HICO-Group

The BARC Spotlight Study GenAI in Business Intelligence & Analytics sheds light on the uncertainty in the industry with regard to the use of GenAI: concerns about data security, a lack of technical expertise, poor data quality and the complexity of integrations dominate the picture. There is no clear focus when it comes to implementing the technology; instead, the various approaches are more of a patchwork quilt: is the industry overwhelmed by GenAI?

As an independent consulting and research institute, the “Business Application Research Center”, BARC for short, has specialized in examining the software market for business intelligence, data & analytics, corporate performance management and enterprise software since the mid-1990s.

The spotlight study GenAI in Business Intelligence and Analytics was conducted by BARC on behalf of HICO and Qlik – and as the title suggests, it sheds light on the adaptation and use of GenAI in current events. For the survey, 238 decision-makers in the field of business intelligence in various industries were asked about their assessment of the adoption status, benefits, risks and best practices around GenAI in BIA. 

For its part, the picture that presents itself casts a spotlight on an insecure guild.

Small minority way ahead

9% in implementation phase for GenAI in BI

Data experts, a cohort to which anyone who wants or needs to generate insights from data turns, are by no means a homogeneous knowledge community, especially with regard to GenAI. On the contrary, the BARC Spotlight Study paints a picture of a highly insecure guild.

For example, the gap between progressing early adopters and laggards in the adaptation of GenAI is tellingly contradictory: a majority of 73% of respondents state that they are experimenting with the implementation of GenAI in BI & Analytics or are already in an advanced implementation phase.

At the same time, a comparatively high 39% stated that GenAI was not an issue in their BI organization and a full 60% that they were still in a discussion phase. That makes a total of 170% – how can that be?

This becomes a clearer when you look at the previous BARC study (see below) from spring 2024. In this study, a distinction is made between data leaders and data laggards.

Although this makes the statistical picture much clearer, it underlines all the more the enormous differences in the way GenAI is used, as it becomes clear that the introduction of GenAI in the BIA is – despite all the prophecies of doom – still in the early stages.

Only a small percentage of companies are actually in the process of implementing GenAI in their BI practice. Around 29% of respondents are in a discussion phase with their teams, 9% are evaluating possible use cases and 22% are still experimenting with GenAI.

Overall, the BARC authors summarize that only around 9% are in any kind of implementation phase.

Why is that the case? And is that a bad thing?

In the spotlight: data (in)security

The first question can be answered, at least in part, with the usual resentments: risks such as data protection, qualification deficits, compliance, data quality and a certain bias form the basis for considerable concerns.

The study does not answer whether these are justified or not. In our experience, it depends entirely on the case. For example, the much-cited bias against new technologies seems to be a widespread phenomenon that is often explained by internal organizational demographics. Whether these are genuine findings or rather statements motivated by company policy will have to be worked out elsewhere.

More important – and more tangible – are the uncertainties with regard to data protection and compliance issues as well as data quality, which is therefore the most important prerequisite for even starting a meaningful implementation of GenAI in the business intelligence environment. “Shit in – shit out”, the central aphorism of all data-supported processes, is particularly true when dealing with generative AI.

Churches, sows, villages and a pragmatic recommendation

While people in the phlegmatic Rhineland often prefer to leave the proverbial church in the village, more activist areas (LinkedIn) tend to chase trends

Both appears to be misguided. Kevin Petrie, lead author of the BARC study, puts things in perspective and points out that the market for GenAI in the BIA is still at an early stage of development. The fact that early adopters are leading the way here is one of the great constants of economic innovation cycles.

What distinguishes the leading companies in terms of GenAI in BI in particular is that they already have an advanced AI/ML infrastructure and are also more satisfied and optimistic about the benefits of GenAI.

Adoption trends indicate that companies with strong data science programs and larger organizations are more likely to adopt GenAI.

Smaller companies see GenAI as a way to expand their analytics capabilities, while larger companies focus on increasing the efficiency of existing projects.

Does it all sound quite interesting?

We think so too! So why not make an appointment right here and discuss everything else with the HICO experts?