The AI Hype Machine: Hype, Hope, or Headache?

ajustus Feb 26, 2025 AI, AI Solutions, Artificial Intelligence, Data Analytics

By Andrej Hudoklin, Executive Head of Data & AI

Why This Article?

I’ve been in the data & analytics business for over 25 years, working on interesting AI solutions across different industries for the last decade. And yet, after all this time, I still find myself facing the same challenge – explaining to business leaders what it really takes to implement AI properly and create real value.

The problem? The hype!

With the explosion of Generative AI, many believe AI is as simple as plugging in a model and watching it work magic. But the reality is far more complex. GenAI is just one part of AI – an exciting one, yes, but not the whole story. The real question is:

How do businesses go beyond the hype and actually make AI work for them?

That’s what this article is about – cutting through the noise and giving you a practical, business-first perspective on AI.

AI is everywhere, but is it actually working for businesses?

AI is like the ultimate buzzword. Everywhere you look, companies are racing to implement AI, vendors are selling it as the magic wand, and investors are throwing money at anything with “AI-powered” in the tagline. According to Gartner, AI adoption in enterprises has grown by 270% over the last four years.

But here’s the catch – most AI projects don’t deliver on their promises!

What is AI, Really? (Simplified)

AI, or Artificial Intelligence, is about teaching machines to perform tasks that typically require human intelligence. These tasks can range from understanding language and recognising patterns to making recommendations and even automating physical processes.

AI is not just one thing. It’s an umbrella term for various technologies, each with unique applications.

AI Has Been Around for Decades

The current hype makes AI feel like a brand-new phenomenon, but in reality, it’s been working behind the scenes for decades. If you’ve ever used Google Search, relied on a credit card fraud alert, or gotten a recommendation for a show on Netflix, you’ve already interacted with AI.

Imagine AI as a diligent assistant that has quietly been working across industries long before ChatGPT made AI feel “human-like.”

Manufacturing: AI-powered robotics and automation have optimised assembly lines for years.
Retail & E-Commerce: AI has long powered demand forecasting, warehouse automation, and personalised recommendations.
Finance & Banking: Fraud detection systems and algorithmic trading have relied on AI-driven pattern recognition for decades.
Healthcare: AI has helped analyse medical scans and optimise hospital operations long before the rise of ChatGPT.

So why all the hype now? Because AI is finally stepping out of the background, is more powerful, more visible and accessible to everyone.

Why is AI Different Now? The Big Shift

Several key factors are making AI more disruptive today than ever before:

Computing Power & Cloud Infrastructure: Advances in GPUs and cloud computing have enabled faster AI model training and deployment at scale.
Big Data Availability: AI thrives on data, and businesses now have access to vast amounts of structured and unstructured information.
Breakthroughs in Deep Learning & NLP: Models like GPT-4 and DeepSeek can now generate human-like text, create images, and even automate creative work.
Low-Code & No-Code AI Tools: AI is no longer limited to data scientists. Business users can implement AI solutions without deep technical knowledge.

Why the “ChatGPTs” of the World Created a Misleading Perception

One of the biggest misconceptions about AI today comes from the rise of generative AI models like ChatGPT.

These models make AI feel easy, instant, and ready to use – just type a prompt, and you get a response – and most people just take it as 100% truth.

Generative AI (like ChatGPT): Already trained on massive datasets and available as ready-to-use services, making AI seem effortless.
Traditional AI & Machine Learning: Requires months (or even years) of data collection, model training, and integration before it delivers business impact.

This difference is critical for business leaders. AI-driven fraud detection, predictive maintenance, or supply chain optimisation cannot be deployed overnight. They require strong data, alignment with the business, and continuous iteration. Expecting every AI project to be as fast as ChatGPT sets unrealistic expectations and leads to failed AI initiatives.

The Two AI Races: Innovation vs. Tangible Value

Right now, we are witnessing two AI races happening simultaneously:

Tech Vendors: Companies like OpenAI, Google, and DeepMind are pushing AI capabilities to new heights, constantly releasing new models that are smarter, faster, and more capable.
Businesses: On the other hand, many organisations are still struggling to implement AI in ways that deliver real business value.

While vendors are in a race to push the limits of AI, businesses are facing a different challenge:

How do we move beyond AI experiments to actually making money or improving efficiency?

The “AI Outcomes Race” – Setting the Right Pace

Businesses don’t need to keep pace with AI research labs; they need to focus on outcomes. Organisations should set their own AI pace based on:

Maturity of their data infrastructure: Do you have clean, structured, and accessible data?
Internal AI expertise: Do you have the right people to implement AI effectively?
Business problem alignment: Are you solving real challenges, or just chasing AI for AI’s sake?

McKinsey Insight: AI adoption is moving past the hype phase into a period where businesses must focus on execution, scaling, and driving real returns. Those who prioritise operational integration over experimentation will benefit most.

Cutting Through the AI Hype: What Actually Works

So, how can businesses avoid falling for AI hype and instead drive real impact? Here’s a pragmatic approach:

1. Solve Real Problems, Not Just “Do AI”

AI should be a tool to enhance existing processes, not a project for innovation’s sake.

Start with clear objectives!

2. Focus on Data Quality Before AI

If your data isn’t AI-ready, fix that first. Invest in data management, master data governance, and integration.

3. Start Small, But Plan for Scale

Rather than jumping into complex AI, begin with focused use cases. But always have a roadmap for scaling successful pilots.

4. Engage Business Teams Early

AI adoption isn’t just an IT project. Sales, marketing, operations – everyone must be involved from the start to drive success.

The Future of AI in Business: Hype or Hope?

AI is neither a miracle nor a disaster; it’s a tool.

Companies that approach AI pragmatically, with a focus on data, business impact, and adoption, will see real returns! Those chasing hype will burn budgets with little to show for it!

If you’re thinking about AI and not sure where to start, let’s connect and figure it out together – data@smollan.tech

Getting there: Key success factors for retailers

ajustus Jun 10, 2022 Data Analytics, PredictRetail, Retail

Success Factors: Using data analytics to become truly customer-centric is now within reach for retailers, but they need to adopt a focused approach with a clear plan − or risk failure.

By Tanya Long, Chief Operating Officer, Argility Technology Group

This article builds on the argument made in my two previous articles that data analytics, and especially the emerging capability to perform predictive analytics, offer retailers a genuine opportunity to reinvent themselves as customer-centric organisations, deepening customer loyalty and enhancing profitability in a volatile marketplace.

As always, though, nirvana isn’t reached in a day, and there are numerous pitfalls along the way. In this final article, I want to look at how to successfully make the move to becoming a data-driven organisation, rather than one that relies on intuition.

Based on my experience in the industry, the following key success factors need to be integrated into the plan:

Avoid grand projects and keep an open mind. It’s imperative to move away from grandiose projects led by IT, to which some corporates appear addicted. A much better approach is to focus on iterative projects led by the business that are less risky and allow the organisation to learn as it goes. This in turn means being open to changing tack as the data dictates − we have to be open to what the data is telling us, and act accordingly.

Understand data needs. It is also important to recognise that most retailers already have all or most of the data they need − don’t waste time and money trying to get at the data you want (or think you want).

In line with my advice in the first point − take baby steps, survey the existing data and work with that, at least in the beginning. That said, it is vital to take the necessary steps to ensure the data to be used is clean and reliable; the old “garbage in, garbage out” mantra holds good.

Set goals − actionable insights are key. A related point is that it is easy to fall in love with the data and embark on data-related projects that are interesting but that don’t deliver any real benefits. A disciplined approach is vital, and a data project must be aimed at generating insights that are actionable.

Knowing everything there is to know about a customer segment for itself is ultimately counter-productive. A better approach is to identify what information is needed to reach a strategic goal or make a better decision.

A disciplined approach is vital, and a data project must be aimed at generating insights that are actionable.

It may be interesting to know a customer’s preferences in the abstract, but it is only valuable when aiming to leverage that knowledge in order to get the customer to buy while they are in the store or on the website. An important element is speed: the actionable insight needs to be generated rapidly so that action can be taken in real-time.

Change the corporate culture. If the retailer is going to become customer-centric, just acting on insights is not enough. The whole organisation has to change its focus or predisposition − everything everybody in the company does or says must be founded on the customer.

For that to occur, a vital first step is for everybody to understand the direct link between customer-centricity and the bottom line (and thus, in turn, on benefits, job security and the rest of it). Segmenting customers in terms of their lifetime value to the company, and how much it costs to acquire and then service them, will help to make the business case for data projects. It’s particularly important that the CFO is involved so the return on investment for specific data projects is well understood.

Another central part of the new culture is a shift towards making decisions based on evidence only, not on emotion.

Use the growing understanding of the customer intelligently. My main point here is to ensure the customer experience is well designed in light of this knowledge and is constantly being refined via a feedback loop that links into strategy and operations. Every interaction with the customer, including those undertaken by software, must be linked to data.

Leadership must be on board. A profound change like this will not succeed without strong leadership. It is a vital step to get the leadership team on board and motivated.

Pay due attention to talent management. The customer-centric retailer needs staff who have the right skills. Specifically, this means access to specialists like data scientists but generally an ability to solve problems and follow logic becomes critical.

Make sure the technology is in place. As should be clear by now, technology is not a silver bullet, but it does need to be in place. Storing, processing and analysing fantastically large amounts of data depends on technology, and the building blocks must be solid.

Develop a pilot to demonstrate the value of predictive analytics to the organisation. There’s nothing quite so powerful as a successful project. It remains important to make the case for data’s role in helping the company to become more customer-centric and why that would be beneficial but give the project the best chance of succeeding by carefully designing a pilot project that can be relatively quickly deployed to show what you mean.

Retailers face a set of tough challenges in the short- to medium-term. Only those that harness the power of data to help them understand the challenge and craft effective responses will survive − as will those that understand this is not a quick dash, but a journey undertaken with an open mind.

Published: IT Web

Tanya Long, Chief operating officer, Argility Technology Group.

Long has 30 years of industry experience. Her career in the IT sector started in 1988 in IT support for point-of-sale solutions. She moved formally into software development with UCS/Argility in 1989 as a developer and progressed through to team leader, account management, project and development management roles, which led her into various industries and corporations.

In 2017, Long returned to Argility (having previously worked at the company in a technical capacity) as human capital executive, where she reunited her retail, IT, leadership and HR knowledge to drive her zeal for transformation.

In her capacity as COO, Long is responsible for overseeing operations, with a specific focus on human capital, sales and marketing, and ensuring the company culture and vision shows up daily for customers through an engaged technical team of experts.

Is customer-centricity in sight for retailers at last?

ajustus May 31, 2022 AI, Customer Centricity, Data Analytics, PredictCustomer

Data analytics—and particularly predictive analytics—can give retailers a significant edge in an industry that is rapidly changing.

By Tanya Long, CEO, Argility Technology Group

In my first article, I looked at data and the opportunity it offers to businesses and other organisations in general terms.

I argued that by learning to use data effectively, companies could not only understand the past better, they could also develop strategies with a much higher chance of success because they are based on extrapolations from real facts and not feelings or gut instinct.

I also made the point that retail is a sector in which this kind of data-driven decision-making holds particular promise. In this article, I want to explore why this should be, and what the benefits are.

First, let’s take a look at some of the retail trends and how they are in turn driving the shift to a data-centric approach.

As we emerge from the COVID-19 pandemic, it’s becoming clear that the move online which we have seen building over the past decade or more has considerably strengthened. Four of the five most valuable retail brands in the world are now online merchants – Amazon, Apple, Google and Microsoft, with the fifth, Visa, also heavily involved in the online world.

Real-world shopping remains important, of course, but the enforced move to digital channels during lockdown has had an effect: in the past six months, 65% of consumers made a purchase from a physical store – but 78% had purchased from Amazon, 45% from a branded online store, 34% from eBay and 11% from, of all things, Facebook.

The move online looks irreversible, especially when one considers these amazing figures relating to each generation’s preference to buying goods from a physical store: Baby Boomers 31.9%, Millennials 31%, Generation X 27.5% and Generation Z 9.6%.

In other words, Generation Z represents a step change, and it’s hard to imagine that the cohort that follows it will not continue along the same trajectory.

Blurring boundary

Retailers can draw two major conclusions from these and similar findings. The first, obviously, is that they had better have an online strategy; the second is that the line between the virtual and real is blurring in consumers’ minds.

By this, I mean that consumers not only want to move seamlessly between real and virtual channels and have the same experience on all, they no longer see a reason why the convenience and personalisation they are already getting online should not be replicated when they visit the store.

In the end, the retail experience can slowly become a collaboration between retailer and customer.

Among the multiple implications of all this, data emerges as a common thread. The digitalisation of the retail environment online means that a considerable amount of data is generated, which can be used to refine the customer experience and to improve the business processes.

But the same information can be used by retailers with real-world stores to adapt steadily their in-store environment by leveraging the lessons of online.

For example, beacons can be used to make the in-store customer experience much more personalised, mimicking to some extent the online experience. Beacons are essentially communication points that prompt customers to log in via their mobile devices to receive useful information such as where to find a certain item, or what specials are running.

Mobile phones have a crucial role to play in the digitalisation of real-world shopping and in bringing the virtual and real together. Research shows that half of shoppers use their phones as an in-store research advisor.

Mobile devices can also be used to offer more convenient, highly personalised ways to pay—it’s a fact that a large percentage of customers (almost 30%) abandon the purchase during the payment phase, so improving it makes a lot of sense.

A more sophisticated approach would be to allow customers to upload their shopping list so that the most efficient routing through the store can be prompted via the beacon, offering special deals at appropriate points.

As in the online world, this data can be used to refine the whole environment, making it more efficient and profitable for the company, and much more engaging for the customer.

In parenthesis, I want to emphasise the important role that data can also play in running the business, thus indirectly contributing to serving them better. The use of digital technologies across the supply chain is another topic, but again will generate a lot of useful data that can be used to improve the business in all sorts of ways.

Another avenue to explore would be the BOPIS (buy online, pick up in store) model, which some customers prefer.

Becoming customer-centric

All of these digital innovations have one thing in common: they generate large amounts of data about customers that offer retailers a golden opportunity to learn more and more about how to serve them better and deepen their loyalty.

In the end, the retail experience can slowly become a collaboration between retailer and customer.

At a very practical level, data analytics will enable much more accurate segmentation, moving away from the crude living standards measure (LSM) to customer lifetime value (CLV), customer acquisition cost (CAC) and cost to serve.

The latter three measures allow not only a more granular understanding of customers but also of their value/ potential value to the business, and so how much to invest in them.

Once one starts thinking about data and how it can help retailers truly transform the way they do business, to put the customer at the centre of everything, it’s clear that the sky is the limit.

However, as always, there are some caveats to bear in mind and I will look at those in my final article.

Published: IT Web

Tanya Long, Chief Operating Officer, Argility Technology Group.

Boards of Directors have spoken! What challenges will data and analytics leaders face?

ajustus Jan 13, 2020 AI Strategy, Data Analytics, Data-driven Decisions, Digital Transformation

By Andrej Hudoklin, Executive Head of Data and AI: Europe

From accelerating digital business to much-needed cultural changes, data and analytics will face serious challenges in 2021. Given that every digital business moment is powered by, or held hostage to, data and analytics, it is no wonder that chronic issues have become more acute.

3 Trends that will impact data and analytics leaders in 2021:

*The Heat Is On – Boards of Directors Place analytics and AI as the No.1 and No. 2 Priorities; source: https://www.gartner.com/en/publications/data-analytics-top-priorities-for-it-leadership-vision-2021

Chief Data Officer will embed data and analytics in business strategy. They are responsible for guiding decision-makers within an organisation using data insights.
Data literacy increases. It is about developing a common understanding of what goals and outcomes are important across the organisation.
Data and Analytics will change management. You must base your data and analytics organisation on collaboration, cooperation, and problem-solving. Let us spread the word about how data and analytics can help drive business outcomes.

Data and analytics leaders will face many challenges in 2021, such as embedding data and analytics into business results and addressing perceptions of data and analytics within the organisation.

The best actions for data and analytics leaders are to create a data-driven culture, develop a data and analytics strategy, and stand up a data governance program. They must implement adaptive data and analytics governance. Adaptive data and analytics differ from traditional approaches that tended to be IT-driven and focused on data and standards. Adaptive governance uses business outcomes to prioritise the work and helps you govern the least amount of data with the greatest business impact.

Leaders must answer 4 crucial questions:

Build a data and analytics strategy and operating model that can support and scale, with a measurable business impact?
Apply trusted data and analytics in their decision-making?
Create a business case to support data and analytics monetisation and innovation?
Develop a data and analytics-driven culture, literacy, and employee behaviour?

So, the heat is on. According to Gartner, the top game-changer technologies to emerge stronger from the COVID-19 crisis are Analytics, Artificial Intelligence and Autonomous Things.

Leading in 2021 will look different from leadership in 2020, but understanding and planning for unknowns and continued disruption across the IT organisation is critical for moving forward.