The threat of AI on traditional BI and Analytics tools
AI also changes software domains like business intelligence (BI) and analytics. Earlier in my career, I spent over 15 years as head of development for two software companies, building and delivering business intelligence, data warehousing, and analytics solutions.
BI and analytics vendors have long marketed their solutions as easy to use, but we must ask ourselves how effective they have been for end users. I would be tempted to say that investments have never paid off, as users found the solutions too hard to use. I have seen this too often during my career, and no matter how much training is given to end users, some tools are just too complex to use.
I ran across a fascinating article by Joe McKendrick (a contributing writer at ZDNet). In it, he discusses AI's impact on traditional business intelligence tools and questions whether "artificial intelligence leads to the decline or the rebirth of business intelligence."
Software vendors like Microsoft have invested heavily in AI, embedding Copilots into their solutions, such as Power BI, Microsoft's flagship dashboard and analytics tool. Other vendors, such as Qlik, follow in their footsteps by adding natural language processing (NLP) via generative AI to their tools. This allows users to interact with data using everyday language. In addition to graphics, adding AI will provide additional capabilities, such as presenting complex data in an easy-to-understand format.
The rise of AI is grounded in the development of large language models (LLMs), which have helped software vendors amend their solutions to include structured and unstructured data. According to Chida Sadayappan, the managing director of Deloitte Consulting, LLMs "enhance data interpretation, improve decision-making, and automate processes, allowing organizations to derive deeper insights and create more value from their data."
However, it is not just the end-user-facing analytics solutions that will be impacted by AI; the traditional extract, transform, and load (ETL) will also change the landscape of how information is integrated into the analytics/data warehousing solutions. It was an eye-opener for me when I used ChatGPT for the first time to "massage" a CSV (comma-separated file). I could not believe how easy it was to use natural language processing (NLP) to change the file format to what I needed.
The role of some of the coders will also change with AI-driven analytics tools. According to the ZDNet article, the business development manager Quang Trinhat Axis Communication is stating the following:
"We will move from static reports to dynamic data products, providing real-time, actionable insights embedded directly into workflows to drive decision-making at every level."
GenAI promises that all analytics tools should allow users to ask questions in natural language and provide easy-to-understand descriptive answers. Analytics vendors have long talked about data "democratization," but I don't think that has really happened regarding how it was marketed. Those of you who have worked within the analytics and BI domain know that technical skills, such as SQL and an understanding of data integration, are needed.
But one thing has not changed. If you have bad data, your results will be bad. So, the famous term "garbage in, garbage out" will still apply to AI-driven analytics and BI solutions. Sadayappan states the following of how LLMs and modern platforms can help:
"Modern data intelligence platforms with LLMs help by facilitating seamless integration, enhancing data quality, and automating insights, thus providing a comprehensive view of customers and operations."
If you represent an organization that uses traditional analytics and BI solutions, you must ask yourself whether your software vendor will be relevant in the future. The pace of technology is moving faster than what potentially smaller vendors can follow and absorb. The question is whether there will be a time when maintaining traditional solutions that have already been deployed will be more costly than buying and applying new AI-driven solutions that are easier to deploy and maintain and provide an easy-to-use analytics interface driven by natural language processing (NLP) to end-users.
This would be an exciting and scary time if I were still running the two independent software companies and product development. The question would be ensuring I invested in the right technologies at the right time. Software vendors need to be careful how they apply technologies. Some of the ones that I know from two years ago and built solutions using AI, the same technology has become part of the large vendor AI platform. As of this, their technology has become irrelevant overnight.
As I have written multiple times in previous LinkedIn newsletter articles, AI is impacting each and every one of us in one way or another. We who work in technology need to learn and absorb new things continuously. The ones that are end-users of AI technology need to be trained on how to use AI efficiently. No one, I mean no one, can escape the impact of AI.
I would love to hear your thoughts on how you see analytics and BI evolving with AI. Do you think users will benefit from it, and if so, what gains do you see happening?
Yours,
Dr. Petri I. Salonen
PS. If you would like to get my business model in the AI Era newsletters to your inbox on a weekly or bi-weekly basis, you can subscribe to them here on LinkedIn https://www.linkedin.com/newsletters/business-models-in-the-ai-era-7165724425013673985/