Sometimes you’ll be collaborating around existing data, perhaps from public records. Other times you’ll be gathering data or tips from readers or potential sources using an online form builder like Google Forms. How do you know when a project is a good candidate for the latter?
When you're considering an idea for a collaborative crowd-powered data project, there are some questions you will want to ask. See the end of this report for a worksheet with a list of these questions.
It's important to consider what information you're hoping to get. It's not enough to define a topic you want to report on; you need to spell out exactly what you want to find out. For example, it's not useful to just say you want to report on an election. Some questions you can ask are: What kind of problems do voters face? Which groups of voters face the biggest hurdles to voting? What cities or states are experiencing serious voting problems?
The questions will ultimately shape the form you'll build to survey your audience.
Keep in mind that crowdsourcing can be your main avenue of sourcing, but it can also be one of several.
You'll want to evaluate how you can gather the most complete dataset or source pool possible. Would this be better accomplished through public records requests, existing datasets or gathering news clips? Or would those be useful in addition to crowdsourcing?
Crowdsourcing can be a tool to help find people affected by the issue you’re covering. In turn, those new sources can help deepen the story in order to tell better, more relevant stories.
When we say “audience” in this instance, we mean the people who will have the highest incentive to fill out your crowdsourcing form. Those are the people you want to reach.
This is essentially an exercise in sourcing. Is your audience defined by geography? Shared experiences? Profession? Ethnicity? Anger at the same institution, workplace or system? It may be more than one group; for example, you may want to reach eligible voters in Florida or undocumented immigrants in detention. You must define this audience early on, after you define the questions you want to answer.
In other words, do you aim to build a complete dataset or are you more concerned with unearthing new story leads? If you need a complete dataset, you will almost certainly need additional branches of reporting and to incorporate other sources, like public records.
If you’re hoping to say something about the overall population based on the responses you get from your online form, you’ll need to launch a more scientific survey using random sample techniques. Without that, you’ll be limited in how you can characterize the overall response to your form.
But if you're hoping for leads — on sources, potential patterns to follow-up on more scientifically, etc. — engagement journalism and crowdsourcing can help you do that.
The next thing to consider is if the information you need is simple or condensed enough to fit into a single set of questions. When it comes to building a survey form later on, you'll want to keep things simple. Circulating more than one form will reduce response rates, especially if you're working within a collaboration.
It's almost always a good idea to translate your tip form and distribute it in other languages. (That said, you shouldn't translate the form into a language if you don't have any participating journalists who can read the responses and communicate with tipsters.) If you’re doing reporting across multiple countries or trying to reach immigrant communities, it’s mandatory. You'll also need to find partners that can help you distribute the translated form to the appropriate community that speaks that language.
There are a number of factors you'll want to consider when thinking about partnerships around crowdsourced data:
Which organizations cover and serve the communities you want to reach?
Which organizations have expertise in reporting on this subject or community?
Which journalists would be interested in and willing to do this kind of work?
You will want to partner with news organizations that are able to distribute the tip form to the audience you want to reach. That could be geographic or relate to specific communities. You'll also want to consider who would be most interested and best prepared to report these tips out.
That might mean local outlets covering local news or journalists with expertise in that particular beat. But you also need to consider which news outlets are willing to collaborate and able to be good partners, which can be a challenge for larger companies given the levels of hierarchy decisions require.
You’ll also want to talk about who’s reaching out to participants and how. Pick partners you trust to be considerate and careful.
When considering partners, you should consider the most likely product outcomes from your data. Will the stories likely be text, audio or video? Will they include visualizations? If the topic is particularly sensitive and tipsters will be hesitant to use their names or show their faces, video and broadcast may be a more challenging route.
You should establish a system for following up on tips and make sure that partners are on the same page. (We have suggestions in this guide about how we've done it so that you can divide and conquer.) You should also think about how you plan to engage people sending in tips. Are there other ways you can keep them engaged, like keeping in touch on a regular basis, sending updates when you publish a new story or starting a private newsletter or Facebook group?