Sweet Data O’Mine, and your 3 quirky problems
If data were a rebellious teen, how would you resolve these 3 issues to allow him/her to invent algorithmic literature and much more?
Data has brought us numerous marvels, yet my all-time favorite remains an NLP-generated Harry Potter chapter by Botnik that was crafted by feeding all seven volumes of this wizarding saga into a predictive algorithm. Take a look at a snippet from this breathtaking marvel called Harry Potter and the Portrait of what looked like a large pile of ash:
Harry tore his eyes from his head and threw them into the forest. Voldemort raised his eyebrows at Harry, who could not see anything at the moment.
Want to read this story later? Save it in Journal.
Interested in feeding articles of your choice and crafting the next masterpiece? Take a look at this algorithmic text emulator and become the next J. K. Bowling.
Onto the rebellious teenager that is data (or at least while the parents don’t understand it)
Let’s call our teen Datina. She has brought us numerous marvels ranging from natural language processing (NLP) that paved a path for more advanced chatbots and speech recognition technology to algorithms capable of generating faces of people that don’t exist, and allowed us to become a whole lot more animal-like through filters.
Data and teens have quite a bit in common — sometimes it might be troublesome to find the right way to communicate with them or they might surprise you with some of their suggestions. Although, if the appropriate methods are chosen, they can bless you with truly ingenious outcomes, such as the reinvented wizarding saga example above. Looking for their inputs might often lead to unique solutions, but it is not always the right way to go.
If you don’t find the right way to communicate with Datina, you’ll eventually find that much of your data-backed problem solving is futile.
Most of us, regardless of industry, aim to make decisions with the assistance of this rebellious teen, which is terrific, but, unfortunately, this often leads to flawed decision making. Here are some of the most common issues when trying to get along with Datina.
1. Sometimes it is better to use your experience to do the work rather than asking Datina
Even though we’ve advanced quite a bit with natural language processing and similar tools, it should be kept in mind that some problems are not informed by data. Activities such as brand messaging are still often best to be left to the marketing rather than data specialists. It is tough to imagine what analyses should be conducted to come up with slogans such as “Just do it” or “got milk?”.
Mitigation
Identify which decisions require exceptional creativity, intuition, and human ingenuity, then spice it with some qualitative analysis. Complete the task at hand and analyze its effectiveness using quantitative analysis.
2. You and Datina are having some communication issues and you can’t get an answer from her
Trying to make data-based decisions without having an understanding of what might be the unknowns, e.g. lack of competitor data, lack of specificity in the “other” data category, and other miscellaneous data intricacies might lead to erroneous assumptions that will lead to overall poor decisions.
You could try deciding on how many treadmills should be installed in your gym by analyzing the number of people running with their dogs in the neighborhood — the action of running is similar in itself, but there is an obvious lack of incentive analysis.
Mitigation
Some of the potential solutions could be sending users follow-up surveys to eliminate data blackspots, redesigning current systems to collect more accurate (better segmented) data, or buying data from third parties. It is crucial to evaluate whether making a long-term investment, such as redesigning systems, is worth it or maybe a one-time contracting of third party data providers would be enough.
3. Datina knows the answer to your question, but you do not listen to what she is saying
The data team (or anyone knowledgeable in data) is not involved in the decision making, hence, Datina’s full potential is not utilized. The person, e.g. the manager, that requested an analysis may have been too specific with the analysis request, requested a very limited data sample or chose to nitpick the decision supporting data. In such a case, the decision could be well-informed, but due to some hiccups in the process — it is not.
Mitigation
The data team should be utilized by involving it from the problem conception stage to develop a more effective solution. The data team is equipped with the crucial skills to understand what kind of data is needed, what kind of data could be accessed, and what are potential data black spots. Furthermore, the data team could be consulted when deciding on the key KPIs due to their understanding of the accessible data.
Conclusion
As with teens, many of the problems arise due to a lack of understanding, which, in our case, would be data illiteracy. Data illiteracy comes in various shapes and forms.
Some, like in the first case, expect data to deliver more than it can, acting anywhere between steroid (accelerating growth) to panacea (fixing everything), while others view it as a “Used data” box to be checked off.
In the second case, it was the example of assuming that any decision informed by data must be a good one. Lastly, in the third case, the decision might be made too early and nitpicked data is used just to gather additional perceived legitimacy.
Many of the aforementioned issues could be tackled by having a person that is good at communicating with this rebellious teen and would ensure that its benefits are employed responsibly and effectively. Moreover, the responsibilities in utilizing data for problem-solving could range from determining what data is needed to what should be collected or generated if it is not readily available. Such a role is often assumed by the Head of Data or members of the data team whenever the problems in question are of lower scope.
📝 Save this story in Journal.
👩💻 Wake up every Sunday morning to the week’s most noteworthy stories in Tech waiting in your inbox. Read the Noteworthy in Tech newsletter.
