data management

data management

 

Christophe De Greift provides a post designed to help you get it right with these five tips for a more data-driven 2021.

2020 accelerated several trends and one of them is the need to be ‘data-driven’ in decision making. From vaccine testing to last-mile logistics in e-commerce, data analytics is part of the path to making the right decision and generating value.

Having collaborated and talked with dozens of Peruvian companies this year on their greatest analytical challenges, I share below my recommendations for a 2021 full of analytical successes.

Team

Turning data into value requires such a diversity of knowledge and skills that it cannot be achieved without effective teamwork. However, collaboration in analytics does not occur naturally, as the business user seeks a practical solution to their problems and does not know about machine learning, while the data scientist prioritizes rigorous analysis and knows little about the business. A solution that several Peruvian companies have successfully implemented is the incorporation of an Analytics Translator to the team , connecting business users and data scientists.

  1. Strategy second

I have been a staunch advocate of strategic planning for business, despite mounting criticism. Data analytics also requires a plan, but that doesn’t mark the beginning of the analytics journey. Indeed, a minimum of knowledge about artificial intelligence, data governance and technology is required to think strategically about analytics and an organization must experiment first to develop this knowledge. An online course is not enough. If you are just starting out, choose a use case following the advice in the next point.

 

Tips include:

  • Overcoming cognitive impairment
  • Using minimal viable outsourcing
  • Moving with speed

 

Read the full article, 5 tips for a more ‘data-driven’ 2021, on christophedegreift.com.

 

 

When building a bicycle for his daughter, Azim Nagree was reminded of the importance of two key components of best practices: process and documentation.

Last week, my daughter turned 4 and I found myself, late at night, trying to build her new birthday bike. The task would have been made easier if the instructions were decent, but unfortunately, they were written poorly so I ended up just trying to figure it out myself. What should have been a one hour project ended up consuming 3 hours of my time, as well as most of my patience and sanity (why would part A connect to part F – doesn’t it make sense for A to connect to B?!?)

As I struggled with the joining the “G-Connector Bracket” to the “U-Slide” but making sure that the “Circle Washer” was in the right place, I realized how this same struggle applies to the workplace. When someone is faced with doing something for the first time, we oftentimes do not set them up for success – instead, we let them either figure it out on their own or rely upon the dissemination of tribal knowledge (i.e. they ask one of their peers who gives them verbal guidance on how to do that particular task).

 

Included in this article examples of:

  • Processes
  • Documentation
  • Implementation

 

Read the full article, How Building a Bike Reminded Me of the Importance of Scaling, on LinkedIn.

 

 

Tobias Baer explains why the lack of randomized testing hampers businesses and raises the costs of the COVID-19 pandemic. 

The lack of randomized testing again and again hampers businesses because it means executives need to make decisions half blind – and Covid-19 is no different, only that in the case of Covid-19, the cost of not doing randomized testing literally might run into the trillions of dollars.

Randomized testing is nothing new – and widely considered best practice in the business world:

 

  • Marketing executives should run A/B tests to make sure that ads and other marketing outlays actually influence purchasing decisions and isn’t wasted on customers who would have bought the product anyhow (or worse, even discourage buyers);
  • Banks and insurers randomly approve a small sample of credit or insurance applications rejected by their policy rules because otherwise they cannot know how many errors these rules make and if they lose any profitable business;
  • Online sellers use randomized pricing experiments to test how much more or less revenue and profit they would make by raising or lowering prices a bit.

 

Read the full article, Why Covid-19 illustrates once more the need for randomized testing: Paying the price for biased information, on LinkedIn.