Analyzing fan sentiment of music NFT drops

tl;dr: With the goal of developing a more objective, big-picture understanding of the state of fan sentiment around music NFTs, we developed a timeline mapping fans’ reactions to hundreds of music NFT drops on Twitter between October 2020 and November 2021. In stress-testing commonly-held industry assumptions about what social-media messaging tactics do or do not work with music NFTs, we found that Twitter has not proven to be an effective platform for artists to address fan concerns around NFTs and move the needle on sentiment relative to the general trend. We also found that Twitter data ultimately does not provide a complete picture of a fan’s experience with music NFTs, especially from a long-term onboarding perspective.

This is Part V of a five-part, collaborative research report that the Water & Music community has put together over the last two months on the state of music and Web3. Contributors to this research thread on fan sentiment analysis of music NFT drops are listed at the bottom of this page, sorted by role. You can view the current state of our report rollout, and a full list of our member-contributors, by visiting

Read our previous installments here:

I. Will music NFTs ever get their PFP moment?
II. Defining music NFT ownership, from the digital to the analog world
III. The state of music/Web3 tooling for artists
IV. Behind the headlines: Analyzing Web3 onboarding strategies for music fans

Despite the astounding amounts of money that artists have raised through music NFTs in the last year alone — and the hope that parts of the industry have placed in them as the future of creative monetization — NFTs remain the subject of much debate. Mixed reactions to music NFTs are commonplace not only among industry professionals but also among music fans themselves, with key concerns ranging from environmental unsustainability to high financial barriers to entry and the perception of being scammed.

Given that fans are a crucial source of support for artists across all forms of consumption, it is critical to understand their higher-level attitudes and reactions towards this emerging medium. The reason for this is simple: Fan sentiment has a significant impact on the choices that artists make. Psychologically and financially, many celebrities’ entire livelihoods and self-perceptions are innately tied to the feedback they receive from their fanbases. It’s no surprise then that an adverse reaction to an NFT drop doesn’t only affect an individual artist — it projects a message to other consumers and artists alike, informing their perception of the viability of this new trend.

In 2021 alone, we have seen multiple examples of negative fan reactions impacting artists’ NFT strategies. Jacob Collier and A.C.E., both beloved by their fans, have pulled back entirely on their NFT drops, mainly owing to fans’ concerns about the environmental impact of blockchain technology; DistroKid Sellouts faced similar backlash. The list of poorly received drops goes on featuring the likes of Charli XCX, Interpol, Crazy Frog, Method Man and A$AP Rocky. In response, many artists have removed the term “NFT” entirely from their marketing copy, opting instead for alternatives such as “collectibles” or “digital art.”

In light of this ongoing challenge, we were interested in developing a more objective, big-picture understanding of the state of fan sentiment around music NFTs. To that end, we have developed a timeline mapping fans’ reactions to hundreds of music NFT drops on Twitter between October 2020 and November 2021. This analysis aims to piece together a bird’s-eye view of fan sentiment around NFTs, study how this sentiment might have changed over time and identify any underlying patterns driving positive or adverse reactions. This piece is also a large-scale, quantitative, data-driven companion to our piece on fan onboarding strategies, which has a more qualitative focus guided by in-depth surveys and case studies.


We chose Twitter as the primary source of data to build our fan sentiment timeline. Besides being a massively extended social network that hosts tens of millions of artists and fans, Twitter is an incredibly active gathering place for Web3-native thought leaders and enthusiasts. It, therefore, seemed like the best single source for a large dataset that could accurately capture fan sentiment around music NFTs.

The Water & Music community has curated a seed dataset of over 200 tweets announcing music NFT drops. Our dataset focuses mainly on electronic music (roughly 75% of our list), and covers other genres including hip-hop, K-pop and rock. For each NFT drop tweet, we also sourced all of the reply and quote tweets, resulting in a total dataset of about 6,000 tweets.

We then aggregated these tweets to produce two descriptors for each NFT drop:

  1. A % positive sentiment value, ranging from 0 to 100, with 0 representing no positive sentiment and 100 representing a project that received only positive sentiment. To do this, we obtained a positive/negative sentiment flag for every tweet in our dataset by applying a Twitter-specific sentiment analysis library to the tweet text. We then computed a single sentiment value for each drop representing the percentage of positive-sentiment tweets across all of the tweets that quoted or replied to the initial drop tweet.
  2. An interaction count, defined as the total number of likes and retweets received by the initial drop tweet and all subsequent reply and quote tweets.

Utilizing these data points for each NFT drop, we were able to plot our timeline. A preliminary version is pictured below:

Each dot in this graph represents an NFT release. The horizontal axis represents the date of the announcement, and the vertical axis represents the percentage of interactions that registered positive sentiment. The size of each dot illustrates overall engagement on Twitter, and its color categorizes the genre of the artist, as pulled from the music analytics tool Chartmetric.

We also used the media monitoring and social listening platform Meltwater to extract demographic and online behavior data as a complementary source. We handpicked eight high-profile NFT drops to run through their tool, based on their engagement values and overall market relevance. By looking at some of these relevant examples, we aimed to uncover some of the factors that might correlate with more positive or negative sentiment around NFT drop announcements for fans.


Perhaps the first thing that stands out when observing our timeline is the uneven spread of NFT drops over time, mainly concentrated in February and March 2021. This period is followed by roughly six months of considerably low activity, after which things start to pick up again in October. This trend is consistent with the  notion of a perceived “NFT boom” in early 2021, followed by a period described as the “crypto winter” that summer. It is also consistent with our reporting earlier this year on monthly music NFT sales dropping 90% in Q2 compared to Q1.

The fan sentiment values themselves are spread across most of the spectrum, with a heavier concentration of neutral or slightly positive values. This is roughly the case for both the first wave in early 2021 and the uptick in recent months, and goes a bit against our initial hypothesis that there would be more negative fan sentiment, especially around bigger artist drops.

Looking into specific genres, we see that electronic music — the most represented genre in the dataset (and in the music NFT market at large, by sales) — follows this general trend, with most drops having neutral to slightly positive sentiment.

Hip-hop follows a slightly more positive trend, although the genre has produced only a handful of drops in the last few months.

Engagement buckets

One of our initial intuitions was that NFT drops from artists with smaller, more targeted fanbases, and therefore with lower total engagement numbers, might have better reactions from their fans. We hypothesized that negative sentiment towards an NFT drop is more likely when an artist looks to monetize a larger, disconnected audience, whose understanding of NFTs might be more superficial quickly. Conversely, we suspected artists with niche but more carefully developed audiences might be better positioned to onboard fans onto Web3 and build a more cohesive narrative around the NFT drop as a meaningful contribution to their creative work.

However, our findings don’t support this hypothesis. We split the NFT drops into three equally large buckets according to their engagement values (most, medium and least engagement). As pictured below, we plotted the fan sentiment distributions for all three buckets side-by-side. In this plot, the horizontal axis represents the % positive sentiment value, while the vertical axis represents that sentiment value’s likelihood.

We found that fan sentiment values are distributed very similarly for all three levels of interaction analyzed. That is, an NFT drop with high levels of interaction from fans is roughly just as likely to have a positive reaction as a drop with shallow interaction. In other words, with our data in hand, we cannot say with confidence that smaller artists’ NFT drops are better received.

A caveat to this reasoning is that it essentially equates total interactions with audience size. Namely, we’re assuming that the larger an artist’s audience, the higher the total number of likes and retweets will be for both the initial NFT drop tweet and any subsequent replies and quotes. But that might not always be the case: Artists with a smaller but more developed audience might have a higher engagement rate, while artists with a larger but less targeted audience might have a comparatively lower engagement rate. While our analysis shows that overall engagement with a drop on Twitter isn’t a good indicator of the sentiment of those replies, we’re limited in our ability to assess the relationship between audience size and the success of an NFT drop given the imprecise splitting of releases by audience size.

Addressing fan concerns has little impact on sentiment

In the fan onboarding chapter of our collaborative report, we present a four-part framework for understanding fans’ top concerns with Web3 — arranged on a spectrum from which concerns artists have the least ability to address themselves to those that artists are most able to address and mitigate in their own onboarding approach.

Environmental concerns and high financial costs were among the top mentioned concerns for fans in our research, especially for those who were not Web3-literate. Interestingly, this also lines up with our tweet data. We split our dataset into two buckets of positive-sentiment (defined as 50% or more) and negative-sentiment drops, and pulled the top 30 keywords that appeared in fans’ Twitter interactions; the words “environment” and “money” are among the top themes on the list, alongside a lot of other colorful language.

Unfortunately, these also happen to be the two problems that artists arguably have the least amount of agency to address themselves. The underlying issues behind these concerns relate to the limitations of blockchain networks as a whole, for which it would be unreasonable to hold artists responsible.

Further, our sentiment analysis shows that when artists do try to address issues of environmental concerns and financial barriers to entry in their tweets or press releases, it tends not to impact sentiment in a meaningful way — perhaps further exacerbating this feeling of artists not having control over the situation.

To complement our qualitative framework of top fan concerns above, we scanned all the NFT drops in our dataset (both the official announcement tweets and the original project pages) for both mentions of charitable and environmentally friendly causes or intentions.

In the context of our dataset, mentions of charitable causes or intentions have little to no impact on sentiment relative to the general trend, as pictured below. In fact, a handful of NFT drops with charitable elements actually registered negative sentiment — such as Charli XCX’s drop on Foundation back in March 2021, which donated a portion of proceeds to Girls Make Beats. This sentiment roughly matches the broader music industry (and within the Water & Music community) that merely tacking a charitable element onto an NFT rollout will not do enough to quell the entire range of fan concerns about the format, and certainly does nothing to address inherent financial and technological barriers to entry that new crypto users will face.

When we tag NFT drops/projects by mentions of environmentally conscious or carbon-neutral initiatives, the average positive sentiment seems to lean slightly higher than the general trend (see below) — but we have too little info on carbon-neutral drops in our dataset to know for sure whether this difference is statistically significant. And similar to the charitable tack-on issue discussed above, leaning too heavily on sustainability rhetoric in marketing an NFT drop risks framing sustainability alone as the drop’s only utility, at the expense of the quality of the actual fan experience. As we cover in our fan onboarding article, much of the surface-level messaging around environmental concerns in NFT drop tweets and press releases glosses over additional externalities and barriers to entry that fans might face as they onboard into Web3, such as needing to purchase ETH on Mainnet and bridge it over to a secondary, lower-cost chain (like Polygon or Tezos) in order to purchase an otherwise “carbon-neutral” NFT.

The importance of fan backgrounds in driving sentiment

Capturing a complete picture of fan sentiment around NFT drops arguably requires analyzing both push and pull factors from a fan perspective. Namely, many of the above charts are gauging fans’ reactions to top-down messaging from the artist, particularly around carbon-neutral or charitable initiatives. However, to make this data actionable, we also need to understand these fans’ bottom-up motivations and backgrounds, which might inform their propensity to engage openly with Web3 in the first place.

In this vein, we used Meltwater to pull demographic, psychographic, and interest-based data on the audiences who engaged with some of the year’s highest-profile, highest-earning music NFT drops. An immediate takeaway from our analysis is that the core audiences for several high-profile music NFT drops on Twitter were already “Web3-native” — in the sense of using self-identifying terms like “collector,” “crypto,” and “NFT,” and actively following several high-profile crypto accounts. In other words, they were avid crypto enthusiasts rather than your “typical” music fan.

Let’s look at the overall audience data for 3LAU’s Faces drop on Nifty Gateway as an example (tweet announcement link here). 3LAU has several drops featured in our dataset with a disproportionately high positive sentiment, averaging at around an 80% positive rate. This is not surprising as 3LAU could be considered a “crypto-native” artist by multiple measures; he has been involved with blockchain from a very early stage (2018 at the latest), and, as a former finance major in college, is comfortable talking openly with his audience about the ways that music, finance and crypto intersect.

The Meltwater audience visualization for 3LAU’s Faces drop, pictured above, shows clear audience clusters that imply a more positive or at least more engaged relationship with crypto. Names of these clusters include “trade/accounts,” “nft collectors/degen,” “opensea/creative” and “investor/ape.” Similarly, the top keywords these audiences use to identify themselves on Twitter lean heavily on the crypto art landscape (29.5% for “NFT,” 15.0% for “NFTs,” 10.7% for “collector,” 10% for “art,” and 9.3% for “crypto”).

In the Affinity section of the visualization, the most relevant accounts that the audience for this drop collectively follows are all related to either the crypto or NFT art markets (Elon Musk, Nifty Gateway, Beeple and Pak). Last but not least, the engaged audience for this drop is 78.38% male. At the risk of overgeneralizing, all of these factors almost perfectly fit the bill of the “crypto bro” stereotype — vast majority male, and skewing towards crypto-native self-identifying terms on social media — who is more likely to embrace a given NFT drop with positive sentiment.

A handful of other projects we analyzed had the same combination of positive sentiment and a crypto-heavy audience makeup:

In contrast, the audience engagement around the launch for Grimes’ WarNymph collection on Nifty Gateway in March 2021 (tweet announcements here and here) did not center crypto nearly as much, as is represented in the chart below. This audience’s top keywords included “love,” “music” and “art,” while “crypto” ranked fourth. Each of these keywords was shared by only around 2% of the drop audience — in stark contrast to drops like 3LAU’s Faces and Dirtybird Flight Club, where over 20% of the audience self-identified with “crypto,” implying a heavy Web3-native bent. Furthermore, the WarNymph audience was slightly less male-skewed (63%) than other NFT drops listed above more diverse range of personas emerged from the audience clusters on the right of the visualization, far beyond just crypto and investing (“student/technology,” “culture/politics,” “grimes/stan”).

All of these factors together suggest that engagement around the WarNymph drop was driven more by Grimes’ long-term fan base than by “crypto bros” per se. Interestingly, according to our analysis, this did not dramatically pull down the sentiment values around the drop on Twitter — which was still generally positive, although at a relatively lower rate compared to some of the drops mentioned above (62% positive rate in Grimes’ case, versus 80% for others).

Based on these graphs, we cannot conclude with confidence whether having a more crypto-native audience is a consistent indicator of more positive sentiment for an NFT drop. That said, what we can glean from this preliminary analysis is the highly personalized nature of fandom unique to each artist. Further, we can speculate on the importance of incorporating audience data based on core identities, interests, and concerns for a cohesive fan sentiment model.


Based on our analysis above, we unsurprisingly have emerged with more questions than clear answers. In stress-testing commonly-held industry assumptions about what social-media messaging tactics do or do not work with music NFTs, we’ve arrived at two main arguments:

1. Twitter has not proven to be an effective platform for artists to address fan concerns around NFTs and move the needle on sentiment, especially around environmental and financial concerns. This may be because, at large, Twitter is not exactly the best platform for nuanced discussion of complex topics (who knew…).

2. Twitter data ultimately does not provide a complete picture of a fan’s experience with music NFTs. As we discussed in the intro section of this article, the fear of adverse Twitter reactions alone can certainly dissuade an artist from experimenting with the technology. That said, negative sentiment doesn’t necessarily guarantee poor financial performance for an NFT drop. For example, Charli XCX’s drops on Foundation, which received a wave of backlash from fans, ultimately sold for nearly 20 ETH in total. Nor does positive Twitter sentiment necessarily guarantee a successful long-term fan onboarding and community-building experience around NFTs. Moreover, “success” for an emerging artist may not mean the same thing as success for a more prominent artist, especially regarding social media engagement.

Fan sentiment ultimately comes back to the strength of the artist-fan relationship — a concept at the center of Web3’s role in the music industry — and to fans’ reception of the work in context of the narratives circulating in the world around them. It’s possible to be mindful of externalities primarily outside of our control, such as the environmental cost of blockchain technology, while still prioritizing levers within our control (authentic creation, community building, intention and value), making those the primary focus around a release.


Alexander Flores (A, B, C, F)
Dani Balcells (A, B, D, F)
Ana Carolina Laurindo (C, F)
Brodie Conley (D, E)
Diana Gremore (B)
Raul Guerrero (B)
Henry Chatfield (C)
Cherie Hu (D)
Yung Spielburg (D)
Brandon Landowski (D)
Andres Botero (E)
Oliver Dawson (E)
Levi Downey (E)
Ivo Dimchev (E)
Sam Minkin (E)
Garrett Perez (E)
Katherine Rodgers (E)

(A) Research project leads
(B) Data analysts
(C) Core dataset contributors
(D) Writers/editors

(E) Other dataset contributors
(F) Visualization