Fantastic Web 3 Advocates and Where to Find Them - Part 2 of 2

Part 1

Part 2

3. Fantastic Web 3 Advocates and Where to Find Them

4. Key Takeaways and Concluding Remarks

Important Disclaimer

This content is provided for informational and educational purposes only, and should not be relied upon as legal, business, investment, or tax advice. Furthermore, references to any securities or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Factomind cannot be responsible for your use of the information provided in this content. Factomind has established, maintained, and enforced strict internal policies and procedures designed to identify and effectively manage conflicts of interest related to its business activities.

3. Fantastic Web 3 Advocates and Where to Find Them

We measure influential factors and interaction changes for the sample projects dependent on the shoutouts from advocates, and we deem the shoutouts are ‘effective’ if the numbers are meaningfully higher. Or in other words, if shoutouts from an advocate on a project A led more people to interact with the project A’s Twitter than before, then we conclude that the advocate was ‘helpful in building a community.’

The question of “why should we run Twitter accounts” and “why should we get shoutouts from third parties” were already discussed in the last paper, so for this time, we would like to focus mainly on the two topics:

“Who” should get advocacy shoutouts?
“From whom” should they get advocacy shoutouts?

Throughout the paper, we will spotlight which metrics should founders look for when deciding marketing partners, along with our subjective recommendations in advocate selection.

3.1. “Who” Should Get Advocacy Shoutouts?

We largely have three criteria for projects in need of shoutouts.

3.1.1. Undiscovered Projects

Projects which are undiscovered by the majority should use a hand from advocates. Even common sense tells us that an influencer spotlighting an obscure protocol with a $10M MC draws much more comparative growth in attention to the project than spotlighting Bitcoin, for example.

Table 2 – Influential Factors per Project Status

_{* Undiscovered : Relatively unknown with sub-40 organic interactions and below-5,000 bot-adjusted followers
** Mainstream : All projects except ‘undiscovered’ projects}

The growth in Twitter interactions are more effective for undiscovered projects than mainstream projects, especially in the influential factorD1, so of course it is better to do paid marketing in the undiscovered phase rather than in the mainstream phase; a view in alignment with common sense.

We defined the ‘undiscovered’ status as having sub-40 organic interactions. Followership numbers are tainted with bots thus it is difficult to draw a round-up figure, but usually they revolve around 1,000 to 5,000 followers adjusting for bots. Price-wise, they would typically be south of $100M in FDV.

On the other hand, the ‘mainstream’ status would have above-130 organic interactions.

Note that being ‘mainstream’ is not equivalent to being ‘successful’; it is merely a state of NOT being ‘undiscovered.’

3.1.2. Sectoral Differences

Some projects, because they are more community-friendly, are more sensitive to advocacy shoutouts slightly more than other projects.

Table 3 – Influential Factors per Project Sector

3.1.3. Timeframe

Projects that began in the bear market actually have a competitive edge in building a community. When bears are striking, investors revert to quality projects that could endure prolonged market depression. We define the bear market as all time periods succeeding May 2022. Below is a table on the effectiveness of shoutouts before and after the bear market.

Table 4 – Influential Factors per Project Foundation Date

* Bear : All time periods ensuing May 2022

Shoutouts are, somewhat surprisingly, more effective during the bearish environment. Note that despite the differences in the influential factors, the interaction changes are barely different; as though some people are leaving the market for now, and the survivors are leaning towards shoutouts.

To sum up, if a project is undiscovered, is an NFT/Game project, and has started building during the bear market, it is advisable to get shoutouts from advocates.

3.2. From Whom Should They Get Shoutouts?

3.2.1. The Characteristics of Good Advocates

The desirable attributes of partner advocates are: High Followers, frequent Postings, and being affiliated with Major L1s or VCs.

To begin with, we already defined ‘good advocates’ as those that cause a high growth rate (influential factor) or high growth (interaction change) in interactions of projects’ Twitter timelines. We regressed Influential FactorD1 and Interaction ChangeD1 on other characteristics of advocates to see which characteristics are connected to advocates being ‘good.’

For the exact method of regression analysis and collinearity tests, please refer to Appendix.

Table 5 – Summary Statistics for Interaction Change_D1

* Major L1 : Polygon, Arbitrum, Optimism, Solana, Avalanche, BNB

Table 6 – Summary Statistics for Interaction Change_D1

For the influential factor_D1, the significant factors were found to be Follower, Listed, Posting, L1, and VC. Although the p-value for Listed is significant at 10% level, there is no way to check for the Listed counts without using API so we would not consider it as an important criterion.

For the interaction change_D1, the significant factors were found to be Follower and Posting. Please do remember the Posting variable above is NOT equivalent to shoutouts.

Overall, it is advisable to find advocates that

– has lots of followers
– posts often (spoiler alert—70% of VCs fail at this characteristic!)
– are affiliated with Major Layer 1s or VCs

Nonetheless, the first factor—follower count—could be easily manipulated, so it should always be taken with a grain of salt.

Also note that the ‘Major L1’ factor above only comprises a very specific set of L1s—namely Polygon, Ethereum L2s (Arbitrum and Optimism), Solana, Avalanche, and BNB (excluding Binance CEX), and discards all other alt-L1s. If we embrace all L1/L2s, of which we will give a comprehensive data below, the factor becomes quite insignificant, implying the community powers of L1/L2s are asymmetric.

Nonetheless, the first factor—follower count—could be easily manipulated, so it should always be taken with a grain of salt.

3.3. L1/L2 Advocate

3.3.1. The Influence Map for L1/L2s

Polygon has the strongest community influence, closely followed by Ethereum L2s and Avalanche.

The power of community is always a consideration for founders to choose L1/L2 partnerships, but they yet have no quantitative metrics to gauge the performance of the respective partnerships, or to evaluate whether the L1/L2 foundations actually help with building a community.

From hereunder, we provide a cheatsheet for the influence factors of L1/L2s. We introduced three conditions to filter L1/L2 tickers:

having more than 10 shoutouts to rule out sample size effects (Aptos, Sui);
existence of a specified entity effectively controlling and running the chain for business (Ethereum);
being free of controversy (Waves). Through the above factors, we managed to prune down the original sample of 26 chains to 8 chains as shown below.

Please keep in mind that we at Factomind are not affiliated in any way with the L1/L2s mentioned below, nor recommending L1/L2s in the order; Figure 2 is a cross-sectional data for a very specific characteristic of the chains’ components: Influential factors_D1.

figure2 – The Influence Map for L1/L2s

* Please do note that the data is as of Nov 2022.

If a Chain is plotted further to the right, it does shoutouts more often and is more supportive; if plotted further to the above, its shoutouts are more effective. Having both qualities are desirable, but having one is good enough.

We separated Binance CEX from BNB Chain as these two are isolated entities, and interconnections regarding listings are very weak.

In short, the results are straightforward:

– Quality Shoutouts : Polygon and Arbitrum
– Frequency Pick : Avalanche and Solana

Putting Binance CEX aside, we can easily deduce that Polygon, Avalanche, Solana, and Arbitrum should be the favorites in terms of community strengths.

Among these four candidates, shoutouts from Polygon and Arbitrum are more effective than the other two, yet at the same time they do shoutouts less frequently than peers.

Avalanche, although their shoutouts may not be the most efficient, still they most wholeheartedly support all onboarded projects and churn out marketing materials for partnering protocols. The similar story goes with Solana as well.

3.2.2. The VC Chain

Solana was once notorious for being ‘the VC chain,’ a terminology coined to jeer at the artificially boosted chain activities at the expense of VC capital. While this paper is not about the appropriateness of the terminology, but is about whether there actually exists a VC chain—or, an L1 or L2 that is more likely to be advertised by the VCs.

Table 6 – VC Shoutout Percentage per Chain

* Please do note that the data is as of Nov 2022.

The VC Shoutout % metric indicates the proportion of VC shoutouts to total shoutouts including retail influencers. A higher percentage shows the chain is more advocated by VCs rather than the retails.

Solana actually turns out to be a VC-preferred chain, but only by a marginal amount. Or rather, Polkadot easily tops the list ahead of Solana; the result is not that astonishing considering the recent debate on the chain activities on Polkadot, implying that not that many retail users are actually involved with the chain yet.

Figure 3 – The Polkadot Debate

Please keep in mind that Table 6 by any chance does not represent the highly ranked L1/L2s to be more likely to get VC funding; it is just a proxy indicator for Twitter activities.

3.2.3. The Sectoral Differences

Other than VCs, shoutouts from the foundations of L1/L2s are again good sources of community interests. Yet, out of 50 self-claimed DeFichains and Gamechains, which ones are truly supportive and strong in the respective communities?

If the entire activities of a chain is dependent on a single project, and thus the foundation cannot spend any resources for your own DeFi project, that L1 may not be a suitable choice for you even though it has the highest TVL. Similarly, it might not be desirable to launch a DeFi product on a prominent Gamechain.

Table 7 – The Influential FactorD1 Ranking Table for DeFi and NFT/Game

Polygon stands out in both sectors, being coherent with its current reputation as ‘the best BD chain.’

For DeFi projects, partnering with Ethereum L2s would be the best way (aside from Polygon) to maximize the effect of L1 advocacy marketing, which is natural if we take into account the diverse DeFi projects popping up in Arbitrum such as GMX and its derivatives.

For NFT/Game projects, Avalanche and Solana have been the most influential in terms of influential factors (aside from Polygon), so they might be worth a call if you are running a NFT/Game studio.

3.4. VC Advocate

3.4.1. Twitter Activities of VCs

Just like we did for L1/L2s, we could draw the same map for VC advocates and examine the picture. Nonetheless, there is an obstacle in the course—not every VC aggressively operates their Twitter accounts. In fact, only 30% of the VCs, regardless of being major or not, help with building Twitter communities.

Table 8 – Twitter Activities of VCs

* As mentioned in 2. Data Specification, a Twitter account explicitly mentioning a VC’s twitter handle in its description

Even if founders get fundings from the so-called A-tier VCs, 70% of the VCs will not actively do shoutouts for the funded projects. In this case, it is suboptimal to just stay idle after all the efforts, as getting big VCs have as much impact on the community as minor VCs if the news is not spread.

If the investors are the 70% non-Twitter-active VCs, founders themselves should indirectly advertise on their Twitter accounts that they got fundings from those A-tier VCs. The community impact would be less than direct shoutouts, but a small impact is better than null.

If the investors belong to the active 30%, they would happily do shoutouts for the founders at the cost of polite requests, or even better, they would proactively do so and define the investment theses on the projects, fomenting the growth of the projects’ communities.

3.4.2. The Influence Map for VCs

Due to conflicts of interest, we masked most of the VC names in the graph above; please refer to us personally to discuss the unmasked version.

Figure 4 – The Influence Map for VCs

Simply spamming the Twitter timelines would be worthless if those messages cannot influence the crowd, so we would want to partner with 1) VCs that actively do shoutouts for partner projects (further to the right) and with 2) VCs that can earn significant growth in interactions through their shoutouts (further to the above).

Generally, partnering with the VCs near the border would be beneficial to community growth of projects.

3.4.3. VC Tiers vs VC Community Tiers

Currently, founders judge VCs based on their reputations. Although vague, strong reputations from backers could grant founders powerful tools for advertising. However, some VCs cannot live up to their names, sometimes because they are completely inactive on SNS, or sometimes they are reluctant to publicly announce their connections to certain projects. Either way, their reputations do not truly represent their support in building a community.

Table 9 is a brief table on whether the reputations of VCs actually coincide with their influential factors. Bringing in A-tier VCs does not necessarily guarantee strong community feedback, but bringing in D-tier VCs certainly makes building a community much harder, even if they are committed to supporting the projects.

Table 9 – The VC Reputation-Community Matrix

* Reputation Tiers are calculated based on @taka_eth’s Crypto VC List

** Community Tiers are calculated based on Influential Factors_D1

Long story short, getting good VCs as backers to boost your legitimacy is still important regardless of their support in advocacy marketing. If they do it for you, it is just better.

3.5. Retail influencers

3.5.1. Mega-influencers

The ranking goes: Mega-influencers > VCs, L1s > Retail Influencers.

Actually, most advocates are retail influencers. Although we defined VCs and L1/L2s as broadly as possible by labeling advocates affiliated in any way as VCs and L1/L2s, we still have 2,934 retail influencers in our data set and only 868 for VCs and L1/L2s together.

Generalizing retail influencers into a single statistics is quite challenging as their characteristics are much more diverse; according to our metrics, @elonmusk with 120M followers is tagged ‘retail,’ as he is not affiliated to a crypto institutional entity, and a random guy with 400 followers is also tagged ‘retail’ as well.

Again, followers are manipulable and thus cannot be a determinate yardstick, but we use the term ‘retail influencers’ to generally refer to individual Twitter accounts with between 783 and 38,409 followers, which would be the interquartile range of all the ‘retail’ influencers in our data set. We intentionally narrowed the follower range to distinguish between mega-influencers and ordinary influencers.

Table 10 – Retail Influencers vs VCs, L1/L2s

* Mega-Influencers : Top 10% retail influencers (> 151,043 followers)

Retail influencers, or those who are not in any way affiliated with VCs or L1/L2 foundations, have much lower influential factors. This fact could be attributed to many reasons, but we will resort to only two.

First of all, retail influencers are more inclined to ‘undiscovered’ projects. Such undiscovered projects of course are less legitimate, less funded, and less managed; therefore the impact of the advocacy shoutouts are naturally much smaller. Out of 16,864 shoutouts from all the advocates, VCs or L1/L2 foundations did 3,506 shoutouts for ‘mainstream’ projects, but did only 542 shoutouts for ‘undiscovered’ projects.

Even if VCs/L1s and retail influencers write for the same project independently, retail influencers are usually limited in ability to carry out due diligence for the advertised products; thus, the broader audience must approach retail shoutouts much more cautiously, resulting in lower influential factors.

However, mega-influencers behave differently. They have lower influential factors than VC, L1 advocates, probably as they only do shoutouts for legitimate projects to protect their reputations, but their influence on the interaction changes are unmatched; they might well be dubbed ‘community whales.’

4. Key Takeaways and Concluding Remarks

For discussing L1/L2 partnerships, the rule-of-thumb would be Ethereum L2s for DeFi, and Polygon for Game. Nonetheless, other chains, notably Avalanche, are still in the list considering their devotion to support the onboarded projects.

Of course, the numbers are never concrete, and we remind you again that we removed multiple supportive L1s just because they are young projects, or they underwent some controversies.

Once again, we are not affiliated with any of the abovementioned L1/L2s.

Other than sectoral differences, another deciding metric could be the current size of the projects’ communities. We could measure the size by using ‘organic interactions,’ and we could divide the organic interactions into three segments:

– ‘Undiscovered’ : <40 organic interactions
– ‘Neutral ’ : 40
– ‘Mainstream’ : >120 organic interactions

If a project is in an ‘undiscovered’ state, it should go for advocates with high interaction changes rather than high influential factors—quantity over quality.

If a project is in a ‘neutral’ state, it should go for advocates with high influential factors as the initial bootstrapping of a community has kicked in, but it still requires enough fuel to grow further.

If a project is in a ‘mainstream’ state, it could use some help from advocates, but it is not a must.

For the full version with the comprehensive unmasked data, please contact us personally.