Photo by fabio on Unsplash

Delegating on The Graph: Choosing an Indexer

(A summary of various resources to analyze indexers on The Graph)

Ian Evans
5 min readJul 2, 2021

--

Disclaimer: This is not financial advice, and I am not a financial advisor. This is for educational purposes only and all readers should do their own research and due diligence.

For those that want to put their GRT to use but are not savvy enough at coding to be an indexer, there is the option to delegate the GRT to an indexer to earn passive income. What this means is essentially you can give your GRT to an indexer for them to use in securing and running their subgraph. In return, they may split the query fee and the rewards with the delegators (if you choose your indexer wisely, more on this below). So you have your Metamask setup and then you head over to The Graph’s network and you see all the indexers that you can delegate to but then the big question looms, which indexer do I pick?

Unfortunately, the technical nature of understanding indexers and researching them has muddled the line between who to trust and who not to trust because there may be a stark difference in the passive income that you, a delegator, receive depending on which indexer you go with. When it comes to vetting indexers, one should look to these four resources to perform indexer due diligence:

https://network.thegraph.com/participants/

https://thegraphportal.com/indexer-profiles/

The Graph on Discord

https://graphlets.io/

By utilizing these four resources one can learn more information about an indexer before delegating them GRT. Thankfully a fail-safe mechanism within the GRT protocol makes it so a delegator is not punished if an indexer acts nefariously, so we, the delegators, don’t have to worry about slashing and burning as penalties, however, we do have to worry about indexers who are taking a larger cut than they should be and if the indexer does act nefariously this may hurt future query fee volume and rewards because that subgraph will not be trusted to be used by developers as a source to query from.

Using each Resource:

https://network.thegraph.com/participants/

We must consider the proportion of the indexer’s delegation to the delegator’s delegation. This generally gives us more meaning and understanding of the varying query fee percentages because indexers will want to tweak this percentage to reflect the amount of their capital being used. For example:

If we look above at figment-lemniscoop.eth vs p2p-org.eth, we see that the query fee cuts and effective reward vary greatly. The indexers set the fees but to properly interpret these percentages we need to look at the indexer’s stake owned compared to the delegated amount and the total allocated amount. Here we can see that p2p-org.eth only owns 36.21M of the total 504.39M that is allocated so they own 7.2% of the allocated amount and therefore are taking a smaller indexing fee compared to figment-lemniscap.eth which owns 74% of the total allocated amount and this is reflected in there 76.6% query fee. Furthermore, the spread between the query fee and the proportion of stake owned by the indexers is interesting to observe. This spread for Figment-lemniscap.eth is about 2.6% (76.6%-74%) while p2p-org.eth has a spread of 7.42% (14.62%-7.2%).

https://thegraphportal.com/indexer-profiles/ and The Graph on Discord

Both the indexers’ profile breakdown, and discord group for The Graph, give us insight into the personal or “human-side” of the indexers. Using the URL we can read summaries about the indexers, their experience, and their goals. By entering the indexer’s wallet address we can find their page where they describe themselves to the delegators. The lack of an indexer’s profile on this page is something to consider before delegating to an indexer. Below is an example of an indexer’s profile:

Furthermore, we can do better by also looking at the names of the indexer profiles and using these names, not the wallet addresses, to search within the Discord channel. By researching and finding the indexer in the Discord channel one can read their past posts and how they have interacted with The Graph community to date. An interesting conversation amongst indexers on Discord below:

https://graphlets.io/

Lastly, this link shows the historic ROI that delegators can expect to receive. One just needs to input the wallet address of the indexer into the search bar and they will get back information regarding historic ROI delegators received. Other important information on this page to consider is the time they’ve been indexing for as this adds to the indexer’s reputation.

Here we can see that I entered the wallet address for p2p-org.eth and was able to see that the return on investments based on delegated funds has been around 6.35%.

Other factors to consider about delegating:

Another caveat is that delegators need to consider the projects themselves that the indexer is indexing subgraphs for because this will also influence the total amount that they receive to delegate and the query reward fee structure. Subgraphs for Defi protocols that will become popular may see a rise in query fees compare to other indexers indexing other subgraphs, so this is also something to consider when deciding which indexer to go with.

Using these tools together to complete indexer due diligence should allow a delegator to make a more educated decision on which indexer to delegate to. The Graph’s ecosystem is way more complex than I previously thought and I’m sure there are more contributing factors out there and tools to learn about and use, as well as more to learn about the underlying protocol as it is evolving and growing. For more in-depth explanation The Graph foundation has a forum page, which is the first link, and these are other links I found useful too:

So which did I pick and why? I chose the delegator suntzu because of memetic reasons

https://thegraphportal.com/choosing-an-indexer/

https://stakingfac.medium.com/the-graph-delegator-cheat-sheet-2a16c1243513

https://stakingfac.medium.com/the-graph-staking-guide-5ec1455f4783

--

--