In a recent report, experience optimization specialist AB Tasty analyzed the behavior of 6,000 online shoppers across 6 countries, noting the end of the “search bar monopoly.” Here are 4 key takeaways from an emerging transformation.
1. Generative AI is redrawing the discovery experience
For roughly fifteen years, Google has reigned nearly unchallenged over product discovery. But that monopoly may be eroding. 60% of shoppers still begin their online purchasing journey through the famous search engine — a figure down 3 points from last year.
Not quite a seismic shift. Yet generative AI tools (ChatGPT, Gemini, Claude) have seen their usage jump from 8% to 14% of discovery interactions in just one year — a 75% increase.
That figure, however, masks a generational divide: 32% of Gen Z finds AI useful for shopping, compared to just 13% of baby boomers.
Younger consumers no longer type disembodied keywords like “red sneakers” — they ask conversational questions: “Find me a cool pair of sneakers for a festival, under €80,” the report illustrates.
The challenge for brands? A new kind of invisibility looms. Queries sent to conversational AI are a “black box” — it’s impossible to know exactly what a customer is asking, or to buy keywords to target them. Winning brands must therefore optimize their presence on these generative AI tools (known as GEO) to be “read” by the LLMs themselves, through structured data, demonstrated expertise, and authentic reviews.
2. The explosion of impulse buying
Another major trend: the surge in impulse purchases, which have tripled in a year, rising from 3% to 10%. Meanwhile, only 3% of shoppers say they never buy spontaneously.
Trust in online shopping has grown, and mobile interfaces have made “add to cart” frictionless. “Most interestingly, AI recommendations are making impulse purchases feel ‘smart’ rather than reckless, by validating the purchase before it’s made,” AB Tasty explains.
Furthermore, 74% of consumers fall into a middle zone: they plan most of their purchases but occasionally opt for spontaneous browsing — a vast market of upsell opportunities.
Best practices for optimizing conversion rates? Setting free shipping thresholds, creating “frequently bought together” product bundles, and offering time-limited deals at checkout. Turning “maybe” into “yes” is the real work of modern e-commerce.
3. The era of text-only reviews is coming to an end
Unsurprisingly, high-quality reviews remain the top purchase driver.
But their format is changing radically: text alone (58.9%) now rivals photos and videos (58.6%). The gap has all but disappeared.
Shoppers trust their eyes. They don’t want studio-retouched photos — they want raw, unedited, user-generated reality. They want to see fabric move, fit adjust, and products in their natural environment.
Star ratings, meanwhile, have become a baseline expectation, not a differentiator.
On the flip side, the top sources of frustration are pop-up windows and slow page load times.
4. Personalization — more illusion than reality
When it comes to personalization, there’s intention… and then there’s reality. 57% of shoppers cite “personalized offers and promotions” as the most important factor in making an experience feel personal — ahead of saved preferences or product suggestions.
Yet only 10% feel their favorite brands understand them “perfectly,” while 60% describe their experiences as “moderately” personalized and 16% feel no personalization at all.
Shoppers are telling us that ‘relevance’ is synonymous with ‘value.’ They feel understood when you help them save money on items they actually want to buy. They prioritize relevance over noise. They don’t want more emails — they want better ones. They’re willing to share their data (and give you their loyalty) if you offer them real value in return.”
Methodology:
For this report, AB Tasty surveyed 6,000 online shoppers split equally across the United States, the United Kingdom, France, Italy, and Australia. To ensure a panel of regular online shoppers, participants were asked: “When did you last make an online purchase?” Those who answered “less than 3 months ago” continued the survey, while those who answered “more than 3 months ago” were excluded — until a panel of 4,000 respondents was reached.
training.isarta.com 
