Long tail theory:what Anderson said and what AI changes
By Enrico Giubertoni
Trusted by
Worldline
Alba Leasing
L'Erbolario
3M
The Long Tail by Chris Anderson (Italian edition, Codice Edizioni)
Long tail theory says that, online, the sum of many rare requests can be worth as much as the few requests that are very frequent. The Long Tail by Chris Anderson is one of the books that explain the economy we live in: the end of the mass market in favour of many small markets. Reading it, you recognise your daily reality, not a vision of the future.
The book matters for three reasons:
It moves marketing into a different dimension. The offer is no longer built for the average customer but for many specific ones.
It shows how much the internet has changed society. Knowledge has become a source of added value for the economy, and the book warns against strategic mistakes such as starving universities and scientific research of funds, which Italian governments have made repeatedly.
It warns companies tied to the past. A business and marketing system still built on the criteria of the 1990s and of the dotcom era (the boom of internet companies at the turn of the millennium) risks becoming obsolete quickly if it does not give up certainties that no longer exist.
The practical questions for a marketing manager are two. How does our e-commerce (online shop) offer services consistent with long tail economics? How does our corporate blog open a dialogue and a relationship with customers?
Two videos help to see the theory at work: the official promo of the book (aptstudio, 2012) and a short explanation of what the long tail is (Business School 101).
Does the long tail still hold in the age of AI assistants?
The long tail is the model in which, online, the sum of very many rare requests weighs as much as, or more than, the few requests that are very frequent. Yes, it still holds, and it returns on the demand side: people ask long, precise questions, and each one is rare.
With assistants based on artificial intelligence (programs such as ChatGPT or Gemini, where you write a question in natural language and receive a written answer) whoever decides range and content in a company meets again the chart of more than twenty years ago, with one difference: the tail is now also read by a machine that answers in place of your website.
What did Chris Anderson claim, and who objected?
Chris Anderson described the idea first in an article in Wired (a technology magazine) in 2004 and then in the book of 2006. In a physical shop, shelf space is expensive, so it holds only what sells in volume.
Online, keeping one more item in the catalogue costs almost nothing, so niche products become profitable and, added together, can be worth as much as the hits (the few best-selling products).
The best-known objection comes from Anita Elberse of Harvard Business School, in the Harvard Business Review of July-August 2008. Analysing sales of music and video, she observed that the tail was getting longer but remained thin, and that the hits kept collecting the largest share of revenue. Anderson replied that the result depends on where you draw the line between head and tail. For a manager, the lesson sits in the middle: the tail exists, and the head still pays. The case of Blockbuster (the video rental chain) shows what a company risks when it builds its offer only on big hits.
What changes when an AI assistant gives the answer?
The place where the tail is read changes.
The Pew Research Center (a US research institute) followed the browsing of 900 American adults in March 2025 and published the results on 22 July 2025. Out of 68,879 Google searches, about 18% produced a summary written by AI, the AI Overviews (the box at the top of the results that condenses the answer). The summary appears in 8% of searches of one or two words and in 53% of searches of ten words or more. When the summary is present, users click on a traditional result in 8% of visits, against 15% when it is absent.
Taken together, the two figures say that long questions, that is the tail, are the ones AI answers most often, and that where AI answers, people click less: more searches end without a visit to any website, the so-called zero-click searches. The link between the two figures is my own reading, not a conclusion of the study, and I have not found a direct measurement of how the length of questions relates to lost clicks.
The wider picture points in the same direction. On 20 May 2025 Elizabeth Reid, head of Google Search, wrote on the company blog that people come to Google with more questions, including more complex, longer and multimodal ones (mixing text, images and voice). And on 25 October 2019 Pandu Nayak, Google’s Vice President of Search, wrote that 15% of the queries (the questions typed into a search engine) Google receives every day are new, never seen before. A tail that does not repeat itself is a tail that no keyword list can exhaust.
How does a B2B company cover the tail of questions?
Work on three levels, in the order in which it pays to do them. B2B (business to business) means a company that sells to other companies.
Real customer questions. For each product or service, list the specific questions a buyer asks before buying. A manufacturer of industrial valves, for example, receives requests such as “stainless steel valve for steam, with food-contact certification”: each one is rare, and together they describe the real market.
Pages an assistant can cite. A two-line definition at the opening, headings written as the questions users ask, every figure with a source and a date, and structured data (information written in the code of the page in a format machines read without interpreting). This work is called generative engine optimisation, meaning writing so that an AI assistant cites your page.
A measure that looks at citations. The click remains an indicator, but alone it is no longer enough: check also whether your brand appears in the answers that assistants give to the questions of your own tail.
The head is not abandoned: the few very frequent requests remain the core of revenue, as Elberse shows, and the tail complements them.
Sources
Pew Research Center, “Google users are less likely to click on links when an AI summary appears in the results”, 22 July 2025: pewresearch.org
Anita Elberse, “Should You Invest in the Long Tail?”, Harvard Business Review, July-August 2008: hbr.org
What should a manager take away from reading The Long Tail?
The recognition of a daily reality: the mass market gives way to many small markets. Reading it, you do not get the impression of a vision of the future. The book invites you to check whether your offer, your catalogue and your content still follow the criteria of the mass market.
Why is The Long Tail important for marketing managers?
Because it moves marketing into a different dimension. It shows how much the internet has changed society, and how pervasive knowledge has become as a source of added value for the economy. It pushes you to ask how your e-commerce can offer services consistent with long tail economics, and how your corporate blog can open a dialogue with customers.
What do companies risk if they ignore long tail economics?
They risk becoming obsolete quickly and being overtaken. A business and marketing system still tied to the criteria of the 1990s and the dotcom era loses ground if it does not give up certainties that no longer exist.
