This is the document that Cindy and Christine provided to challenge grant officers, including me (Nicole) in 2015-16 when freelance officers were brought on to launch challenges. This is the basically the schedule that I used to get Zoos launched in late 2016.

This document needs updating as of 1/27/21

We can feasibly get grants launched in 4 weeks now, but I can’t emphasize this enough: get started early! It’s always best to afford more time because researchers’ responses to emails are both unpredictable and much slower than what we’re used to (sometimes people email us two weeks later asking if they can “start the process”).

Week 1-2 is scraping, contacting, and fielding calls with interested people. Start of week 3 is first draft submissions, end of week 3 is final draft. Spend week 4 doing campaign prep. Launch at end of week 4. What we’ve been seeing is that these dates get pushed back a few days each, even a week or more, due to researchers needing more time to submit, and our team also needing more time to manage the projects.

What will make this faster is if people come in already knowing about Experiment or interested in trying a new way of fundraising. I’d say a lot of my time at the start of the four weeks goes into explaining and convincing people to try this. Many people are initially interested because of the word “grant” and sometimes this puts the wrong idea into their heads. Emphasize that this is crowdfunding, an exercise in sharing excitement about science, and communicating what they do, rather than submitting a proposal and hearing back about being funded or not.

Grants Program Overview

A. Getting started: identify a field of study for your grants topic

It should be narrow enough, such as the focus on a conference, to limit the scope of the people you’ll contact. But the group shouldn’t be too small, because you can expect that at minimum, only about 10% of people will reply to your email and 3% will end up committing/submitting. So if you want to get 20 people to commit (assuming around 5 drop out or can’t end up submitting in time for various reasons), you should try to collect around 500-600 emails. Many of these won’t end up being valid, so always aim higher. It’s better to have too many great projects to choose from than struggling to recruit projects or researchers that aren’t the best fit for this model.

Also think about what topic makes sense, budget-wise. We mentioned $5000 budget in the email since this is what we’re comfortable funding (“average successful project on the site raises $4-5k”), so if it’s a field that generally uses more expensive procedures and equipment, it’ll be hard to find enough people to commit in time for smaller budgets. Insects turned out to be great for this. Many of them were excited at even the idea of $2-3k, and all of the projects submitted turned out to be under $5k. We did a lot of ecology/charismatic animal challenges, which work well enough, but don’t limit your ideas to just that area!

You also don’t want to saturate the entire conference abstracts list. People forward the email and share it, and they might get suspicious if it seems like everyone got the same email (though if they ask, I say something like, “We read through the abstracts from XYZ conference and hand-selected ones we thought were interesting” which is true most of the time). Also, in an attendee list, many of them are international or work for government-funded agencies such as NOAA or the USDA. I filter these out before starting, usually by first filling going down the list and filling in states/gov based on institution name. Our success rate there will be much lower if they’re on government funding, and it’s not worth the time to scrape for those emails.

B. Scraping for names and emails

Use import.io to scrape websites for names of attendees. Try to get the primary author on an abstract, as well as institution, title, and abstract name.

If the abstracts list is similar to the marine mammals and entomology conference apps where the abstract is hidden inside a name card, you’ll need to do a nested scrape — first scrape all of the abstract presenter names and their links, then input the list of links to a second scraper that collects the text data information on the abstracts.

Copy and paste it all into a Google doc and remove any abstracts that don’t look appropriate (General informational stuff, educational initiatives, international, sometimes titles are missing, etc.).

Here’s the one I did for entomology abstracts (separate out fields we found emails for to the second sheet): https://docs.google.com/spreadsheets/d/1iLOlO2wsm5U7NnxUsZaV4aohO4BufQf6Ie1ynvzLlNk/edit#gid=0

Jeff the ODesker is pretty fast at collecting emails. He can do average around 15-50 emails an hour, and you can ask him to either try to get as many as possible, or just focus on getting all of them. If you have a long list that is much more than 600, I tell him to get one email from each abstract (avoids emailing everyone on that talk and having them all check with each other).

Make sure to double check his emails because if we send too many bounced emails, it apparently hurts our email domain quality (?). You can also run it through Kickbox.io ($5 for 500 email checks. I have an account currently linked to my card), but you can save a couple just by having a quick run through. Sometimes copy and paste doesn’t capture the “u” at the end of “.edu” or switches “mail.cornell.edu” to “cornell.mail.edu” or accounts for having to change “___ (at) ____ (dot) edu” into a usable address. It’s kind of amazing how many people mess up their own email addresses!

Kickbox usually advises you to only end up sending like a little more than half of emails, depending on the list. I fed a list of 977 emails to them and they said around 600 were “not risky”. Still not sure whether it’s worth it to skip out on the emails it calls “risky”, but when I skipped these for the insects batch, we were guaranteed that none of them would bounce, and they were right.

Before sending out any emails, make sure you’ve finalized a timeline for deadlines (first draft, final draft, launch date, when the grant is awarded, when projects will end -- you can see examples on the grants info docs for how we spaced it out to give ourselves enough time to do calls, review project proposals, and send pre-campaign prep emails: marine mammals, insects). More about thinking about timelines in section E.