable primary information.
Information that already exists in some physical or digital form is called secondary information. This can be data from previous surveys, data residing in databases, data from sensors, public information services, etc.
Secondary information can be structured (i.e. if it resides in an existing database) or it can reside in XML files, text files, or even digital audio or video files.
3. Data Cleaning
The primary data collected in the previous step never comes ready to use. It is raw data, and like raw food, it must be prepared before it can be consumed. There is always incomplete, duplicate, corrupt, or poorly formatted (i.e. unusable) data. For that information to be usable, it must be cleaned.
There is no one-size-fits-all data cleansing method. The cleanup task varies for each project and even for each data set. It is common for this task to require scripting in a language – usually R, Python, or SQL – with the power to manipulate datasets. Or analysts might employ a data cleansing tool, such as OpenRefine (formerly Google Refine), WinPure Clean & Match, or Trifecta Wrangler.
4. Data Modeling
This is the phase of the data analysis process that has the most in common with database design work.
The previous stage of data cleansing usually results in a usable data set stored in an easily-accessible data repository. You could perform your analysis tasks directly on that repository. But if you do some data modeling before the analysis, analysis tasks can be done much more quickly. For example, if you design a data warehouse schema that houses the collected data, you can employ data visualization tools that directly access that warehouse. The data warehouse can be kept up to date with new information, so that the analysis tasks are always done on current data.
At this point, data analysts should put on their database designer's hat and use the knowledge and tools we have already described for this area – or ask a database designer to help them in the data modeling tasks.
5. Data Analysis
Here is where data analysts must demonstrate their knowledge of applied mathematics and statistics to obtain answers to the questions developed in the first step of the process. In general, the task of data analysis involves finding patterns and trends that predict future events or expose the behavior of a population or universe. (In statistics, a population or universe is the entire group of units in the study.)
Analysts use spreadsheets (although you might want to read about alternatives to spreadsheets) and programming languages such as R, Python and/or SQL. They also use software packages like:
- SPSS or Stata for the statistical analysis of quantitative data.
- ArcGIS for geospatial data analysis.
- NVivo for qualitative (textual and audio-visual) data analysis.
6. Interpretation and Presentation of Results
After extracting the insights by means of the analysis process, the data analyst must know how to present them. This often involves some kind of data visualization (e.g. plots or charts). There is a wide range of options in terms of visualization tools, such as Tableau, QlikView, and Power BI.
At this point it is important not only to present the results of the analysis, but also to know how to explain – in simple and understandable terms for those unfamiliar with mathematics and statistics – how these results were obtained. There is an entire discipline called data storytelling specifically dedicated to this task.
Let’s revisit the example mentioned above on real estate data. After completing the process, the data analyst should be able to show the following results:
- A dynamic graph of real estate purchase intent completion percentages, grouped by demographic dimensions:
- Age groups
- Socioeconomic level
- Geographic area
- A dynamic graph of percentage increase in property values, grouped by geographic region.
Each graph and each result displayed should clearly indicate the collection method or data source used, as well as the population studied and the time period covered.
Choose Your Path: Design vs. Analysis
You probably now have an idea of the database designer vs. data analyst roles. To help you choose a path, let's look at some pros and cons of each of these professions. But these pros and cons are rather subjective, since what is a pro for one person may be a con for another.
Responsibilities
The results of a data analysis process are commonly used to make critical business decisions. For example, the result of a data analysis can be used to target a marketing campaign or a sales action to a certain market segment. But being a data analyst involves a great responsibility; a misinformed result can cause a company to suffer substantial setbacks. If the results of the data analysis are wrong, millions of dollars can be lost.
The responsibilities of a database designer are also important. However, a mistake in a designer's work is more easily salvageable: there is always the possibility of deploying a backup or rollback. Even if a mistake was made in the design of a data warehouse and it has to be rebuilt from scratch, the cost may be high, but it is unlikely to reach the level of the cost of errors in a poorly-crafted data analysis.
Income
Because data analysts' responsibilities reach higher levels than those of database designers, their average income is also higher. According to data provided by Glassdoor, data analyst salaries in the United States are about 10% higher than database designer salaries.
Work Environment
A database designer is typically part of a software development or application maintenance team. This is a pro if you enjoy working in a team, interacting with peers, validating your deliverables with them, sharing responsibilities, and coordinating your schedule with your teammates.
But if your work style is more individualistic, you'll be more comfortable with the way a data analyst works. You will have to be able to shoulder an analysis project and be the main (or the only) person responsible for its success or failure. You will have to interact with different areas of the company: those who expect the results of your analysis (financial area, marketing area, general management) and those who can help you with the technical aspects of obtaining and processing the data (database designers, MIS/IT employees, etc.).
Career Paths
As a database designer, you can move up the ranks within the IT and software development team. Once you have a few years of work experience as a database designer, you can aspire to a position as a database engineer, then a database architect and, if you have enough leadership qualities, you can become a systems manager or even a CTO.
As a data analyst, you can work your way up to different analyst positions. These depend on the industry segment you are in, e.g. business analyst, financial analyst, risk analyst, marketing analyst, or Business Intelligence analyst. But if you want to stay in the field of multipurpose data analysis, there’s also the possibility to go from data analyst to data scientist.
Want to Do Database Design and Data Analysis?
If you manage to cover both sides of the data work universe – that is, you master both the world of databases and data structures and the world of information analysis – you can aspire to a relatively new position called Chief Data Officer, or CDO.
The CDO is a corporate officer responsible for the governance and utilization of information across the enterprise via data processing, analysis, data mining, information trading, and other means. In simple terms: A CDO is a leader who creates business value from data.
If you choose that path, the journey will not be easy, but the rewards will be great. So, best of luck and happy learning!
