Understand the Differences and Identify the Right Talent
By far the most common challenge in People Analytics is identifying, attracting and retaining the right data experts. For organizations at all stages of their people analytics maturity, it remains the most central challenge to solve because it is a prerequisite for impactful people analytics.
If your organization is struggling to identify, attract and retain the right talent to achieve mission success, you are not alone. Most HR executives in organizations that have ventured into people analytics will recognize one or more of the following challenges when attempting to achieve a high-functioning people analytics capability:
“It is hard to identify what type of data expert to search for, and what specific skills they should have.”
“The recruiting process was long and expensive. In some cases, our search failed entirely, leaving us with no talent to help implement our people analytics plan.“
“We had to pay far more for the talent we hired than we had planned or we feel we can afford.”
“The data expert we hired is great, but it turns out they are overqualified for what we need them to do at this stage, and I think we would have been better off with someone less advanced”.
“The data expert we hired was great, but they quickly left because another organization offered them a more exciting position, and a lot more money than we could offer.”
These are the top five common issues we hear from clients and colleagues regarding unsuccessful HR data job searches.
The good news is that solving the very first concern on the list - properly identifying what data expertise your organization actually needs to hire - will drastically lower the likelihood of experiencing any subsequent issues.
Check out this companion [client case study] that breaks down how we helped a client find more effective data expertise talent, at a much lower cost and much faster than they typically experience by carefully defining what kind of expert they actually needed.
Determining what kind of data expert to hire
In a nutshell, there are two different types of data experts in HR: Data analysts and data scientists. Both roles involve working with data to generate insights and inform decision-making, but they require different levels of expertise and training, and they (should) serve different needs.
Data analysts are typically responsible for collecting, cleaning, and analyzing data to identify patterns and trends. They often use tools like Excel, SQL, and Tableau to work with data and generate insights. They may also be responsible for creating reports, dashboards, and visualizations to communicate their findings to stakeholders.
Data scientists, on the other hand, are typically responsible for more complex tasks like machine learning, predictive modeling, and data mining. They work with both structured and unstructured data and use tools like Python and R to work with data and generate insights. They may also be responsible for building predictive models, developing algorithms, and designing experiments to test hypotheses.
Here is a quick breakdown of how they tend to be different.
It’s tempting to think that it would always be better to have a data scientist, because they appear to do everything a data analyst does, except better, and then they can do a lot that a data analyst can’t do. But this fallacy is the most common reason for “Recruiter’s remorse” that many HR executives will tell you they’ve felt after successfully recruiting an advanced data expert. In fact…
…the right data expert for HR is almost always a data analyst.
The most obvious reason is that data scientists tend to be a lot more expensive to recruit and hire. A good senior data scientist will often command a higher salary than the chief HR executive in the organization, and an entire team of data scientists could cost more in salary than the entirety of the HR function combined in a mid-sized organization.
But there are also several other, albeit more subtle reasons why it is important not to hire a data scientist if you don’t need one - even if you can afford to. Not taking these into account risks your organization derailing People Analytics and not getting the value you need from it.
Employees who are not challenged in their day-to-day work tend to become unengaged, and they have little opportunity to develop. Best case scenario they abruptly leave, while worst case they may choose to stay while underperforming compared to what should be expected from their price tag.
Data scientists are, by definition, more technically skilled. However, in many cases soft skills are at least as important, if not more important, for a people analytics expert. While it's not always the case, data scientists often excel in technical expertise more than soft skills.
It is natural for a data scientist to attempt to apply their advanced skills, even in areas where more basic analytics would be more appropriate. The consequence tends to be that products are overbuilt and fail to meet the basic needs of the intended audience. The analytics will be advanced, but not as actionable as insight from more appropriate analytics would be.
Justifying the cost of People Analytics to the rest of the organization is often the biggest obstacle for organizations getting started. Spending the first resources on hiring a data scientist increases the likelihood that people analytics will under-deliver on value, compared to the cost.
Highly advanced expertise can deliver tremendous value, but typically only in tandem with other advanced expertise, and supportive functions. Rocket scientists deliver the most value with a team around them, while a car mechanic can deliver most of their value without a team.
In other words, only search for a data scientist if…
…you need skills you cannot find in a data analyst
…you can (and want to) afford the added expense
…you are confident they can deliver and demonstrate value to the organization significantly above and beyond the cost they entail
…the organization has (or will have) the technical tools and the general data maturity necessary for the data scientist to be effective
You can also read this [client case study] where we break down how and why our client went from wanting to recruit a Senior Data Scientist to actually looking for a Junior Data Analyst.
Understanding What Data Expertise you Need
The goal here is to help you understand whether you generally need the skills of a data analyst or of a data scientist.
Both data analysts and data scientists are part statistician and part business expert. That is, both will have a background in statistical methods, and both will have and apply domain knowledge. That is, they will both need to have and apply knowledge of HR, your organization and other relevant subject matter.
What generally separates a data scientist from a data analyst is technical expertise in computer science. Advanced skills in programming and data management primarily - using tools like Kubernetes, Apache Spark and TensorFlow. They will also typically have advanced skills in common programming languages like R or Python.
Don’t buy into the common fallacy that data scientists are more advanced in statistics and business expertise than data analysts.
Sure, it tends to be true that on average data scientists are more advanced in statistics than data analysts, but that is far from always the case. A data scientist whose expertise is primarily data management, for example, need not be a strong statistician to be effective, nor do they necessarily need more than cursory business knowledge. They are mostly computer scientists with some statistics and business expertise.
Here are some general rules of thumb for deciding on a data analyst versus a data scientist:
If you do not need your data expert to know advanced coding languages or database management expertise, chances are that you will be much better off with a data analyst.
Coding languages whose entire or main purpose is data analysis, like R, SPSS, Stata or SAS are common skills among data analysts and not generally sufficient for someone to identify as a data scientist. Only if you need the data expert to use these languages to do highly advanced machine learning models, data management, or similar, should you consider a data scientist. Even then, you may want to consider whether that skillset could be achieved by a data analyst who already knows the language and simply needs to learn how to apply it in a new way.
If your data expert will work exclusively with descriptive and investigative people analytics, or with basic predictive and prescriptive analysis, hire a data analyst - conducting those analyses do not require computer science expertise to implement. Just make sure the analyst has the statistical skills to match the level of analysis.
Use this resource to understand which level of analysis makes sense for the role you’re hiring for. You will also learn why getting the level right - not just for this role, but for your organization as a whole - is crucial for your organization’s success in People Analytics.
For advanced predictive and prescriptive analytics (where most artificial intelligence and machine learning lives), a data scientist is usually the appropriate choice because implementing the methods behind that kind of analysis requires advanced coding. Very few organizations are ready for and require this type of analytics.
Finally, it’s important to understand that if the expertise you need will be rooted very deeply in subject matter knowledge - of HR in general, your organization, your industry, behavioral science or something else - it is usually far better to hire primarily for that, and for technical skills second. In fact, you may want to find someone internal, with an affinity for numbers, to skill up to meet your needs. Foundational data science skills are far easier to build than to train a data analyst or data scientist in deep subject matter knowledge, which can sometimes only be gained through experience and not training.
You should now have a good sense of whether a data analyst or a data scientist is the best option for the role. Take the next step and learn how to attract the right data analyst or data scientist for the role and your organization, and how to maximize the chance of finding the right candidate.
If you're uncertain about the role you're hiring for because it falls somewhere in the gray area between the two types of data experts, don't worry. You can use the resources mentioned above to gain a better understanding of what each type of data expert entails. If you're looking for more in-depth or customized advice, our team at Opterion Consulting is here to assist. You can reach us via email at info@opterionconsulting.com, or by calling +1 979-985-9714 in the Americas and +45 6168 9714 from the rest of the world.
Comments