A New Approach to Social Service Data
At One Degree, we’re making community resources more easily discoverable by organizing agencies’ services into plain, action-oriented statements of opportunity, so people can find exactly what they need. Here I discuss the issues with the current state of social service discovery, our new “opportunity”-based approach, and the challenges that lie ahead. (Originally published in three posts: parts one, two, and three.)
Summary
The way people find social services today is antiquated and time-consuming. Online versions of resource directories provide little more guidance than a simple Google search. This organization-based approach focuses on giving people information about agencies, such as their website, phone number, general category, mission statement, funding sources, etc. But sifting through jargon-filled descriptions and irrelevant information requires a lot of time and effort on the part of the individual who is already struggling to make ends meet.
One Degree started as a pilot project that mimicked this directory-like structure, but we found that simply pointing people to organizations was not enough. Furthermore, there was no real way to measure whether or not we were providing a valuable service to people. Now we’re trying something new by going a step further by pulling out the “opportunities” that each organization has available. If you search for “child care”, rather than returning organizations, such as Mission Neighborhood Centers and Holy Family Day Home, and their general descriptions, we give you the actual opportunities that exist at these organizations, such as, “Get affordable childcare for children ages 4 months to 6 years.” We also give you the next steps so you are truly one degree away from finding the help you’re looking for.
Due to the range of services organizations offer, connecting people with opportunities is also inherently more measurable than the organization-based approach. The latter, which is the industry standard, gives people information and leaves it at that. By organizing data in atomic opportunity units, we have devised a system to track whether or not an individual finds that piece of information useful, thereby bringing measurability to an area that has had no clear metrics in the past. However, tracking information in and of itself is not useful. We want to measure whether or not people are actually able to improve their lives with the information we’re giving them. Our approach is to ask them directly. We know this is not a perfect system, but we think it is better than not measuring the outcomes of automated resource connection at all, which is the situation traditional online directories currently find themselves in.
Our hypothesis is that if we can provide better, more relevant information, then families will be able to more quickly and consistently improve their lives.
Part 1: The organization-based approach
In the United States, we are fortunate to have a huge network of organizations that are sometimes referred to as the “social safety net” — nonprofit and government agencies that serve families who need some kind of comprehensive or supplemental help in meeting their basic needs. Unfortunately, figuring out where to go for help when we need it is still much more complicated than it needs to be.
80% of us will face poverty or near-poverty at some point in our lives. That means making a low income, being unemployed, or generally needing assistance to overcome obstacles is something the vast majority of us deal with in our lifetimes. It happens to nearly all of us. There’s no shame in utilizing the network of organizations set up to help us. That’s why they exist.
Frustratingly, however, as life gets more difficult, so does the process of finding help. So far, efforts to make navigating social services have been focused on incremental improvements on the “resource binder” concept. Since the advent of the copy machine, agencies have kept literal binders of photocopied lists of organizations and their contact information. In the last couple of decades, these binders have been turned into online databases, but they have the same design flaw.
Let’s say you are a single mom, working two part-time minimum wage jobs, and facing some emotionally difficult circumstances. You’d like to try group therapy, but there’s no chance you’ll be able to afford the high-priced programs you’ve heard about. Finding a group that you can afford, for your demographic, and at a time that works with your schedule is not so easy. It requires a lot of digging and, most often, a helping hand from someone who knows the terrain.
Like everyone else, you start with Google, and get back a list of organizations and their websites. Then what? You’d need to sift through each organization, determine which seemed relevant — probably based on their name, at first — and try to visit their websites and call around to figure out exactly which organization offers what and when.
There are a number of organizations that have put the “resource binder” online. We think tools, such as United Way’s 211, are useful to certain audiences: case managers, social workers, and other practitioners already well-versed in the network of local organizations and what they offer. But for an actual individual looking for help, they provide little more guidance than a Google search.
Here at One Degree, we often say we’re creating a Yelp for social services because of the rating and review system we have. While that’s a good starting point for discussion, the analogy only goes so far.
Yelp results contain companies in discrete categories, whose purpose is clear: at a restaurant you eat food; at a dentist’s office you get your teeth cleaned; at a salon you get your hair cut. But now imagine if Yelp returned a list of company names, mission statements, and jargon-filled paragraphs of mostly-useless information. More importantly, each company in the list might serve food, cut hair, clean teeth, serve alcohol, or repair automobiles — and some might do all of those things. (This is in no way a criticism of Yelp; it does a great job at what it intends to do. I’m just using it as a point of reference.)
Social service organizations don’t fit into discrete categories because their services are so varied. Saying that an organization is a “housing organization” or a “food bank” tells you very little about what exactly it offers, whereas saying that a company is an “Italian restaurant” at least gives you the basics of what to expect.
In the case of a search for social services, returning organization names, jargon-filled descriptions, or basic categorization is really only helpful to other service providers or social workers who are already familiar with what each organization does, and who can interpret the jargon to find something useful.
We believe it’s time for a solution to this problem that is built for the customer, not the intermediary or service provider.
Part 2: An “opportunity”-based approach
In Part 1, I talked about what the current landscape looks like for finding information about social services. To recap: It’s a mess, focused on online directory tools that give superficial information about organizations.
One Degree started as such a website. For our pilot, we had a directory of organizations, where people could find basic information about each one. But after talking with users and taking a hard look at whether or not we were making a difference, we realized we needed to try something new. We had a useful database, but it probably wasn’t going to change people’s lives. And from a practical perspective, we had no way to even measure our impact.
So we evolved. Now, we’re going a step further: rather than just giving out information about organizations, we’re serving information about what an individual can do or get. We call these “opportunities,” as they are precisely that: action-oriented, rather than information-oriented. We think this is more aligned with how individuals look for help.
Too much information is a bad thing. Our approach is to only provide information that is relevant to finding help. Most resource directories and organization websites are filled with extraneous information, such as an organization’s mission statement, tax-exempt status, staff bios, and funding sources. If I’m looking for an after-school program for my kids, I don’t care about when an organization was founded or the history of its management team. On One Degree, we show you only what you need to know to get help.
For example, the St. Anthony Foundation in San Francisco has over a dozen fantastic programs for the community, including computer classes. It has one of the better nonprofit websites we’ve seen because it has so much information about its services. Yet the site is overwhelming, and it’s only one organization. An individual would need to browse through dozens of organizations’ websites — most of which are written for funders or supporters, not customers — before finding the help she needed.
If you’re looking to improve your computer skills, you could use One Degree to find computer classes. Rather than just giving you the name and basic description of the St. Anthony Foundation, we tell you exactly what you can do there — you can “take a class on basic email skills this Thursday morning”, or “take a workshop on Microsoft PowerPoint Saturday afternoon.” We also tell you how to sign up, where to go, and at what time.
We think this is a simple, yet fundamentally better way to connect people with community resources.
Measurability
Another advantage of this approach is that it’s inherently more measurable than the organization-based one. The latter, which is the industry standard, gives people information and leaves it at that. Agencies — or funders, for that matter — have not been able to track whether or not someone looking for information was able to find it.
By organizing data in atomic opportunity units, we have devised a system to track whether or not an individual found that piece of information useful, thereby bringing measurability to an area that has had no clear metrics in the past.
With One Degree, we are highly focused on using data to drive what we do. We want to see if this approach works and if it doesn’t, then try something else. Baked into our application is a mechanism to securely track what people are searching for and what they add to their list that interests them, while respecting their privacy and personal information.
However, tracking information in and of itself is not useful. We want to measure whether or not people are actually able to improve their lives with the information we’re giving them. Our approach is to ask them directly. We follow up with users, asking them to indicate whether or not they took an opportunity and if it fulfilled the need they had.
We know this is not a perfect system and recognize some of its inherent flaws: some will be reluctant to give us detailed information about their behavior, they may not be candid with their responses, and others may not know whether an opportunity has actually been immediately useful. But we think this system is better than not measuring the outcomes of automated resource connection at all, which is the situation traditional online directories currently find themselves in.
Our hypothesis is that if we can provide better, more relevant information, then families will be able to more quickly and consistently improve their lives. We’re still at the beginning of this experiment, but we are excited about our opportunity-based approach to data, in particular because we will be able to measure its results.
Part 3: The data collection challenge
In Part 1, I detailed the way families look for social services today, and in Part 2 I introduced One Degree’s opportunity-based model, which we think brings more useful and measurable information to the table. Although we think this is a better approach, there are a number of challenges to making it work.
Serving data that goes a step further than traditional directories means having information in a specific format whose collection is not easily automated. A computer script, scouring organizations’ websites for their most basic information is not sufficient. Furthermore, we do not yet have the artificial intelligence capable of perusing organizations’ websites, talking to staff on the phone, and then automatically writing content in the format of opportunities described here.
Indeed, our data model requires a much more human-driven collection strategy. For now, we are relying on volunteers to help us sift through local San Francisco nonprofit websites and write out opportunities. This is a time-consuming, painstaking process that requires editorial judgment and a decent understanding of social sector code words so they can be translated into a user-friendly style.
(While many organizations do incredible work, the information they provide is clearly not directed at their clients. For instance, these direct quotes from the websites of two effective and well-respected agencies in San Francisco paint a confusing picture: “a comprehensive, integrated framework of evidence-based services” and “reducing harm is possible through the provision of accessible, non-judgmental drop-in and outreach sessions.”)
We recognize that our current efforts to aggregate this opportunity information is not scalable with volunteers, and the information itself will quickly become outdated and inaccurate. In the long-term we will likely need to build some semi-automated processes, which would require human editors, and also a staff that can work within communities we serve to collect and constantly update information, structured as opportunities, on an ongoing basis.
Moreover, we need to demonstrate that our model is an effective one, and eventually enlist the help of agencies to use opportunities as the basis for sharing their information, either by proactively adding their information to One Degree by hand, or by allowing One Degree and others to consume their opportunity information through a computer-friendly medium. This will require advocating for policy changes in government and nonprofit agencies to supply data in this improved format. To facilitate the compilation of this data from third parties, we are maintaining the opportunity information we collect in an open database, accessible via an API, that any individual or organization can use or add to.
Just getting started
At One Degree, we’re taking a new approach to connecting people with help to improve their lives. There are so many great resources out there for people who want it, and they should be easier to find. By giving people information about the opportunities at organizations, rather than simply the organizations themselves, they will be one degree away(!) from finding what they need.
Our vision is a world where everyone can find the help they need quickly and easily to lift themselves out of poverty. This is a lofty goal, but we believe that, in our modest way, we can help move our society toward that end. We have a lot of work to do, but at the outset of this experiment, I wanted to share with you this new approach to helping families improve their lives, and why we’re really excited about it!
If you have any feedback, questions, or ideas, please email me at eric@1deg.org.
Our CEO, Rey Faustino, signing up a new user for One Degree (smart phone is on the clipboard).
