All these numbers are the annual average percentages from our projections of the next five years.
Instant Usage Percentage
The percent of Usage that is available instantly -- available as Open Access, Backfile, ASNs, or Subscription. Or to look at it the other way, all Usage that isn't via ILL or Other (delayed)
The percent of Usage that is available from Academic Social Networks like ResearchGate and Academia.edu. These are not considered open repositories, so papers hosted here are not considered Open Access. These can be excluded in the settings.
The percent of Usage that is available as Green, Hybrid, or Bronze Open Access. Some of this content can be excluded in the settings (see the Open Access page for more info).
The percent of Usage that is available from your backfile -- the content that you still have access to because it was published while you had a subscription. If your institution doesn't have Perpetual access you can exclude Backfile in the settings. In a Paid Account you can also upload a customized list of perpetual access availability by journal.
Percent of Usage that is fulfilled via subscription in this scenario. Of course if you've subscribed to a journal you can actually fulfill all of it via subscription, but we model that you fulfill via free, instant sources when possible, and only subscription for the remainder of the paywalled content (costs and instant percentages aren't affected by this decision, it mostly just shows up in the fulfillment bar graphs showing fulfillment breakdowns).
Percent of Usage that is fulfilled via ILL in this scenario. This all Usage that isn't available via a free, instant route multiplied by the ILL request rate (adjustable in the settings, though the default is based on an extensive literature search of other institution's experiences).
Percent of Usage that is fulfilled via Other (delayed) paths in this scenario. This all Usage that isn't available via a free, instant route and the researcher does not choose to place an ILL request, but instead fulfills their information need another way -- asks the author for a paper, asks a colleague, finds another similar paper that is good enough for their purposes, etc.