Data Analytics in Blockchain

Author

Milos Bojinovic

Published

May 7, 2022

Executive Summary

Data Analytics is the science of analyzing raw data related to a specific problem and extracting all of the necessary information in order to make conclusions about as well as derive approaches for solving it.

In the context of Blockchain, Data Analytics revolves around the process of collecting and parsing of raw transaction data thus transforming it into usable and actionable data. Parsing of those transactions requires knowledge about the chain specifics as well as internal workings of Smart Contract that that are of interest which is an extremely time consuming process - all of the data on the Blockchain, while it may be public and unchangable, is unstructured.

Blockchain data, however, holds all of the chain’s history since its inception, making it possible to see past interactions between addresses and/or Smart Contracts. This data can then be segmented to include only a wanted subset for which the analysis will be performed. With this aggregated information it is then possible to gain insight in the past trends for a set of Non-Fungible-Tokens(NFTs) and Decentralized Finance(DeFi) related applications as well as general Crypto related trends and potentially predict future ones.

Data Analytics platforms in the Blockchain space are gaining users as the whole crypto ecosystem evolves and is becoming harder to navigate. With substantial recent investments in this area,12 it is clear that the investors are showing interest in what these platforms can do and the potential they have in shaping the future of Blockchain.

Introduction

This research focuses on the type of information that can be extracted from the blockchain, elaborates on where that information can be used and signals the utility that lies in it - especially in the NFT and DeFi related applications that dominate the crypto ecosystem.

Out of the scope of the research falls the aquiring/storing/organizing/displaying of the raw data which are extremely hard and complex processes that also open questions about the decentralization of those processes and the integrity of the collected data.

Instead, it casts a light on the current state of the Data Analytics / Blockchain intersection, top platforms that operate in it and the tools that they provide. These platforms/tools are then categorized based on area where they are used and their use cases briefly explained.

Goals & Methodology

Data Analytics platforms/tools enable users to make sense of this rapidly changing space and so the importance of this explorative research lies in the fact that they are relatively new with a huge potential and as such must be investigated further.

The goal of this research is to:

  • explore, explain and categorise Data Analytics platforms and their tools

The methodology of the research includes:

  • taking into account only platforms that support Ethereum Virtual Machine (EVM) compatible chains

  • discussing where a platform/tool can be used

  • categorizing a platform/tool as one of the following:

    • General Purpose - contains both of the other categories as well as some additional functionality
    • NFT specific - focuses on providing information about a collection/NFT or the NFT market in general
    • DeFi specific - focuses on the extracting metrics for specific DeFi projects and DeFi market in general

Results & Discussion

With those three categories explained above, platforms can be investigated and then grouped. Regarding the distribution of categories that will be elaborated on it is evident that most platforms/tools are not general purpose ones. They require a lot of time to develop and so there is an emerging trend of developing specizalised ones that are easier to develop and target a single niche.

In the following text categories are listed in order, with each category being divided into “Tools” and “Platforms” sections. In the “Tools” section, each entry describes a specific functionality and the “Platforms” section focuses on illustrative platforms that have those functionalities (among others) incorporated into them.

Platforms/tools from the same/different categories are not mutually exclusive and can be combined.

General Purpose

Tools

Tools in this category query information about the complete transaction activity of a specific address (both Externally-Owned-Accounts (EOA) and Smart Contracts) and groups the extracted information into human readable metrics.

Contract interactions

For any choosen contract it is possible to extract general information about the:

  • transaction count
  • unique addresses
  • token inflow/outflow

for a specific timeframe. These values can then be organized and monitored over a larger time periods to provide information about the latest trends and changes in the number of users, who these users are, etc.

They can also be used to detect contracts that were recently deployed that are gaining popularity as to investigate the project with which those contracts are associated with.

More valuable information would be tied to how the contract is being used - what methods are being called, their sequence, etc. To extract meaningfull data, as it was discussed above, there would need to exist a parser with a specific domain knowledge.

Address Profiler

For any user address of interest, it is possible to extract the information about the :

  • portfolio (all of the assets that the address holds)
  • estimated portfolio value (sum of values of NFT* and ERC20 holdings)
  • recent token trades (both ERC20 and NFTs)
  • addresses that the user has interacted with

All of these values can also be monitored since the beginning of the chain’s history and addresses can be grouped together to provide some form of live feed for those that are most interesting either to the user or to the platform itself.

*Estimated value of an NFT is platform specific (i.e. see3)

Alerts

Alerts are delivered to the user, via a communication channel of choice (Telegram, Text Message, Discord, …), when a certain customly defined condition is met - some address buys an NFT, collection’s floor price has increased/decreased by some margin, …

Platforms

Two of the top plaftforms in this category are Nansen and Dune Analytics which can be used to gather and analyze similar information but take two drastically different approaches - user oriented and business oriented, respectively.

Nansen.ai

This paid platform doesn’t require or demand from user to have technical knowledge and it provides detailed non-customizable dashboards for General purpose, NFT and DeFi specific tools. Almost all of the tools (from the three categories) listed in this paper, in one form or another, are supported by Nansen making it the most comprehensive and beginner friendly platform.

An interesting additional feature that Nansen provides is labeling of some addresses as being “Smart Money” (addresses that were early adopters and/or have made smart decisions in the NFT and/or DeFi space)*. There exist specific dashboards/sections where it is shown what the “Smart Money” is doing - what are they minting/buying/selling, with whom they are interacting, etc. This information can be used by the user to decide what they think is a good strategy for them when investing and can be combined with custom alerts when a certain condition is met.

*See4 for more details on the labeling of these addresses.

DuneAnalytics.com

Dune Analytics translates raw on-chain transaction data into SQL databases such that the information can be requested using SQL queries. Custom vizualizations (charts, graphs, …) and dashboards can be created from those queries which can then be embedded into other websites.

Additional benefit of the platform is that there exists an active community of members who can create dashboards for which both the visualizations and the SQL queries are publicly avaliable. This enables them to build upon on another’s work, making a powerfull snowball effect.

There exist, however, two drawbacks to the platform:

  1. only the platfrom itself can perfom the parsing of smart contracts (users can request a contract to be parsed)
  2. doesn’t provide an API (though paid users can export results as a CSV file)

NFT specific

Tools

Market Overview/Trends

Contains information about the whole NFT market and specific marketplaces, such as:

  • number of distinct users (minters/buyers/sellers)
  • trading volume
  • average price of all NFTs sold
  • floor price (taking into account all NFTs listed)
  • trending collections

These values are then used to analyze the percentage share of a marketplace compared to the whole market which is useful to determine the top marketplaces and capture the moment when there is a drastic shift in the leaderboard.

The tool also helps in discovering new trending NFT collections and on which marketplace is the most of the trading activity for that collection happening thus answering the question where to go to when considering to invest in it.

Collection Breakdown

Contains information tied to a specific collection and involves:

  • basic information (number of distinct holders, average price, volume, price range, number of trades…)
  • balance changes (how many NFTs were bought/sold/minted by an address)
  • rarity stats - what traits are the rarest and thus more valuable
  • recent mints/trades
  • similar collections

This data can be used to determine whether the majority of NFTs from the collection are held by few addresses which is a bad position for other holders as those addresses can quickly unload the NFTs, selling them at lower prices and so driving the floor price for the entire collection down.

It can also be used to assess the confidence of NFT holders in the collection by seeing if long term holders are suddenly started selling or if the collection is gaining momentum (for example, a lot of trades by different addresses in a short period of time).

All of this can be used by the users to develop unique NFT trading strategies, making this tool extremely informative.

NFT Breakdown

This tool focuses on a specific NFT from a collection and displays

  • history of trades - changes in ownership and price
  • similar NFTs (based on traits)

One of the obvious use case is tracking the price movement of that NFT but another one is tracking at the same time the price movement of similar NFTs and buying those that seem undervalued (ofter reffered as “sniping”).

Platforms

All of the previously listed tools are supported by Nansen but there are specialized alternatives that perform limited subset of those functionalities in a same/slightly different way. Some of the popular ones are icy.tools, moby.gg and NFTNerds.ai.

DeFi specific

Tools

Total Value Locked (TVL) Tracker

TVL s the overall value of crypto assets deposited in a specific DeFi protocol – or in DeFi protocols generally. It is often analyzed to determine the oportunities across chains and protocols. When analyzed over longer periods of time it can help in discovering new project trends.

Recent Activity Tracker

This tool gathers in real time the latest transactions that happened on Decentralized Exchanges (DEXs), Lending/borrowing and Derivatives platforms. Using it, it is possible to detect and take into account large funds movement by an address that is of interest.

Staking/Lending/Liquidity Metrics

These metrics revolve around the number of lenders/borrowers/stakers of a DeFi platform, their current and past balances (including deposits/withdrawals/liquidations) as well as the distribution of token holders.

Platforms

Dune Analytics is very useful in this area as the most useful information is DeFi platform specific and needs to be analyzed in a different way. For general, comparable information there are DeFi Pulse and DeFi Llama.

Conclusion

Using Data Analytics platforms, it is possible to track and analyze the performance of a portfolio for any address and to learn from their past experiences. They are also useful in determining the differences between the most sucessfull and least successful addresses as well as what they have in common. These groups of addresses can then be monitored for future transactions and can alert an user when they happen. Users can then assess the new information and act accordingly.

The bottleneck of this process lies in the human factor that does information assessing manually and so a lot of time is spent on it. Improvements can be made by automating the decision process that leads to the action. Since it would be dealing with real funds, a machine could suggest a potential sequence of actions and the user would still need to aprove it by signing the corresponding transactions.

For example, in NFT trading, a bot could monitor all trades in realtime and detect an increase/decrease in the popularity of a collection which would signal it to buy/sell an NFT or a set of NFTs and suggest the price of it - at what price to list an NFT or what offer to make.

This, and other use cases from specific niches would need to be studied further, as part of a separate research. There exists a question on the utility of those tools, however, their functionalities are in a large measure dependant on the Data Analytics and so their whole processing pipeline would need to be carefully designed.