How to Become a Data Analyst

I’ve been working with data for some 20 plus years as of the writing of this post. In the video below I captured my thoughts on the required hard and soft skills it takes to succeed as a data analyst. If you are looking to start your career in data as someone who has not yet graduated or as someone with tangential work experience, then this video will serve you well.

Do You Need a Computer Science Degree to be a Data Analyst?

This question is frequently asked by people such as yourself looking to make a move into data. The answer is no. You do not need a computer science degree to have a very successful data career. In the video I give my thoughts on computer science, but the reality is that although it may be helpful from a “getting a first job” perspective, it is not a requirement to succeed. Although I have an undergraduate computer science degree from Clark Atlanta University (shout-out to HBCU alums), some of the brightest minds I’ve worked with in the data space do not have a computer science degree. Bottom line; a formal computer science degree certainly helps but it is by no means necessary. All you need is the willingness to learn the tools and the perseverance to get your first data opportunity.

Hard Skills Required (View Video)

I’ll give you a hint, data visualization skills are a must and Tableau is the tool of choice for me.

Soft Skills Required

I’ll keep it short here and simply state that you should always look for ways to differentiate yourself and not just be seen as an interchangeable commodity worker. To paraphrase famed Harvard professor Michael Porter, a differentiation strategy advocates that a business must offer products or services that are valuable and unique to buyers above and beyond a low price. In this metaphor, think of yourself as a business and you bring multiple skill sets to your employer (other than being a single focus technical employee who can be easily outsourced for a lower price).

To be a differentiator, do not think of yourself as just being a tool specific analyst. Learn how to take requirements, communicate well, develop exceptional writing skills for business emails and documentation. Finally, learn how to present your analyses to people several pay grades above yourself when required. You want differentiation to be your competitive advantage. You do not want “low cost” to be your advantage for obvious reasons (if you’re like me, you want to be paid fairly for the value you provide).

Future Career Paths

In our jobs we desire mastery, autonomy and purpose. After a certain point in your career you may want to take a leap from the descriptive analytics path and move towards a predictive analytics path. Descriptive analytics (think data analyst or traditional business intelligence reporting analyst) deal with what has happened in the past while predictive analytics focus on what will most likely happen in the future. In order to level up in predictive analytics, you will need python, statistics, probability, and/or machine learning skills.

If you want to make the leap from data into management, you can consider obtaining an MBA or a masters degree in Management Information Systems. I happen to have an MBA from the Georgia Institute of Technology and a masters degree in Information Management from Syracuse. This may seem like a bit of overkill but I work in consulting where credentials are overly appreciated by clients (and I am a lifelong learner).

Interact with my Tableau resume here.

Conclusion

A career in data can be fun (in the early learning phases) and lucrative (mid to late career). In my case it has been a fulfilling career ever since I started work as a data analyst at General Motors many years ago. I turned myself from a commodity to a differentiator by not only learning the basics but also adding business understanding and a willingness to share what I know on this blog and my YouTube channel. I know that you can do the same. If you put in the time to learn along with the perseverance to land that first data role, you won’t need much luck at all to accomplish your goals.

Looking to land that first role or trying to move ahead in your current role? Then check out this post for the Keys for a Successful Career as a Data Analyst.

-Anthony Smoak

All views and opinions are solely my own and do not necessarily reflect those of my employer

I appreciate everyone who has supported this blog and my YouTube channel via merch. Please click here

Thank you!!

The Data Quality Face-Off: Who Owns Data Quality, the Business or IT?

This article is also posted on LinkedIn.

Data is the lifeblood of organizations. By now you’ve probably heard the comparative slogans “data is the new oil” or “data is the new water” or “data is the new currency”. A quick type of “data is the new” into the Google search bar and the first result delivered is “data is the new bacon”. I’m not sure how apt this slogan is except to say that both can be highly fulfilling.

With the exception of a well-known Seattle based retailer, most enterprises experience substantial data quality issues as data quality work is typically an exercise in “pass the buck”. An Information Week article shrewdly commented on the risks associated with the lack of data quality:

“There are two core risks: making decisions based on ‘information fantasy,’ and compliance. If you’re not representing the real world, you can be fined and your CFO can be imprisoned. It all comes down to that one point: If your systems don’t represent the real world, then how can you make accurate decisions?” [3]

The Handbook of Data Quality: Research and Practice has reported that 60% of enterprises suffer from data quality issues, 70% of manufacturing orders contained poor quality data and that poor data quality management costs companies roughly $1.4 billion every year [4].

Organizational Hamilton/Burr style face-offs occur in which IT and the business are at loggerheads over the state of data quality and ownership. The business typically believes that since data is co-mingled with the systems that IT already manages, IT should own any data issues. With the high costs of poor data quality I just cited, and the risks of nimble disrupters utilizing data more efficiently to attack incumbents’ market share, both IT and the business need to be on the same team with regard to data quality for the organization’s sake.

“The relationship between IT and the business is a source of tension in many organizations, especially in relation to data management. This tension often manifests itself in the definition of data quality, as well as the question of who is responsible for data quality.” [5]

Anecdotally, IT units do not have the desire to be held responsible for “mysterious” data and/or systems that they had no hand in standing up. In my opinion, the enterprise IT mindset is to make sure the data arrives into the consolidated Enterprise Data Warehouse or other centralized data repository; and if downstream users don’t raise concerns about data quality issues, all the better for IT. Garbage-In, Garbage-Out. If the checksums or record counts from source to target match, then it’s time to call it a day.

The developer or analyst related mindset is to immediately dive in and start building applications or reports with the potential to deliver sub-optimal results because the data was misunderstood or misinterpreted as the “golden copy”. Up-front data profiling isn’t in the equation.

Gartner has suggested that the rise of the Chief Data Officer (particularly in banking, government and insurance industries) has been beneficial towards helping both IT and the business with managing data [2]. The strategic usage of a CDO has the potential to free up the CIO and the enterprise IT organization so they can carry on with managing infrastructure and maintaining systems.

However, most experts will agree that the business needs to define what constitutes high-quality acceptable data and that the business should “own the data”. However, IT is typically the “owner” of the systems that house such data. Thus, a mutually beneficial organizational relationship would involve IT having a better understanding of data content so as to ensure a higher level of data quality [5].

From a working together perspective, I find this matrix from Allen & Cervo (2015) helpful in depicting the risks arising from one sided data profiling activities without business context and vice versa. It illustrates how both “business understanding and data profiling are necessary to minimize any risk of incorrect assumptions about the data’s fitness for use or about how the data serves the business” [1]. Although originally offered in a Master Data Management context, I find the example fitting in illustrating how business and IT expertise should work together.

picture1From: Allen, M. & Cervo, D (2015) Multi-Domain Master Data Management: Advanced MDM and Data Governance in Practice. Morgan Kaufmann Publishers. Chapter 8 – Data Integration.
  • From the bottom left quadrant, low business knowledge and inadequate data profiling activities leaves the enterprise in a less than optimal position. This is not the quadrant an organization needs to languish within.
  • The top left quadrant illustrates that business context is high but “knowledge” is unsubstantiated because of the lack of understanding of data quality via profiling exercises. Cue quality guru W. Edwards Deming stating, “Without data, you’re just a person with an opinion.”
  • The bottom right quadrant illustrates the opposite problem where data profiling without business knowledge doesn’t yield enough context for meaningful analyses. Cue a “bizarro” W. Edwards Deming stating, “Without an opinion, you’re just a person with data.”
  • The “Goldilocks” quadrant in the upper right yields appropriate understanding of data quality and the necessary context in which to conduct meaningful analyses.

The more engaged that the business and IT are in finding common ground with respect to understanding existing data and solving data quality issues, the better positioned the organization is to avoid regulatory fines, departmental strife, threats from upstart firms and overall bad decision making. Refuse to become complacent in the data space and give data the attention it deserves, your organization’s survival just may depend upon it.

References:

[1] Allen, M. & Cervo, D (2015) Multi-Domain Master Data Management: Advanced MDM and Data Governance in Practice. Morgan Kaufmann Publishers. Chapter 8 – Data Integration.

[2] Gartner, (January 30, 2014). By 2015, 25 Percent of Large Global Organizations Will Have Appointed Chief Data Officers
http://www.gartner.com/newsroom/id/2659215

[3] Morgan, Lisa. (October 14, 2015). Information Week, 8 Ways To Ensure Data Quality. Information Week.
http://www.informationweek.com/big-data/big-data-analytics/8-ways-to-ensure-data-quality/d/d-id/1322239

[4] Sadiq, Shazia (Ed.) (2013). Handbook of Data Quality: Research and Practice: Cost and Value Management for Data Quality.

[5] Sebastian-Coleman, L. (2013). Measuring Data Quality for Ongoing Improvement: A Data Quality Assessment Framework. Morgan Kaufmann Publishers. Chapter 2 – Data, People, and Systems.

Picture Copyright: wavebreakmediamicro / 123RF Stock Photo

The IT Department Needs To Market Its Value or Suffer the Consequences

This article is also published on LinkedIn.

By now it’s an all too common cliché that the IT department does not garner the respect it deserves from its counterpart areas of the business. This perceived respect deficiency can manifest itself in the lack of upfront involvement in business strategy (we’ll call you when it breaks), unreasonable timelines (do it yesterday), rampant budget cuts and layoffs (do more with less) and/or limited technical promotional tracks (promotions are for business areas only).

IT pros tend to believe that if they’re adding value, delivering difficult solutions within reasonable timeframes and providing it all in a cost efficient manner, the recognition and gratitude will follow. Typical IT and knowledge worker responsibilities fall under the high level categories of “keep things running” (you’re doing a great job so we don’t notice) or “attend to our technical emergency” (drop what you’re doing).

It’s fair to say that there is a perception gap between the true value and the perceived value of what IT brings to the table. Anecdotally, there certainly seems to be a disconnect between the perceived lack of difficulty in business asks and the actual difficulty in delivering solutions. This perception gap can occur not only between IT and the “business” but also between the non-technical IT manager and the technical rank and file.

In this era of automation, outsourcing and job instability, there is an element of danger in one’s contributions going unnoticed, underappreciated and/or misunderstood. Within IT, leaders and the rank and file need to overcome their stereotypical introverted nature and do a better job of internally marketing their value to the organization. IT rank and file need to better market their value to their managers, and in turn the IT department collectively needs to better market its value to other areas of the business.

Perception matters, but IT must deliver the goods as well. If the business misperceives the actual work that the IT department provides and equates it to commoditized functions such as “fix the printers” or “print the reports” then morale dips and the IT department can expect to compete with external third parties (vendors, consulting firms, outsourcing outfits) who do a much better job of finding the ear of influential higher–ups and convincing these decision-makers of their value.

I once worked on an extremely complex report automation initiative that required assistance from ETL developers, architects, report developers and business unit team members. The purpose was to gather information from disparate source systems, perform complex ETL on the data then and store it in a data-mart for downstream reporting. Ultimately the project successfully met its objective of automating several reports which in-turn saved the business a week’s worth of manual excel report creations. After phase 1 completion, the thanks I received was genuine gratitude from the business analyst whose job I made easier. The other thanks I received was “where’s phase 2, this shouldn’t be that hard” from the business manager whose technology knowledge was limited to cutting and pasting into excel.

Ideally my team should have better marketed the value and helped the business partner understand the appropriate timeliness (given the extreme complexity) of this win, instead of just being glad to move forward after solving a difficult problem for the business.

I believe Dan Roberts accurately paraphrases the knowledge worker’s stance in his book Marketing IT’s Value.

“’What does marketing have to do with IT? Why do I need to change my image? I’m already a good developer!’ Because marketing is simply not in IT’s comfort zone, they revert to what is more natural for them, which is doing their technical job and leaving the job of marketing to someone else, which reinforces the image that ‘IT is just a bunch of techies.’”

The IT department needs to promote better awareness of its value before the department is shut out of strategic planning meetings, the department budget gets cut, project timelines start to shrink and morale starts to dip. IT workers need to promote the value they bring to the table by touting their wins and remaining up to date in education, training and certifications as necessary. At-will employment works both ways, if the technical staff feels stagnant, undervalued and underappreciated, there is always a better situation around the corner; especially considering the technical skills shortage in today’s marketplace.

“It’s not about hype and empty promises; it’s about creating an awareness of IT’s value. It’s about changing client perceptions by presenting a clear, consistent message about the value of IT. After all, if you don’t market yourself, someone else will, and you might not like the image you end up with [1]”

References:

[1] Colisto, Nicholas R.. ( © 2012). The CIO Playbook: Strategies and Best Practices for IT Leaders to Deliver Value.

[2] Roberts, Dan. ( © 2014). Unleashing the Power of IT: Bringing People, Business, and Technology Together, second edition.