Berns and Moore: brain activity and later song sales

What happened

Gregory Berns and Sara Moore of Emory University published A neural predictor of cultural popularity in the Journal of Consumer Psychology. Thirty-two adolescents were studied and 27 were included in the fMRI analyses. They heard 15-second clips of 120 songs by unsigned or relatively unknown artists, taken from MySpace in 2006, and the authors later compared the brain responses with sales recorded through May 2010.

The psychological idea

The design asks whether brain responses in a small group can predict what a large population later buys, when the group's own ratings cannot. Three regions correlated with the average likability of each song: the cuneus, the orbitofrontal cortex and the ventral striatum (nucleus accumbens). Activity in those regions was then compared with sales.

In practice

Sales data were available for 87 of the 120 songs. Likability ratings did not correlate with sales (R = 0.110), but nucleus accumbens activity did (R = 0.32, p = 0.004; R = 0.27 without one outlier). Using a hit threshold of 15,000 units, a logistic model correctly classified 80 per cent of non-hits but only 30 per cent of hits. The authors suggest record labels might use such scans, and the paper reports no commercial use.

Why it mattered

The authors say truly prospective data on whether neural signals from a few people predict wider popularity had been lacking, and call their result modest. Falk and colleagues later cited it when reporting a similar pattern in smokers: medial prefrontal activity ranked three anti-smoking campaigns in the order of quitline call volume, while self-report did not.

Evidence behind the claims

• Berns and Moore studied 32 adolescents, five of whom were excluded from the fMRI analyses, leaving 27 (14 female and 13 male, aged 12 to 17.9, mean 14.6); the stimuli were 15-second clips of 120 songs from MySpace by unsigned or relatively unknown artists, 20 in each of six genres, and each participant heard 60 songs from their top three genres.

  Experimental findingSource: [1]Where: Material and methods, pages 155 and 156

• Sales came from Nielsen SoundScan, aggregating recorded sales from release through May 2010 including singles, albums and compilations; data were available for 87 of the 120 songs, and only three songs' albums reached the gold standard of 500,000 units.

  Experimental findingSource: [1]Where: Material and methods page 156; Results page 157

• Average likability (R = 0.110, p = 0.313) and familiarity (R = 0.106, p = 0.330) ratings did not correlate with sales, while nucleus accumbens activation correlated with log sales (R = 0.32, p = 0.004), or R = 0.27, p = 0.013 with the far left outlier excluded.

  Experimental findingSource: [1]Where: Results page 157 and Figure 2 caption

• With a hit threshold of 15,000 units the logistic model correctly classified 80 per cent of non-hits and 30 per cent of hits; the authors say the sales figure was dominated by album sales that may reflect band reputation, and that their data are modestly predictive of population effects.

  Experimental findingSource: [1]Where: Results page 157; Discussion pages 158 and 159

• The Crossref record gives the paper as Journal of Consumer Psychology volume 22, issue 1, pages 154 to 160, published online on 8 June 2011 with the print issue in January 2012; its abstract says liking did not predict sales but ventral striatum activity did.

  DocumentedSource: [2]Where: Bibliographic record and abstractChecked at abstract level

• Falk, Berkman and Lieberman recorded fMRI in 30 analyzed heavy smokers (31 recruited) watching three anti-smoking campaigns; medial prefrontal activity ranked the campaigns in the same order as quitline call volume while self-report ranked them differently, and the article cites Berns and Moore.

  Experimental findingSource: [3]Where: Method, Results and Introduction

• Ariely and Berns (2010) say it is still unknown whether neuroimaging can prospectively reveal whether an advertisement will be effective and that single-region readouts are probably too simplistic to be useful in real-life settings.

  InterpretationSource: [4]Where: Revealing hidden information

People involved

• Gregory Berns. Neuroscientist, lead author
• Sara Moore. Co-author

In the Psy101 course

• OpenStax Psychology 2e, section 2.3: Analyzing FindingsChapter 2, Psychological Research. Method. Nearest related section; it does not discuss this topic by name.Nearest related section: it covers correlation and prediction, statistical significance and replication. It does not discuss brain imaging studies.

This is course relevance only. It is not evidence for the historical or corporate claims on this page.See the full course map.

Sources and inspection

4 of 4 sources were opened and read at the level shown. A source marked as not verified supports no claim on this page by itself.

1. [1]
   A neural predictor of cultural popularityBerns, G; Moore, S. Journal of Consumer Psychology 22(1), 154-160 (author's personal copy hosted at gwern.net), 2012. Journal article. Primary source.Relevant text checkedRead all pages of the author's manuscript copy as extracted text and page images: abstract, introduction, methods, results, discussion and references. The copy is not the publisher's typeset version.
2. [2]
   A neural predictor of cultural popularity (Crossref record)Berns, G; Moore, S. Journal of Consumer Psychology 22(1), 154-160 (Crossref record), 2012. Journal article. Primary source.Abstract or publisher record onlyRead the Crossref record, which gives the abstract and bibliographic details.
3. [3]
   From neural responses to population behavior: neural focus group predicts population-level media effectsFalk, E; Berkman, E; Lieberman, M. Psychological Science 23(5), 439-445 (PubMed Central), 2012. Journal article. Secondary account.Relevant text checkedRead the open-access article text on PubMed Central as returned by the fetch tool: participants, campaigns, rankings and the citation of Berns and Moore.
4. [4]
   Neuromarketing: the hope and hype of neuroimaging in businessAriely, D; Berns, G. Nature Reviews Neuroscience 11(4), 284-292 (PubMed Central), 2010. Journal article. Critique or replication.Relevant text checkedRead the first 100,000 of about 115,000 characters of the open-access article as returned by the fetch tool. The article does not mention the Berns and Moore study.

Access note: The paper was read through the authors' manuscript copy hosted by a third party, not the publisher's page. No replication of the song-sales result was found or searched for.

Independent source review: Claude review agent, independent of the research agent, 2026-10-21. Verdict: pass-with-edits.

Related records and elsewhere on Psy101

• The iPhone op-ed and reverse inference in neuromarketing
• The MusicLab artificial music market
• Ernest Dichter and motivation research