Intent centric protocols rely on Fillers to facilitate the execution of orders on their platform. One measure of the health of an Intent centric protocol is the number of Fillers that regularly fill orders and the extent to which these Fillers compete with each other to provide the best possible execution quality for user orders. Having a small number (or even one) of Fillers is a significant risk factor to the protocol.

Whilst most protocols are open and encourage free permissionless participation by any entity wishing to contribute bids for orderflow, some may require Fillers to apply and be whitelisted before they can access orderflow. Having a permissioned system increases risks of centralisation and heightens the risk of Censorship and amplifies the consequences of Filler Failure since it is not straight forward to for a new entity step in to provide bids for 'stuck' orders should no Fillers be available.

Fillers often run sophisticated selection criteria to determine what orders they will fill and under what circumstances. Fillers additionally need to own sufficient inventory of the tokens on the appropriate chain to be able to execute an order they are bidding imposing limits on the competitiveness of filling entities based on their treasure and inventory management practices. As a result the range of orders that are filled differ between Fillers and can cause uncontested orders based on a number of factors including order value. It is not uncommon to see a tail off in Filler competition towards the higher value range.

The distribution of tokens filled by each Filler, based on the USD value of orders. It provides insights into which tokens are most frequently provided by different Fillers and highlights any specialization or concentration in specific tokens. A diverse range of tokens across Fillers indicates a healthy ecosystem, while a high concentration in particular tokens or Fillers may suggest potential liquidity or centralization risks.

Competition between fillers is a crucial element in the intent centric architecture design. Monopolistic filling discourages Fillers from increasing their bids to win order flow from other Fillers and results in reduced execution quality for the user. Another key metric of the health of an Intent based protocol is the degree of competition between Fillers to win order flow. Healthy ecosystems are demonstrated by the presence of multiple Fillers consistently submitting different competing bids for the same orders.

The distribution of order sizes gives insights into how users place orders and the extent to which they trust the protocol. Low trust or protocol airdrop farming activity is often represented by a heavy skew in orders towards the lower value end of the scale, whereas a more organic distribution between order values both demonstrates increased trust and organic behaviour.

How fast a protocol can satisfy an order is seemingly the main differentiator between protocols seeking to win order. The consistency of these order execution times per protocol give some insight into the ability of protocol Fillers to consistently service orders whilst performing inventory rebalancing operations themselves. As such, whilst optimising for speed on it's own doesn't lead to better execution quality (in fact it in some cases ends up with the opposite result), the consistency of execution from the protocol Fillers is an indicator of protocol health. Note that there can be some margin of error in this measurement based on the timing resolution of the block time headers for each chain.

The consistency of these order execution times based on the order value gives some insight into the ability of protocol Fillers to consistently service orders whilst performing inventory rebalancing operations themselves. It's not uncommon to see execution speeds become significantly longer as the order sizes increase as fewer Fillers have inventory (and inventory deployed in the right places) to capitalise on the orderflow and bid to execute.

Order execution times across source and destination chains provides some insight into the ability of protocol Fillers to consistently service orders whilst performing inventory rebalancing operations themselves along with exposing chain preferences. Chain block times and rebalancing frictions often skew Fillers to prioritise some chains over others when filling orders and this can lean to localised minimums and maximums in order execution time.