Depth of participation tracks the channel. Voters mostly met the process at a ballot box in a school, community centre or T stop and were often hearing of it for the first time; in 2014 over 70 percent had done nothing else, and 38.7 percent of votes were cast at six Boston Public high schools. Change Agents, by contrast, gave weekly after-school hours and often long journeys, which Grillos (2014) and Augsberger et al. (2016) both identify as the main cost and the reason committees skewed toward already-engaged youth. Digital channels grew from a marginal 2014 website (nine eligible ideas) to text, social media and digital ballots by 2016, which participants credited with wider reach but found impersonal. Door-to-door canvassing was recommended in 2014 and is not documented as having happened. Translation and interpretation were offered as often as feasible, and youth themselves prioritised Haitian Creole, Spanish and Vietnamese for vote materials in 2016.