What Is System Zero? A New Layer of Consumer Decision-Making
By Dig Insights
- article
- Agile Quantitative Research
- Agile Qualitative Research
- Behavioural Insights
- Brand Research
- Concept Testing
- Diary Studies
- Innovation Research
- Trend Monitoring
- Trend Scouting
- Questionnaire Design
Summarise with AI
A little while ago, my wife and her best friend were out wandering around London. They had been in Coal Drops Yard, then made their way to Soho and found themselves standing in front of a Sephora about to go in. As they were about to enter, my wife’s friend asked her if she had done a colour analysis through ChatGPT. My wife said no, so her friend took out her phone, took a selfie, uploaded it and prompted the tool for her colour profile.
A few seconds later it came back with reddish undertones, brown hues, the whole description and then asked whether she would like to know which foundations matched her skin tone. She said yes. A short list of brands appeared, a mix of familiar and unfamiliar names. She asked which one the tool would actually recommend for her, and without flinching, clicked the first recommendation, paid with Apple Pay and had it delivered home.
It is worth pausing on what happened in that story. My wife’s friend never entered the Sephora. She never spoke to a specialist human. She was not even seeking to make a purchase that day. Everything in the middle, from awareness through to consideration and even the final choice, was delegated.
She did eventually choose to buy something, but she wasn’t even aware of the full list of brands she had chosen from, nor really in control of the final decision. And this wasn’t a one-off, nor is she an early adopter. Realistically, about five per cent of consumers let AI make a purchase decision outright, but significantly, four in five have used an AI tool in their purchase process in the past month.
Behind System 1 and System 2
Most of us are familiar with Daniel Kahneman’s System 1 and System 2. System 1 thinking is fast, instinctive, automatic and associative. I recognise a brand, I reach for it and I put it in my cart. There is no real active thought process, or if there is, it is intuitive and accelerated. System 2 is deliberate. It’s slow and analytic. I compare my options, weigh my choices, check price against quality and make a considered decision. Both of these models assume that the consumer is the one making the decision, that the consumer is doing the cognitive work.
What happened outside that Sephora didn’thappen in either System 1 or System 2. It happened outside the consumer’s head entirely, and it arrived pre-formed. That’s the layer we’ve spent this year measuring at Dig Insights, and we call it System Zero.
It’s not a new mode of human thought, and it shouldn’t be treated as a third psychological system, because it doesn’t happen in the human mind. It sits deliberately alongside System 1 and System 2, and is defined by delegated cognition. I should say we didn’t invent this terminology. It comes from academic literature, and our job has simply been applying it to our field.
Why Traditional Research Needs a Second Lens
In practice, this means I can now ask an AI system to build my consideration set, compare alternatives, explain the trade-offs and tell me what it thinks I should choose. I may still make the final decision, but I’m no longer doing all the work that gets me there.
Legacy research tools don’t capture that, and neither do the decades of surveys we have designed and perfected for traditional consumer research. While traditional research doesn’t need to be replaced, it does need a second lens because without it the research risks misrepresenting reality, or at a bare minimum, representing an incomplete version of it.
System Zero Across Categories
System Zero isn’t confined to one category or one kind of decision, and it doesn’t show up the same way everywhere, but it grows in importance with every new product release. In shopping, AI can narrow choices and recommend products. In financial services, it’s helping people assess risk and compare options. I know, because when I opened a new joint bank account recently, my first stop was a long conversation with Claude, and it kept redirecting me away from the account I originally wanted.
In travel, AI is building itineraries and adapting plans, and Expedia now has an agent function given how quickly that disruption has happened. In customer service, it’s resolving issues before a person ever gets involved, and we know IKEA has made major changes here as well. The same pattern is also showing up in health, work, the smart home and information discovery.
What is important is that a lot of this influence happens out of sight. Consumers don’t always experience it as “I used AI to make a decision”. Some may simply experience it as an easier path, a shorter list, a suggested next step, a default option or simply not clicking a link after a Google search.
For insights professionals, the implication is fundamental. Consumers aren’t the only ones making decisions anymore. They are making decisions with AI, through AI and sometimes because of AI.
A Compressed Consumer Journey
The journey itself has changed shape. It doesn’t necessarily follow the same steps it used to, and some compression is happening. AI can introduce brands a person might never have considered, or hide brands they might have considered. It can summarise a complex category, compare alternatives, narrow the list and recommend an option it believes fits the need.
Historically, consumers have done much of this work themselves: they searched, read, compared, interpreted and decided, which is why we have surveyed them about it for so long. System Zero compresses that process, and as reliance on it grows, the consumer journey becomes partly human and partly machine-mediated. When the machine intervenes, it decides what surfaces, what doesn’t and which brand associations get made, including whether a brand is present in the conversation at all.
What This Means for Brands and Insights Teams
So what changes for anyone working in brand or insights? First, choice becomes pre-filtered. Consumers may still choose, but increasingly from a set that’s already been shaped for them. Second, preference doesn’t necessarily equal performance. A brand can test beautifully with consumers and still lose out if AI systems rarely surface it, or represent it poorly. Third, influence now has a new gatekeeper. Brands are no longer competing only for human attention; they’re also competing for algorithmic inclusion.
None of this means replacing what we already do. Brand equity work, journey mapping and segmentation are still valid and necessary, but they describe a shrinking share of the overall decision architecture. Understanding people remains essential; it’s just no longer sufficient on its own.
Building the Second Lens
Our industry has historically treated the consumer as both the source of preference and the principal site of decision-making. System Zero complicates that assumption. So let’s stop asking consumers to account for decisions they’ve already delegated, and start building the second lens that lets us see where, and how, that delegation is happening.
