How I spent my summer working on TechFinder
A directory of nonprofit technology service providers.
Planning Site: http://tfprototype.aspirationtech.org
Old version: very awkward platform, difficult to search for useful data; had difficulty with vendors vendors and users updated info. It has to be able to be done better
"Search engine for nonprofits to use to find technology service providers"---The 'service provider' version of Social Source Commons.
The redesign process and conversation
- Started by looking at other search engines/services that *work*: Yelp, Kayak, etc.
- Then came up with a list of functionality/features (e.g. search/browse by needs, location, language, etc. ratings/reviewing,
- Did surveys of how people wanted to rate eevaluate nonprofits---what is preferred rating taxonomy/methods (1 click, scaled, multi-dimensional, etc.)
- Q: How do you review the reviewers? ("Was this review helpful?" - Q: How can readers better understand the perspective of the reviewer? - Q: As a service provider, how can it capture my referral networks and endorsements?
- Designing an optimal search box: what's the best entryway into the service? Dropdown of options is great (people don't know what they want). How to facilitate better recommendations e.g. people may not be searching for hardware recycling, but if they're searching for "hardware purchase", they could be presented with how to recycle old equipment. (do more than pander to their immediate need; provide broader context)
- Mapping professional affiliations (e.g. NTEN). Could help demonstrate legitimacy, but also be less-than-inclusive. How to link arbitrary affilitations/membership-orgs (with oversight?)
# Top Service Requests: make it smart so that the requests are related to your geographical/organization-type/etc. # Have a step-through wizard to ask the types of introductory questions that take place at the very beginning of client/provider matchmaking process # Can there be rich profiles for service-seekers? Can service-providers provide reviews of organizations (how to make it informative without locking people out). # How to "do no harm" when providing any sort of ranked list? How do structured reviews actually reflect a rich real-world matchmaking process? Are service providers rateable in the same as "good pad thai"? How to deal with the "Users will usually click on the first item in a list" when maybe the top 5 or 10 listings or commensurable? (randomize them for every search?). Also, "I'm a small NGO in the Bay Area, therefore reviews from other small NGOs in the Bay Area are most meaningful." # [every system turns into a social network, ed.] # "I want a vaguely positive/negative review from a very similar organization/circumstance than a very specific review from a vague actor". Focus more on the matchmaking process: ask people "what feels like a good match?" #Spam, ballot-stuffing, and moderation. # Ratings numbers with decimal points may provide a false authority. Can those numbers be softened and displayed as a shape or dimensional-field?