The right mix of sources can provide more proof, but it's also important to understand when exposure the sources matters, when they'll have the most impact. A few years ago, a colleague and I looked at this exact problem. We notice adoption of a new website. They sent out e-mails, submitted their users, encouraging them to invite their friends to join the new site and so existing users could help spread the word through Facebook. The only way people could learn about the new site was through invitations sent from their friends. We analyzed the data over time to see how receiving invites from friends, as well as how many invites people received, impacted their likelihood of joining the site. Consistent with the value of corroborating evidence, people who got more invitations were more likely to join compared to someone who had only one invitation, for example, potential users who got a second invite were almost twice as likely to sign up. But beyond how many invitations people received, when they received those invitations also mattered. The closer different invitations were in time, the bigger their collective impact. To understand why it helps to go back to that co-worker recommending a show. They tell you how much they love it and another co-worker says something similar next day, it's hard not to at least consider checking out the show. Lots of people are talking about it, soon or further, show must be pretty good. But spread those conversations out a bit more and their effect is muted. One co-worker says something today, another one mentioned it three weeks from now, it's less likely to drive action. It's been a while since you heard about the show, so you're less likely prefer it's wildly popular and you heard about a lot of other shows in the meantime. If enough time passes, you may not even remember about the show in the first place. Our analysis of user growth found something very similar. Each invitation provided some evidence that the website was good or worth joining. Over time, though, it was like some of that proof disappeared or evaporated. Like water steaming off a hot rode, the more time that elapsed until the second invitation happened, the less proof that was left over from the first one. After one month, the invitation provided only 20 percent as much impact as it had initially. After two months, that first invitation had almost no impact at all almost as if people had never even received it. But concentration, putting them closer together, mitigated the decline. Just like hearing the same show from multiple people at once encourages action, we found that receiving multiple website invitations within a shorter period of time catalyzed change. If two people got them in short succession, they were more over 50 percent more likely to join the site. When trying to change minds, not all proof is equal. Concentrating that proof in time boosts its effectiveness. Trying to increase attention for a new service or important social cause make sure that social media hits happen soon after one another so potential supporters hear about it multiple times in a short period. Trying to change the boss's mind after stopping by her office, catalysts encourage colleagues to make similar suggestions right away. Hearing from multiple people in a short period of time increases the impact of those sources.