Everyone now googles what antiques are worth. Why Google gives you ten links and no answer

Recently, opinion writer Brecht Nuyts wrote in the Belgian newspaper De Standaard that junk no longer exists. The flea market of the past, where you bought something without knowing whether it was worth anything, has disappeared. Not because fewer people show up, but because we have all changed our minds. We have become little strategists, with a smartphone in hand and a calculation running in our heads. He is right. And it is precisely there, inside that smartphone, that a problem lies which nobody names out loud.

The piece describes a summer flea market a quarter of a century after the author's youth. Crowd barriers, portable toilets, food trucks, and little tables of branded clothing, neatly sorted by degree of wear. The brocante with its anarchic zest for life has turned into a tightly organized open-air fair for private micro-entrepreneurs. And the most striking detail arrives almost in passing. In the old days, Nuyts writes, the value of an object was not yet systematically validated through searches on our smartphone. When in serious doubt, you could always count on the second opinion of a passing bystander. A fellow soul, an expert on the matter. That advice felt as familiar as a warm blanket.

That warm blanket is gone. In its place, you now stand alone at the stall, phone in hand, trying to google something. And it works only halfway. This article is about why it only works halfway, and about what it takes to make it whole again.

The flea market did not die, it changed its mind

Let us start with where the article is right, because that is bigger than it looks. The change is not in the market but in the people. Where a flea market used to be a public, material confession about surplus belongings and embarrassing collections, it is now an exercise in maximum value recovery. Every yellowed comic album is vintage. Every porcelain shepherd boy comes with a story attached. Every piece of metal has already been valued online, or is being valued on the spot.

The word junk has even become taboo. From now on it is called retro, nostalgia, pre-loved, or collector's item. Supply and demand, but with an emotional package insert. The seller wants the deal to pay off. The buyer wants the deal to pay off. Conversations serve that deal, no longer each other. Everyone has become a trader, including the person who simply wants to get rid of an old cabinet or is looking for a nice little trinket.

Anyone who regularly attends markets and auctions to buy stock recognizes this down to the fingertips. The real dealer, the connoisseur with thirty years of eye, is becoming a dying breed. Not because there is less demand for old things, but because his knowledge has been democratized into a search query. The amateur beside him now has the same reflex. Doubt about a plate, a figurine, a small painting? Phone out, photo, search. The private purchase has become a small business, and the search engine is the tool.

That is neither decline nor progress. It is simply a fact. And it raises a question that is interesting both commercially and humanly: if everyone now checks value on their phone, why does nobody feel any more certain?

The bystander who disappeared

The answer begins with that vanished bystander. In the old days, when in doubt, there was almost always someone standing next to you. An older man who had been coming for forty years. A collector who happened to walk by. A friend who knew just a little more. You held the object up, you said "what do you think," and you got a verdict. No report, no proof, but a verdict. A human being who said: that is a later copy, leave it. Or: that is old, buy it, it is a bargain.

That verdict did two things at once. It gave you information, and it gave you peace of mind. Someone with more experience took the decision onto his own shoulders. That is exactly what a warm blanket is: not only the right conclusion, but the feeling that you are not facing it alone.

The smartphone gave you the information back, in abundance even, but not the peace of mind. You still stand alone at that stall. Only now there is no longer a human being beside you saying "buy it" or "leave it." There is a screen that hands you ten blue links and kindly asks you to figure it out yourself. The bystander has been replaced by a list. And a list places no hand on your shoulder.

Google is brilliant. And that is where the problem begins.

Now an honest word, because this is not an attack on Google. Google is one of the most impressive things humanity has ever built. It is the largest library in history, in your pocket, free, in a fraction of a second. For a thousand questions a day it is the perfect answer. For antiques it is indispensable as a starting point. We ourselves use image search technology in our analyses, so we would be the last to deny the power of search.

But that very power is the problem. Google gives you everything, and in doing so it decides nothing. Type in a description of your object and you get a thousand images that roughly resemble it. Some genuine, some reproductions, some something else entirely. You get forum threads from 2011 where three people disagree. You get marketplace listings running from fifty to five thousand euros for what looks like the same piece. In other words, you get raw material. An enormous pile of fragments that you still have to assemble yourself into something resembling a verdict.

And that is where it pinches. To assemble that pile, you need exactly the thing you do not have, because otherwise you would not have googled at all: a trained eye. The dealer with thirty years of experience has that eye. He can weigh ten contradictory search results in two seconds and pull out the right one, because he is the search engine himself. The layperson cannot. He stares at ten links and still does not know which of the ten is his.

Ten links versus one answer

This is where the whole difference lies, and it is a small difference that changes everything. Google gives you the ingredients. It gives you flour, butter, eggs, and an oven full of other people's cakes. But it does not bake for you. You have to cook yourself, and you have to be able to cook. Whoever does not know the recipe stands there with the ingredients in hand and gets no further.

An antique analysis does the opposite. It chews the material for you and gives you the answer. Not ten links but one verdict. Not a half-hour search but a conclusion in a few minutes. Not "here is everything that resembles your vase, good luck," but "this is in all likelihood a later reproduction, decorative value, no collector's value, and here is why." That is a pre-chewed answer, and pre-chewed is exactly what you need at the moment you have to make a decision.

Think back to the bystander. What that man did was not searching. He gave a verdict. He had done the searching years earlier, in his head, and he delivered only the conclusion. That is what a good analysis gives back: not the search work, but the conclusion. The warm blanket, only now inside your phone, at the moment you need it.

This is also why a price comparison on an online marketplace so often does not help. You see that comparable pieces range between fifty and five hundred euros, but you do not know where yours belongs in that spectrum, because you cannot read the subtle differences that determine it. The spread itself is the proof that the raw data does not save you. Someone has to interpret the spread for you. Anyone who wants to understand more about why asking prices and actual sale prices diverge so widely will find more on that in our piece on appraisal value versus market value.

One vase, three answers, no certainty

An example makes it concrete. Suppose you find a blue-and-white porcelain vase at a brocante. It looks old, it feels heavy, and on the base there is a mark in Chinese characters. You do what everyone does: you take a photo and run an image search on it.

What you get back is a fan of possibilities. At the top, an almost identical vase that sold last month at a major auction for twelve thousand euros. Below it, one on a web shop for eighty euros, offered as a modern reproduction. Somewhere in between, a forum thread in which three collectors passionately disagree about the dating, and a two-hundred-euro listing with no further explanation. Four answers, between eighty and twelve thousand euros, for something that on your screen all looks alike.

Which of the four is yours? That is precisely the question you cannot answer, because to answer it you must be able to read the glaze, the foot ring, the brushwork, and the nature of the mark. If you cannot, you unconsciously choose the answer you most want to hear, usually those twelve thousand euros, and that is exactly how people talk themselves rich over an eighty-euro reproduction. The spread in the search results is not information. It is a trap with the appearance of precision.

An analysis does the only useful thing here: it chooses. It weighs the features you cannot weigh and says where on the scale your vase truly stands, with the reason attached. Not four answers, but one, and the honesty to say how certain that one answer is. Anyone who wants to dig deeper into the pitfalls of this particular porcelain will find more in our piece on how to spot fake Chinese porcelain.

Why the layperson cannot finish Google's work

The difference between the connoisseur and the layperson is not a difference in access to information. Both have the same Google. The difference lies in the ability to finish that information into a verdict. And that ability is precisely what takes years to build, and what most people will never have.

A real dealer buys on his instinct, built up from thousands of objects that passed through his hands. He feels the weight of bronze against spelter. He sees in half a second whether a patina has grown genuinely or been applied. He recognizes the wrong sheen of a glaze that is too recent. That instinct is not magic, it is condensed experience. But it is also not googleable. You can look up the conclusion of that instinct, but you cannot download the instinct itself.

That is the reason why the private little strategist, however well equipped with his smartphone, structurally lags behind the connoisseur. He has the tool but not the hand. He has the library but not the way of reading. And at the moment that it counts, at the stall, with a seller waiting for his answer, he has no half hour to weigh ten links he cannot weigh anyway.

An analysis fills that gap. Not by giving the layperson an eye, no one can do that, but by delivering the conclusion an eye would draw. It is a borrowed verdict, available to those who lack their own. For anyone who honestly wants to know where the limits of such an automated verdict lie, we have written those limits down explicitly in what AI cannot do in antique analysis. Because a borrowed eye is strong, but it is no magician, and you deserve to hear that from us, not discover it only afterward.

What an analysis does that a search query does not

In concrete terms, something different happens in an AntiqBot analysis than in a search query. A search query matches images and text and gives you the hits. An analysis runs through several layers and delivers a conclusion.

First, the object is assessed visually for shape, decoration, material, and stylistic features. Then any visible marks, stamps, signatures, and inscriptions are read and linked to knowledge of makers, periods, and regions. Next, comparable pieces are sought in recent auction results and dealer listings, in order to support an indicative market value. And then the most important thing happens, the thing a search engine never does: all that material is combined into a single report with a clear verdict, a value indication, and a short list of what you could still verify yourself.

That verdict works with five gradations, from authentic to not authentic, with probably authentic, uncertain, and probably not authentic in between. That gradation exists because antiques are rarely black and white. A late nineteenth-century Chinese vase with a Qing imperial mark is almost never simply genuine or simply fake. It is often a later production bearing a mark of reverence, and that deserves an honest naming rather than a reassuring or a dismissive label. Red flags are not glossed over with positive arguments. If there is doubt, the doubt appears in the report, with the reason attached.

The point is not that an analysis is always certain. The point is that it takes a position. Where Google lays ten possibilities before you and stays silent about which one is right, an analysis says where on the scale between reproduction and authentic your object stands, and why. That is the difference between material and verdict. Between a search and an answer.

At the stall, with the phone in your hand

Return to the moment that truly matters. You are standing at a flea market or a brocante. You are holding an object that costs you forty euros. The seller looks at you. You hesitate. Is this genuine, is this old, am I getting a bargain or being taken for a ride?

This is the moment where Google lets you down, not because it is bad, but because it is the wrong tool for this moment. You cannot conduct half an hour of research while a seller stands waiting and three other interested people hover around the stall. You do not need ten links. You need one thing: a verdict, fast, that you can lean on to say yes or no.

That is exactly the place where a verdict in your pocket makes the difference. A photo, a few minutes, a conclusion. Not the definitive truth, but the first, decisive estimate that lets you cut the knot with a settled mind. It is the bystander returning, digitally, at the moment you stand alone. Anyone who wants to get better at buying smart at markets will find practical points of attention in our article on buying antiques at the flea market.

And this is why recurring situations matter more than recurring objects. Most people do not find an old vase every week. But anyone who regularly attends markets, thrift shops, and auctions faces the same decision over and over again with a new object. It is not the object that returns, it is the situation that returns. And in every situation the question is identical: do I reach for my wallet now or not?

For those who live this way, the meaning of an analysis shifts. Whoever walks a brocante every Saturday, steps into a thrift shop monthly, or leafs through an auction catalogue, makes not a single decision but dozens per month. For those people, an analysis is not the aid for that one vase from an inheritance. It is a standing second opinion for every doubt that arises, available at the moment it counts. Not to replace their own judgment, but to have something to place beside it when that judgment falls short. That is a different kind of tool than a one-off check. It is a companion for those who meet the situation again and again.

Where Google wins, and we say so honestly

This would be a dishonest piece if we pretended that AntiqBot is always and everywhere better than Google. That is not so, and it is good to know where the giant remains unbeatable.

Google is broader than anything else. For pure factual questions, for finding a specific maker when you can already read the name, for looking up an auction house, for background on a stylistic period, Google is faster and more complete than any specialized tool. It is free, it is everywhere, and there is no category it does not touch. Anyone who sees a legible mark on a piece of silver and simply wants to know which silversmith it belongs to would do well to just look it up.

So AntiqBot does not replace Google, and does not try to. The two do different things. Google is the library. An analysis is the librarian who opens the right book for you to the right page and tells you what it says. You need both, and it is not a contest to win. It is a division of labor. David does not stand up to Goliath in order to defeat Goliath. David does the one thing that Goliath, for all his strength, does not do.

No David against Goliath to win, but to finish the job

So this is how it ends, not with a fight but with a complement. The article in De Standaard laments, with reason, the disappearance of the warm, human second opinion. The bystander who reassured you. It is a genuine loss, and no screen brings back entirely the human being who stood beside you.

But the function that person fulfilled can return. Not the company, but the verdict. Not the chat, but the conclusion you can lean on. Google democratized the information, and that was a gift. The next step is to democratize the verdict that this information still requires. Because information without the ability to finish it leaves the layperson exactly as uncertain as before, only with more tabs open.

The flea market has become a tightly organized open-air fair for private micro-entrepreneurs, Nuyts writes. That is true. And every micro-entrepreneur deserves the tool the big entrepreneur already has: not more information, because everyone has that now, but a verdict he cannot form himself. That is not an attack on the search engine that gave us so much. It is the finishing of the sentence the search engine left hanging halfway.

Frequently asked questions

Is AntiqBot better than just googling? Not better, different. Google gives you a pile of search results that you have to interpret yourself. AntiqBot gives you one substantiated verdict with an indicative value. For looking up a legible name or a fact, Google is fine. For making a decision under uncertainty, where you cannot weigh the material yourself, an analysis delivers the conclusion instead of the homework.

Can I use it while I am standing at the market? Yes. A photo with your phone and a few minutes are enough for a first verdict. For the best result, it is preferable to take a second photo of any marks, stamps, or the underside, because the depth of the analysis is tied to the quality of the photos.

What does it cost? Your first analysis is free after registration. After that, you buy credits in packs, starting from €0.60 per analysis. There is no subscription and there are no hidden costs.

Does this replace an appraiser? No. For a legally valid appraisal, an insurance policy, or a notarial inventory, you need a certified appraiser who is personally liable. AntiqBot helps you determine which objects are worth such a formal step, and gives you the working document for that first conversation. The last question is answered by a human, the first questions are answered by the analysis.

For which objects does it not work well? Loose gemstones and minerals require instruments such as a refractometer and a loupe, and fall outside our scope. For porcelain, silver, paintings, furniture, clocks, bronze, prints, glass, and African art, photo analysis works well, provided the photos are usable.