Incentives for Full Supernodes

Any network needs to be stimulated to get off the ground. A simple marketplace network might require bringing in sellers when there are not enough buyers, or bringing in buyers when there are not enough sellers. All the prominent networks of today – Uber, AirBnB, eBay, Amazon – faced this issue and were able to overcome it by tapping into existing users or providers of similar services, and by creating extra startup incentives for people joining their platforms.

GRAFT is no exception. It is a complex network with many participants – users, miners, full supernode operators, service brokers, proxy supernode operators, merchants, application developers. Having so many players is both a blessing and a challenge.

Miners represent the first layer of the network (settlement), and they are already incentivized through both network transaction fees and block mining rewards, which do not depend on the number of settled transactions and provide a steady income for miners. Beyond the miners, the next group without which the network will not function are the full supernodes which are critical for real-time authorizations (RTA) and the pay-in/pay-out facilitation.

Full supernodes get paid for performing validations for the network and preventing double-spending while processing authorizations in real time. Their income depends exclusively on the number of transactions that they get to work on as well as the total number of full supernodes in the network. Full supernodes are chosen randomly in each block, with two selected from each of the four tiers for a total eight supernodes, which equally split the 0.5% fee of the transaction they authorize. The problem, however, is that a small transaction volume of the brand new payment network would result in low initial full supernode income, and thus little incentive to run full supernodes before the network volume ramps up.

The good news is that there is a fairly easy way to incentivize the full supernode hosts without resorting to simple airdrop-type incentives or block reward sharing. What we will do instead is send enough RTA transactions across the network to provide a healthy transaction volume until the network is fully ramped up. This way we reward full supernodes for the real work they are supposed to do – validating transactions – unlike most other second layer reward models where masternodes receive passive income for just “being there” (which does little to promote network robustness or self-optimization).

The incentive program is designed to maintain a robust daily transaction volume until merchant-generated RTA transactions reach that level on their own. The number and size of stimulus transactions will depend on overall transaction volume and will be reduced gradually as the network gains momentum. We estimate the cost of this program to range between 50 and 100 million GRFT.

As with any new network, especially one with lots of participant types, the GRAFT Network needs to be stimulated to maintain a level of involvement ensuring stability of the network and availability of network services. We believe that the most critical part needing extra stimulation at this early stage is the full supernode layer. To stimulate the network as a whole, however, we are also considering additional groups of participants that may require stimulus going forward. For example, buyers may need to receive cashback incentives (similar to some credit cards) in order to choose GRFT over other methods of payment available at the checkout.

As always, we’re open to other suggestions from the community.