What AI Actually Knows About You: I Investigated My Digital Shopping Profile

After writing about AI personalization and chatbots, I kept thinking about one uncomfortable question: what does AI actually know about ME? So I decided to find out. What I discovered in my digital shopping profile was both fascinating and unsettling. Here’s my journey into the data that shapes every shopping experience I have online.

I’m going to be honest with you – starting this investigation made me nervous. We all have that vague awareness that our online shopping behavior is being tracked, analyzed, and used to personalize our experiences. But how many of us actually know what’s in our digital profile? What patterns have AI systems identified about us? What predictions are they making?

Last week, I decided to find out. I requested my personal data from several major e-commerce platforms, dove into my analytics profiles, and spent hours analyzing what these systems have learned about me over years of online shopping. The results surprised me in ways I didn’t expect.

This isn’t just an academic exercise. As I’ve been researching AI in e-commerce for The Neurals, I’ve realized that understanding what data exists about us is crucial to understanding how modern commerce actually works. So let me take you through what I found, how I found it, and what it all means.

The Data Request Process: Harder Than It Should Be

Before I could analyze what AI knows about me, I had to actually get my hands on the data. Thanks to GDPR (the General Data Protection Regulation), I have the legal right to request this information from any company processing my data. Under Article 15 of GDPR, consumers have the right to request from companies what data they have collected, who has access to it, and information on how it’s used.

The theory is simple: submit a data subject access request, wait up to 30 days, and receive a comprehensive download of everything a company knows about you. The reality? It varied wildly by company.

Some platforms made it relatively straightforward – buried somewhere in privacy settings was a “Download Your Data” option. Others required me to email specific departments with proof of identity. One platform made me fill out a PDF form, sign it, and upload it back. The experience itself told me something important: companies are legally required to provide this data, but they’re not exactly eager to make it easy.

After two weeks of requesting, following up, and waiting, I finally had data packages from five major e-commerce platforms I regularly use. What I found inside those files was eye-opening.

The Surface Level: What I Expected to Find

Opening the first data package, I found what I anticipated: my basic account information, purchase history, addresses, payment methods (partially masked), and browsing history. Nothing particularly surprising here. I knew these platforms stored my shipping addresses and remembered what I’d bought.

But as I dug deeper, I realized this was just the tip of the iceberg. The real insights weren’t in what I had bought, but in what the AI had inferred about me based on how I shop, when I shop, and what patterns my behavior reveals.

The Hidden Profile: Behavioral Patterns I Didn’t Know I Had

Here’s where things got interesting. Buried in the data files were behavioral categories, preference scores, and pattern analyses I had no idea existed. One platform had assigned me to something called a “value-conscious frequent browser” segment. Another had tagged me as a “evening research shopper with high cart abandonment.”

These weren’t just labels – they were detailed psychological profiles built from thousands of micro-interactions. The AI had noticed that I tend to browse products extensively at night, add items to my cart, but rarely complete purchases immediately. It had categorized this as “research behavior” and adjusted what it shows me accordingly.

Customer Data Platforms can assemble, unify, and reconcile various data types, recording sensitive details such as customers’ names, addresses, and gathering data about buyers’ purchases and engagement activities. What surprised me most wasn’t that this data existed, but how accurately it captured patterns I wasn’t consciously aware of myself.

The Timeline Analysis: My Shopping Rhythms

One of the most fascinating discoveries was seeing my shopping behavior mapped across time. The AI had identified seasonal patterns in my purchasing, down to specific months and even days of the week. It knew that I’m most likely to make electronics purchases in November and December, buy clothing in spring and fall, and rarely shop on Sundays.

But it went deeper than that. The systems had identified micro-patterns: I’m more likely to complete a purchase if I’ve viewed a product at least three times over several days. I tend to abandon carts on mobile but complete them on desktop. I rarely buy the first item I click on, but I often return to products I’ve bookmarked.

By analyzing customer behavior, purchase history, and engagement patterns, AI tools help deliver the right message to the right person at the right time. The result is that 81% of marketers say AI has helped them increase brand awareness.

These patterns revealed something about my decision-making process that I’d never articulated, even to myself. The AI had created a model of how I think about purchases, and it was unnervingly accurate.

The Price Sensitivity Score: What I’m Willing to Pay

Perhaps the most uncomfortable discovery was finding that several platforms had assigned me some form of “price sensitivity score” or “willingness to pay” indicator. While the exact methodology wasn’t spelled out in the data dumps, the implications were clear: these systems have calculated how price-conscious I am and likely adjust what they show me accordingly.

For one category of products, I was scored as “low price sensitivity” – meaning I tend to buy regardless of minor price differences. For others, I was marked as “high price sensitivity” – prone to abandoning purchases when prices increase even slightly. This wasn’t based on my income or any demographic data, but purely on behavioral patterns the AI had observed.

This raised immediate questions about dynamic pricing and whether I’m seeing the same prices as other customers. The data I received didn’t explicitly confirm dynamic pricing, but it certainly contained all the information needed to implement it.

The Preference Engine: Predicting My Tastes

Beyond purchase behavior, the AI systems had built detailed models of my aesthetic preferences, brand affinities, and style tendencies. One platform had categorized my style as “minimalist contemporary” with specific color palette preferences I hadn’t consciously identified but were absolutely accurate when I reflected on my purchases.

Using AI, big data, and business intelligence, these systems analyze customer behavior and preferences to improve customer experience through custom notifications, product suggestions, and tailored marketing campaigns based on purchasing patterns.

The system knew my size preferences across different brands (accounting for how sizing varies), my preferred materials for different product categories, and even seasonal variations in my style choices. It had noticed that I buy more colorful items in spring and more neutral tones in winter – a pattern I’d never consciously recognized.

The Social Graph: My Influence and Influences

Perhaps most surprising was discovering that some platforms had mapped my connection to other users through various signals – shared addresses (gifts), similar browsing patterns, and coordinated purchases. While names weren’t directly associated with these connections, the systems had identified a “shopping network” around me.

This social graph was being used for collaborative filtering – “people in your network also liked these items” – but it also meant my shopping behavior was influencing what others see, and vice versa. My purchases weren’t just about me; they were data points contributing to a broader network of influence and recommendation.

The Predictive Profile: What AI Thinks I’ll Buy Next

The most fascinating section of data was the predictive modeling. Several platforms included probability scores for different product categories – essentially, what the AI predicted I would purchase in the coming months based on my historical patterns and current behavior.

These predictions were specific: 67% probability of purchasing athletic wear in the next 30 days, 34% probability of buying small kitchen appliances in the next 90 days, 89% probability of replacing my phone case within the next quarter.

What made these predictions particularly interesting was their accuracy. Looking at my actual intentions (which the AI obviously can’t read), the predictions were right more often than they were wrong. The system had learned enough about my replacement cycles, seasonal patterns, and life stages to make educated guesses about my future needs.

The Context Layer: Location, Time, and Device Signals

Digging deeper into the data revealed that every interaction I had with these platforms was stamped with contextual information: my location (sometimes precise, sometimes city-level), the time of day, the device I was using, even my network connection type.

This context was being used to understand my shopping behavior patterns across different situations. The AI had learned that I browse casually on mobile during commutes, conduct serious product research on desktop in the evenings, and make impulse purchases on mobile during specific times of day.

One platform had even identified that my shopping behavior changes based on whether I’m at home or traveling, adjusting recommendations accordingly. It knew that when I’m traveling, I’m more likely to make quick decisions and less price-sensitive for convenience items.

The Missing Pieces: What I Couldn’t See

As revealing as these data dumps were, I’m certain they don’t contain everything. The data provided was clearly sanitized – formatted for consumer consumption rather than showing the raw algorithmic models and scoring systems that actually drive personalization.

The “black box” problem in AI personalization is real. While I could see that I’d been assigned to certain behavioral segments, I couldn’t see the full decision tree of how the AI makes real-time choices about what to show me. I could see historical patterns, but not the live predictive models running in real-time during my shopping sessions.

Customer Data Platforms deliver tailored insights and recommendations that help retailers drive growth, improve personalization, and enhance customer loyalty through AI trained on billions of retail touchpoints.

The Privacy Implications: Should I Be Worried?

After seeing all this data, my first reaction was concern. Should I be worried that these systems know me this well? The honest answer is: it’s complicated.

On one hand, this data is what enables the personalized experiences I actually benefit from. The AI knows my size, my style, and my preferences because I’ve taught it through my interactions. When I get relevant recommendations instead of random products, it’s because of this profiling.

On the other hand, the depth and breadth of behavioral analysis feels invasive. The idea that systems are calculating my price sensitivity, predicting my future purchases, and potentially adjusting what I see (and maybe what I pay) based on proprietary algorithms is uncomfortable.

The risk isn’t necessarily in the data collection itself, but in how it might be used. With GDPR and similar regulations, there are legal protections around data security and usage. Companies must implement appropriate technical and organizational measures to protect data, including encryption and access controls.

But beyond legal protections, there’s the question of information asymmetry. These companies know far more about me than I know about them. They understand my decision-making patterns, my vulnerabilities, my predictable behaviors. That knowledge gap represents power, and power can be misused.

The Accuracy Question: How Right Is the AI?

As I reviewed all this data, I kept asking myself: is this profile actually accurate? The surprising answer is that it’s both impressively right and occasionally wrong in interesting ways.

The AI correctly identified many patterns I recognized once they were pointed out. My evening browsing habits, my seasonal purchase patterns, my preference for researching before buying – all accurate. The behavioral segments I was placed in felt apt.

But there were also errors that revealed the limitations of algorithmic understanding. One platform seemed convinced I was interested in a product category I’d only browsed once out of curiosity. Another had misinterpreted gift purchases as personal preferences, suggesting similar items I had no interest in.

These errors are instructive. They show that AI personalization, while sophisticated, is still fundamentally reactive and pattern-based. It can’t read intent, doesn’t understand context the way humans do, and sometimes mistakes correlation for causation. My single curious click became “interest in model trains,” and now I see hobby supplies I’ll never buy.

The Industry Perspective: Why This Data Exists

To understand why companies collect all this information, it’s important to see it from their perspective. AI-driven personalized recommendations contribute to a 15-20% increase in conversion rates, and 80% of customers say they are more likely to do business with a company if it offers personalized experiences.

This isn’t just about invasive surveillance – it’s about creating better customer experiences and running more efficient businesses. The data helps companies understand aggregate behavior, optimize inventory, improve product development, and yes, sell more effectively.

Generative AI models used for e-commerce analysis require large amounts of data from customer behavior, transaction histories, and market trends to create ultimately personalized experiences with dynamically tailored content that resonates with each user’s demands.

From a business standpoint, customer data is crucial for survival in competitive e-commerce markets. Companies that can’t personalize effectively lose customers to those that can. The data collection isn’t malicious – it’s competitive necessity.

But that doesn’t mean it’s without problems. The challenge is balancing legitimate business needs with individual privacy rights and ensuring that personalization enhances rather than manipulates customer decision-making.

What You Can Do: Taking Control of Your Data

If you want to investigate your own digital shopping profile (and I recommend you do), here’s what I learned about the process:

How to Request Your Data

Look for “Privacy” or “Data Protection” settings in your account on major platforms
Search for “Download My Data” or similar options
If not obvious, email customer service with a formal data subject access request
Be prepared to verify your identity through security questions or ID verification
Wait patiently – companies have up to 30 days to respond under GDPR

What to Look For

When you receive your data, focus on:

Behavioral categories and segment assignments
Purchase patterns and frequency analyses
Preference profiles and taste predictions
Interaction histories showing browsing patterns
Third-party data sharing records

Your Rights

Remember that under GDPR (and similar regulations like CCPA), you have the right to:

Access your data – see what’s collected about you
Correct inaccuracies – fix wrong information in your profile
Delete data – request removal (with some exceptions)
Opt out of certain processing activities
Data portability – take your data to competitors

Companies must also provide transparency about collecting, using, and protecting personal data through comprehensive privacy policies that are easily accessible and clearly explain customer rights.

The Future: Where This Is Heading

As I reflect on this investigation, I’m struck by how this is just the beginning. The AI systems analyzing our behavior are becoming more sophisticated, the data being collected more granular, and the personalization more precise.

By 2032, the ecommerce AI market is expected to reach $45.72 billion, with 84% of ecommerce businesses placing AI as their top priority. We’re moving toward a future where AI agents can handle entire shopping processes autonomously, from identifying needs to completing purchases based on learned preferences.

The question isn’t whether AI will know more about us – it almost certainly will. The question is how we ensure that knowledge is used ethically, transparently, and in ways that genuinely serve customer interests rather than simply maximizing extraction of value.

My Personal Takeaways

After spending weeks investigating what AI knows about me, here are my honest conclusions:

I’m impressed by the sophistication. The depth of behavioral analysis and accuracy of pattern recognition is genuinely remarkable. These systems have identified aspects of my decision-making I wasn’t fully aware of myself.

I’m concerned about transparency. While I could access my data through formal requests, this information isn’t readily visible during normal shopping experiences. The asymmetry of knowledge bothers me.

I’m conflicted about the trade-offs. I genuinely benefit from personalized recommendations and experiences. But I’m uncomfortable with the extent of profiling and the potential for manipulation.

I’m more conscious of my digital footprint. Knowing that every click, every hover, every abandoned cart is being analyzed has made me more mindful of my online behavior – not in a paranoid way, but in an informed way.

I’m optimistic about regulation. GDPR and similar laws aren’t perfect, but they represent important steps toward balancing business interests with individual rights. The fact that I could access this data at all is progress.

The Bigger Picture: What This Means for All of Us

This investigation was personal, but the implications are universal. We’re all building digital profiles through our online interactions, whether we’re conscious of it or not. These profiles shape our experiences, influence our decisions, and potentially affect the prices we see and the opportunities we’re offered.

The AI systems analyzing us aren’t going away – if anything, they’re becoming more sophisticated. More than half of consumers say they will likely become repeat buyers after a personalized shopping experience with a retailer, driving companies to invest even more heavily in personalization technology.

But we don’t have to be passive subjects of analysis. We can:

Educate ourselves about what data is collected and how it’s used
Exercise our rights to access, correct, and delete our data
Make informed choices about which platforms we use and how we interact with them
Demand transparency from companies about their AI systems and personalization practices
Support regulations that balance innovation with privacy protection

A Final Thought on Privacy and Personalization

As someone who’s now deep into researching AI in e-commerce, I’ve come to believe that the privacy versus personalization debate is often framed incorrectly. It’s not an either/or choice – we can have both sophisticated personalization and robust privacy protections.

What we need is:

Transparency about what data is collected and how it’s used
Control over our data and how it’s shared
Accountability when systems make mistakes or are misused
Balance between business interests and individual rights

The AI systems analyzing our shopping behavior are tools – powerful ones, but tools nonetheless. Whether they serve us or exploit us depends on the values and regulations we build around them.

This investigation has been enlightening, sometimes uncomfortable, but ultimately valuable. I now shop online with my eyes more open, understanding that every interaction is shaping my digital profile. But rather than making me paranoid, this knowledge has made me more empowered.

I encourage you to take this journey yourself. Request your data. See what AI has learned about you. You might be surprised by what you find – and that surprise itself is valuable information.

This investigation has been both fascinating and challenging. Have you ever requested your data from e-commerce platforms? What did you discover? I’d love to hear about your experiences and thoughts on the balance between personalization and privacy. Share your stories as I continue exploring how AI is reshaping our digital lives.

Next in this series, I’ll be exploring the actual tools and technologies that make this data collection and analysis possible – and whether small businesses can (or should) implement similar systems. Follow along as The Neurals continues investigating the real-world impact of AI in commerce.