Date: 13th January 2026 

Speakers: Dr. Shubhadeep Mukherjee (MSc Business Analytics Programme Director) & Peter Ludgate (Senior Student Recruitment Advisor) 

Introduction 

Peter Ludgate: Hello there. Good afternoon, good morning, or good evening, depending on where you are joining from, and thank you for joining us for today’s introduction to Aston Online’s MSc in Business Analytics programme. My name is Pete, and I am the Senior Recruitment Advisor here at Aston Online. 

Joining us today is Dr Shubhadeep Mukherjee, the Programme Director. How are you doing, Shub? 

Dr Shubhadeep Mukherjee: Not too bad, thank you. Thank you for introducing me. Hello, everyone, and hopefully we will have a really good session with some exchange of ideas and discussion. 

Peter Ludgate: Wonderful. 

About Aston Online and the MSc Business Analytics 

Peter Ludgate: Before we move into the programme itself, I want to say a little about Aston Online more broadly. 

We are a Triple Gold-rated university according to the Teaching Excellence Framework. We are ranked in the Top 5% of global universities according to the QS World University Rankings. We were also voted University of the Year for Student Success last year according to the Daily Mail. 

Now, specifically on the MSc in Business Analytics, the programme is accredited by the Institute of Analytics. It is delivered by our triple-accredited business school. We are also ranked second in the UK and in the Top 10 in Europe for Business Analytics according to Eduniversal 2025. 

So, yes, we are highly regarded, as I have already spent some time saying, and those are the bragging rights for the university. 

In terms of the practical details, the total fee for the programme is £12,500 for the entire programme. That can be paid in monthly instalments if you wish. The total duration of the programme is two years, and it is studied part-time. 

With that comes the study recommendation. We do not absolutely mandate it, but we recommend that you commit between 15 and 25 hours of study each week. That may fluctuate depending on the task at hand. If you are completely new to business analytics, or analytics more generally, you may want to allow yourself a little more time at the start. 

During the programme, you will be learning analytical tool proficiencies, data visualisation and descriptive statistics, predictive analytics, prescriptive analytics and modelling, and importantly, project management. 

Entry requirements 

Peter Ludgate: If this is sounding good to you so far, you are probably wondering how you can join. 

We look for applicants to hold a 2:2 bachelor’s degree, a lower second-class honours degree, in a quantitative, technical, or scientific field, such as computer science, finance, or something similar. 

International students must hold an equivalent qualification and must also meet our English language requirements. That usually means holding something like an IELTS or TOEFL qualification. 

If you do not currently hold one of those, or if the qualification is older than two years, you can still speak to us. We may be able to offer an alternative English language assessment so that we can still assess your capability. 

If you do not meet the standard academic requirements, we can also take a more holistic approach to the application. For example, if you do not hold a bachelor’s degree but have spent a number of years working in business and specifically in analytics, we would still very much like to hear from you. 

So, if you are unsure, that is actually a very good reason to move forward and have a conversation with us. I would recommend booking a call with me or a member of my team, and we will be happy to assist. 

Programme structure 

Peter Ludgate: Now let us look at the structure of the programme before I hand over to Shub. 

There are six modules in total, worth 180 credits altogether, and the programme runs over two years in two stages. Stage one is taught, and stage two is dissertation. 

There are four modules in stage one and two dissertation modules in the second stage. 

Each module is worth thirty credits, and each module lasts 16 weeks. 

That is broken down further into 12 weeks, in which you complete your learning and assessments, followed by at least a four-week break between modules. That break is when tutors assess your work, so we are not monsters, and we do want you to have a well-deserved rest. 

So, you will have study periods across the year with breaks around Christmas, Easter, and summer, and that pattern repeats across the two years. 

Once you complete the taught modules, you then move into the dissertation stage. This is where you begin to put together something like a research proposal. That final project can take a few different forms. It could be a consultancy project, a business plan, or an industry- or theme-based research project. 

Then, at the very end, you complete your final project, receive your award, and are invited onto campus for graduation, with cap and gown, friends and family in the audience, and perhaps a few tears as well. Maybe I am projecting a little there. 

Programme outcomes include designing solutions and evaluating projects, assembling descriptive statistical methods, designing analyses of performance, and designing and evaluating predictive modelling solutions for business problems. 

So really, without wanting to oversimplify it, it is a very comprehensive programme covering everything central to business analytics, and it will put you in a very strong position for both future roles and current opportunities you may already be considering. 

Case study 

Peter Ludgate: That is it from me for now. I will join back in later. 

Shub is now going to take over and talk us through the Latte Launch Backlash case study. Please feel free, Shub, to tell me when to move to the next slide, and I will do that for you. 

Shubhadeep Mukherjee: Absolutely. Thank you, Peter. That was a brilliant introduction to all the things we offer in Business Analytics at Aston. 

I have been the Programme Director for the programme for the past couple of years, and today we want to talk about one of the more practical business problems and how we might look at it from a business analytics angle. 

As Peter explained, if you look at the programme outcomes, the taught journey begins with what we call descriptive analytics, which is about looking at the data and describing it. Then we move into predictive analytics, where we use existing data to make forecasts or predictions. After that we move to prescriptive analytics, where we consider specific scenarios and optimise outcomes. 

Across all of these, there is also the analytical and technical tools element, which supports the whole process. 

So, if you think about stage one of the programme, it is a bit like learning to stand up, then walk, and then run. 

Today, what we are going to look at is more of a descriptive and predictive problem. We are not going to use actual tools in this session because this is more of a primer or introductory webinar, and the case study we are using is called The Latte Launch Backlash. 

What we are trying to do here is look at a specific business problem and then see how data analytics can be used to solve it. 

The Latte Launch Backlash: The business problem 

Shubhadeep Mukherjee: What is the problem? 

This is a very common issue in modern business because of the rise of social media. A brand can be attacked very quickly online, and that can lead to revenue loss and reputational issues. 

This does not just apply to large companies. It can apply to individuals too. If you are a content creator, for example, or someone building a personal brand on YouTube, Instagram, or TikTok, then you will know that something you do may not resonate well with a segment of your audience, and suddenly your subscriptions or engagement start dropping. 

This is simply the environment in which we operate now. We are not here to judge whether that is good or bad. We are simply acknowledging that this is the reality of the business environment. 

So, in our case, we have a popular coffee brand called Bluebox. It is a popular chain with a growing app user base. They launch a new pumpkin spice latte. If you are in the UK, you will know that pumpkin spice drinks are a very seasonal thing in late autumn and early winter, and many coffee companies release them. 

So Bluebox launches their new drink with a large marketing campaign and in-app promotions. They do all the usual things through Instagram, Twitter, TikTok, and other channels. 

However, soon after launch, the hashtag PumpkinSpiceRage starts trending. 

Now, imagine you are the marketing manager. You have put together a large campaign, spent money on it, and feel very positive about the launch. Then suddenly you wake up and see this negative trend online. The immediate question becomes: Why are customers not happy? 

Now, if anyone would like to write in the chat or say out loud what they think might have gone wrong here, feel free. 

Discussion of possible causes 

Shubhadeep Mukherjee: One suggestion is that social media engagement may have been mistaken for purchase intent. Yes, that could absolutely be one possibility. High engagement does not always translate into actual purchase behaviour. 

But here we seem to have something more active going on. Customers seem to be unhappy in a more direct sense, because something like PumpkinSpiceRage is trending. 

Would anyone else like to suggest a possible reason? There are not really any right or wrong answers here. On the internet, almost anything can go wrong. 

Another suggestion is that there could be a pricing issue, or that competitors’ seasonal launches may simply have been more appealing. Yes, that is also a very valid suggestion. 

So now the question becomes: Where do we start the investigation? How do we separate the signal from the noise? 

Framing the problem as business analytics 

Shubhadeep Mukherjee: Suppose you are the marketing manager. You notice that your app ratings have gone down suddenly. Perhaps you were sitting at 4.5 and suddenly you are down to 3.9. There is also a great deal of negative activity happening across different social platforms. 

Where do you begin? 

This is where business analytics comes in. 

And I want to make something very clear here. Business analytics is not just a group of people doing very high-end mathematics or proving statistical theorems. 

At its heart, business analytics is about taking a business problem and helping the right team make the right decision. 

Every organisation has different teams: finance, operations, marketing, sales, HR, and so on. Business analytics is a toolkit that can be applied across all of those teams to understand what is happening and help determine what should be done next. 

So, the goal is not simply to do complicated coding or very advanced modelling for the sake of it. You can certainly do those things, and we do teach the technical aspects, but that is not the central objective. 

The central objective is to identify a real business problem and use analytical skills to move toward a practical solution. 

In this case, the problem is straightforward: Why is customer sentiment dropping after a major campaign, and how do we address it? Can we use public feedback to understand the key issues? 

Unstructured data and why it matters 

Shubhadeep Mukherjee: Now we come to one of the really important parts of the discussion. 

When people think about business analytics, they usually think about numbers. They imagine spreadsheets, calculations, means, medians, and so on. 

But in reality, most of the data around us is not structured numeric data. 

Estimates suggest that around 80 to 85% of data is unstructured, and only around 10 to 15% is structured. In this particular business problem, what are our data sources? Tweets, app store reviews, Reddit threads, and similar content. These are not numbers. They are text. 

That means the nature of the modern business analytics problem is very different. People are communicating through text, memes, video, audio, screenshots, and all kinds of unstructured information. But the challenge is that almost all meaningful mathematical or statistical analysis has to happen on numeric data. 

You cannot calculate an average directly on a paragraph of text. You have to first convert that text into some form of numeric structure. 

So, our first major analytical step is to convert non-numeric data into numeric data. 

Technically, we can say that we are converting unstructured data into structured data. 

Pre-processing text data 

Shubhadeep Mukherjee: So, what does that mean in practice? 

We start with our text sources: tweets, app reviews, Reddit comments, and so on. Then we apply pre-processing. This might include removing things like emojis, URLs, or HTML tags. We then tokenise the text, which means splitting sentences into individual words. We remove stop words, which are common filler words such as “the” or “and” that do not contribute much meaning. We also lemmatise words, which means reducing them to their root form. 

For example, if a post says, “Bluebox ruined fall with the new pumpkin spice, app crashed again,” then after pre-processing we might simplify that to something like “Bluebox ruined fall new pumpkin spice app crash.” 

So, we remove unnecessary noise and keep the essential content. If you do not fully understand all of that on first hearing, that is absolutely fine. The key point is that before we can analyse the data, we have to transform it into a form that can be analysed. 

Sample tweets and what they tell us 

Shubhadeep Mukherjee: Now let us look at some sample tweets and app store reviews. 

Suppose one post says, “Love the new pumpkin spice latte.” That is clearly positive. 

Another says, “App keeps crashing every time I order.” That sounds more like a technical issue. 

Another says, “Pumpkin spice is too expensive.” That reflects pricing dissatisfaction. 

Based on those examples, even in a small sample, we can already begin to see themes emerging. Some people like the flavour. Others are frustrated by the app. Others feel the price is too high. So the challenge is to take all of these comments and convert them into something we can count and analyse. 

Converting text into numeric form 

Shubhadeep Mukherjee: One very simple way to do this is to build what is effectively a presence-or-absence table. 

Imagine we label our tweets T1, T2, and T3. Then for each important word, such as “love,” “pumpkin,” “spice,” “app,” “crash,” or “expensive,” we note whether that word appears in each tweet. 

If the word appears, we put a 1. If it does not, we put a 0. So, if tweet one contains “love,” “pumpkin,” and “latte,” those become ones, while words not present become zeros. 

What this gives us is a very simple but very effective way of transforming text into numbers. 

And once we have numbers, all the tools become available to us. We can count word frequencies, compare trends, group positive and negative terms, and start building proper analytical summaries. We can also assign a sentiment category to each post. If a tweet is positive, perhaps it gets a 1. If negative, perhaps a minus 1. If neutral, then zero. 

Now suddenly we can count how many positive, negative, and neutral posts we have in total. 

From numbers to narrative 

Shubhadeep Mukherjee: Once we do that, we can start producing meaningful outputs. 

For example, we might generate a word cloud, where larger words indicate words appearing more frequently. We may find that “pumpkin” and “spice” are, unsurprisingly, among the biggest words, but we may also see words such as “crash” and “expensive” appearing prominently. 

We can also generate a sentiment breakdown and discover that perhaps 30% of posts are positive, 10% are neutral, and 60%are negative. 

At this point, we are no longer lost in the chaos of raw customer comments. We have transformed all of that into a narrative supported by data. 

We can now ask more focused questions. What are the top negative words? What are the top positive words? What does the sentiment distribution tell us? 

Business insight 

Shubhadeep Mukherjee: And this is where business analytics becomes really useful. 

What does our analysis suggest? 

It suggests that customers generally still like the flavour. So, from a product point of view, the drink itself may not be the issue. 

But the negative words suggest at least two strong concerns. One is app crashing, which points to a service or technical issue. The other is pricing, which points to a pricing or value perception issue. 

So, from a very basic but useful piece of analysis, we have already moved from chaos to clarity. 

The likely issues are not with the core product. The likely issues are with the app’s technical performance and with customer perceptions of the price. 

From insight to decision-making 

Shubhadeep Mukherjee: Now we move from output to business action. 

This is a very important step. The purpose is not simply to say, “Our model is 90% accurate,” and stop there. That is not the value of business analytics. 

The value lies in helping a business make decisions. 

So now we might tell the marketing and operations teams the following. 

There appears to be a pricing gap, meaning there is a mismatch between what customers expect to pay and what they are actually being charged. That mismatch is creating dissatisfaction. 

There also appears to be a service quality mismatch. Customers expect the app to work smoothly, but instead it is crashing, which creates frustration. 

That is already very useful. Without analytics, a company might only have instinct or anecdotal guesses. But with analytics, they have a much stronger basis for action. 

Possible actions 

Shubhadeep Mukherjee: So, what might the business do next? 

It could respond directly to dissatisfied customers on review platforms and social media. It could fix the app immediately or roll back to the previous version if a recent update caused the issue. It could publicly acknowledge the frustration while still reinforcing the fact that many customers do like the flavour. 

It might also provide a discount code or some app-based incentive to regain goodwill and encourage people to come back. 

So, the idea is that the analyst produces the insight, that insight goes to the business team, and then the business team takes action. 

That is exactly the chain of value we want you to understand. 

Monitoring the recovery 

Shubhadeep Mukherjee: After that, of course, we would want to continue monitoring the situation. 

We might establish KPIs and dashboards to track sentiment, engagement, app stability, and related indicators over time. 

Suppose the app is fixed and the messaging is improved. We might then observe that positive sentiment starts to rise, neutral sentiment remains stable, and negative sentiment begins to fall. That would indicate that our mitigation strategy is working. 

So not only are we able to diagnose a crisis, but we are also able to monitor whether our response is actually improving the situation. 

Why this matters for the MSc 

Shubhadeep Mukherjee: So, let me summarise what we have done. 

• We started with a real business problem. 

• We gathered unstructured data. 

• We cleaned and transformed that data into a structured form. 

• We analysed sentiment and patterns. 

• We identified the problem areas. 

• We fed that insight back into the business. 

• Then we designed possible actions based on that evidence. 

That is the journey from business problem to actionable insight. 

And that, really, is what this programme is about. 

It is not about analysis for the sake of analysis. It is about using analytical thinking, data handling, and problem-solving to help businesses make better decisions. 

Peter Ludgate: That brings us to the end of the case study and into our Q&A. 

Before we do that, though, thank you so much, Shub, for walking us through that. It was incredibly interesting. I think this is the second time you have done this type of webinar, and it is always extremely engaging. 

I honestly think that even somebody who did not begin with an interest in business analytics would struggle not to have one after hearing that. 

Shubhadeep Mukherjee: That is very kind of you, Peter. Thank you. And apologies that you have to sit through it every time I do it. 

Peter Ludgate: Not at all. It is like re-reading a favourite book or watching a favourite episode again. 

Disclaimer: 

This transcript is based on a recorded webinar and has been lightly edited for readability, clarity and conciseness. We’ve removed filler words, false starts, and repeated phrases (such as “um” and “ah”) without changing the meaning. All information remains accurate to the recording. 

Tuition fees may be subject to annual increases in line with inflation from 2026, in accordance with UK Government policy and maximum fee limits.