In the race to digitize big data, business use cases are too often neglected. Here are ways to make big data and AI projects work for your company.
Image: invincible_bulldog, Getty Images / iStockphoto
Companies may be better off in their artificial intelligence (AI) and big data business projects if the goal was not so much data-driven but business-driven.
I am thinking of a thought-provoking article in the Harvard Business Review of February 2019 by Randy Bean and Thomas H. Davenport entitled Companies failing in their efforts to become data-driven. The authors say that companies do not become data-driven, although they aggressively pursue AI and big data. The article also said that 77% of managers believe that the acceptance of AI and big data by companies is a major challenge. Even when companies are data-driven, adoption can be daunting.
Here is an example: an IT department of a large financial company that I work with, digitizes and indexes all documentation. The project manager told me that the goal of the project is to get rid of all the paper and to digitize it in a central, searchable repository of data to which applications and systems within the company can connect.
From an IT point of view it is easy to see the logic; however, it strikes me that no one at the company has identified the business processes that can be optimized by this data storage that is being digitized, or considering the analyzes and business insights that the company can derive from this data that would become searchable.
In short, there is no strategic plan or business direction, except for the digitization of the data. The product is really ‘data-driven’ because it produces data, but it is not business-driven, and the company does not position itself to see impactful business results for months.
SEE: Data analysis: a guide for managers (free PDF) (TechRepublic)
Many organizations are going through a similar battle. An inherent challenge with many of these technologies is that you simply cannot connect them as a traditional transaction system and prepare invoices, inventory reports and purchase orders. Instead, most AI, analysis, and big data projects undergo a series of iterative processing and testing until there is a consensus among IT, data scientists, and business users that the results these projects deliver are “true.” Along the way, some of these projects don’t make it and others don’t, and there is a risk that the project work you do will turn into a data-driven exercise that isn’t business-like.
So how do you stay sharp and avoid the pitfall of becoming a data-driven organization instead of a business-driven company? Follow these three tips.
1. Identify your business cases before you purchase technology
This is very important. If your company cannot visualize measurable income, cost reduction, work environment or customer satisfaction from the analysis, IoT or digitization, you do not have to spend your budget.
2. Exit the pilot mode
In recent years, analytics, big data and AI projects have been given a mulligan when it came to producing measurable business results because they were new technologies implemented in pilot experimental modes, provided that the projects would work or not. This honeymoon is over. Management expects big data projects to deliver tangible business results, just like transactional data systems.
3. Communicate project status and methodology
Even if big data projects are now viewed in mature mode and are expected to produce results, this does not change the fact that they are more difficult to implement than transactional data systems.
Big data analysis and AI use algorithms that must be constantly refined until they reach an accuracy of at least 95% before being put into production. That is why there is an iterative approach to big data projects and testing until you reach an acceptable level of accuracy. This repetition can give management the impression that a big data project is not being carried out properly due to continuous changes and retests. That is why it is essential for CIOs and big data leaders to explain the differences between transactional and big data testing methods and projects in plain English, so that management understands the process.
Remember your goals
Big data projects must always be business driven from the start. It is not sufficient to collect, manage and process data for the sake of data, in the expectation that business usage cases will just follow. Keep business goals in mind and you keep focusing on what’s really important.
Big Data Insights newsletter
Master the basic principles of big data analysis by following these tips from experts and by reading insights about innovations in data science.
Delivered on Mondays