But what about AI?

Before getting into this one, a disclaimer: we see AI as being immensely useful now, and in the future for a wide range of use cases. We are actively reviewing how we utilise AI to further improve our services. We are also aware of the potential risks for many businesses. This blog post explores why AI might not be the ideal solution for every company, specifically relating to business intelligence and dashboard initiatives. And with that…

AI is the hot topic at present.

If you have "AI FOMO" you're not alone; Many business are thinking about AI in relation to Business Intelligence and dashboard projects. We take a look at a few points that may ease your mind.

Return on Investment
Firstly, why does a business make any decision? Ultimately, because they want a return on their investment.

Businesses need to think about what that ROI looks like, and the tangible outcomes that an AI project would provide.

This will vary hugely from company to company. For some the complexity and cost of a project would be immense and the outcomes would not be significant. For others it would be a clear decision and a necessary project in order to stay competitive in a marketplace. This is individual to your business, industry, market, etc.

AI can be very effective at automating labour intensive tasks, however a lot of this can already be done with simple solutions that don't need AI. In addition, the insights AI can bring may be very beneficial for some companies, and not others. Each company needs to do their own analysis and act accordingly.

Data Quality, Availability and Integration with Existing Systems
We live in an imperfect world. Most businesses have incomplete, outdated and inconsistent data. This data is typically siloed within organisations and this is going to limit the effectiveness of an AI project. Many companies have outdated and legacy systems which add further problems when trying to integrate effectively.

For some companies, the process of collating data, checking the accuracy and validity, and adding that human intelligence as an overlay, is a very important process. That being said, for companies where data is more readily available, AI becomes a more attractive proposition.

Model Interpretability and Explainability
Many AI models are seen as "black boxes", meaning users can see the inputs and outputs, but the process by which the model arrives at its conclusions is opaque. This can be an issue in companies and industries where it is imperative that data can be trusted, and where stakeholders need to be specialists at interpreting and explaining the figures.

Again, this is not to say AI can't be effective for many companies in assisting that analysis. But as we see it for the medium term, many business stakeholders may be reluctant to trust insights without transparent explanations.

Troubleshooting Issues
What happens when something goes wrong? Who understands the model/data in order to flag that numbers are incorrect? Who is able to tell that the AI has hallucinated? How can you trace back the issue?

Again, these questions relate to trust in the solution and having stakeholders who truly understand the model and the data. These are risks that need to weighed up by each business.

Ethical and Bias Issues
AI models may inadvertently perpetuate biases present in training data. Businesses need to be cautious that this doesn't impact customer trust and brand reputation.

Cost, Resource Constraints, and Skills Gaps
Linked to ROI, AI implementations require investment in infrastructure, software, and talent.

For a company to implement effectively teams would need to be upskilled, there would be ongoing maintenance, data storage costs, etc. With a lack of AI-savvy BI professionals, there could be hiring challenges in the short-medium term.
These are potential risks that impact ROI.

Data Security and Privacy Concerns
At present many companies are concerned about security with AI handling sensitive data (e.g. GDPR). Each company will perceive this risk differently.

Evaluating the Value (ROI)
Coming full circle, back to ROI; For some companies effectively measuring the tangible outcomes provided by AI is going to be extremely difficult, for others it will be clear.

So in conclusion, there is no doubt that AI is going to have a massive impact on the world. For some businesses, industries, use-cases, it is going to transform the way we work forever. But for other businesses the potential return needs to be carefully weighed against the cost and potential risks outlined above.

For many Business Intelligence and dashboard projects, the benefits AI could bring may be relatively small and the cost and risks may be high. Each company needs to analyse and act accordingly.