Picking Up from Last Week
Last week, Kaley wrote about our new direction: a platform for Austinites to voice their feelings on civic topics, which then empowers their voices through a collective display of the city’s voices. The feedback loop for residents who engage was still a lingering piece, and our current vision for this is to connect users with opportunities to take action and seek resources.
This week, we collected a larger data set to start playing with and worked through the details of our platform to bring the idea into higher fidelity.
Call for Voices
Essential questions we needed answered this week:
What kinds of things will people say when asked what in Austin they care about?
Will people’s responses start to fit together when we collect a large enough sample size?
Will people be motivated to contribute their voices in the first place?
To address these questions, we created both an online survey, a call/text option, and an in-person activity. We posted the survey on our respective Nextdoor and Facebook accounts, where we received strong traction. For the call/text option, we designed and posted flyers in libraries, parks, and coffee shops across Austin. Finally, we designed an in-person questionnaire that we worked through with 3 strangers. In total, we collected 100 responses, 96 from the survey, 3 from the in-person interactions, and 1 from the posters.
- On social media, where people know us, people were more inclined to participate. Strangers were more hesitant – who wanted their thoughts and what were they using them for? Valid questions. We need to be more explicit with our intentions to help establish trust.
- One action item for building trust and reaching engaged Austinites will be to go through already-formed groups, such as PTAs, neighborhood organizations, and Facebook community & interest groups. Getting one advocate within each group will enable a transference of trust and increase participation with The Pulse.
- The poster’s call to action was not easy to find or very clear. Simplify the graphics and the message.
- The posters failed – we got one response from a library manager for multiple hours of our time. Social media had a good return, but this is only within our network. A big challenge will be reaching a more representative Austin. While this is our ultimate goal, we have decided that we will first focus on people who are already active in local issues. Have them spearhead the platform and bring others on board.
Digging into the Data
To gain insight into how our platform would work, we simulated the categorization of voices by printing and grouping all the collected responses. We read each one out loud, then proceeded to place it near others that addressed the same topics.
What the Data Showed
- Transportation and Housing are the strong top categories where the people we reached have concerns to voice
- People generally selected the most relevant category for their thought. However, some people covered two separate issues in one response.
- Most of the Other categories fit nicely into our larger category groups – some people just like that specificity.
- Some specific subgroups emerged, but many responses remained at a higher level.
- There’s a wide range of specificity with open-ended responses. Ex. “affordable housing” vs. “I was lucky enough to buy an affordable home that has become unaffordable because of taxes. I am approaching retirement, and can not see how I can remain in my home with the reduced income of retirement. I imagine it is even worse for people who are renters.”
- Some people’s voices may fit within others’ as we categorize them.
- There’s a certain level of conjecture that we have to make at this early stage in deciding where to place opinions.
- Some people were thinking more on a national scale. We corrected this by emphasizing Austin in the messaging.
- Some people’s responses don’t necessarily take a stance – we must prime them to do so.
- Rather than “what about that topic is on your mind?” We are considering phrasing the prompt as “what change do you hope to see?” – Making it more actionable and making them take a stance.
- Some people’s responses felt vague because they had already indicated their category prior. We decided to remove the category because:
- One less click for them.
- They will be more specific in their free text and we will be able to tag it more successfully.
- 19% of respondents selected their own category rather than selecting a preset.
We carried our takeaways from this experiment into creating a service blueprint. The blueprint draws the user’s path through interacting with Pulse of Austin, as well as maps out the other factors at play during their experience.
The activity of creating the service blueprint highlighted steps and connectors between aspect of the service which we had not yet considered. It also clarified our conception of the Pulse’s system requirements, helping us to see what we are getting ourselves into on the technical side. This blueprint also inspired new features and pushed us to make decisions on elements of the user’s experience.
Decisions and Revelations
- We need to decide from what sources to pull updates, events, and resources.
- We can let users contribute to the listed resources to help prevent things from falling through the cracks, but a lot of this will need to be moderated for legitimacy and usefulness.
- We have moments where moderation would be needed: confirming that a submission should be in its assigned group and reviewing both thought and resource submissions for appropriateness and accuracy.
- We can have our power users (early adopters, advocates, activists) lead charge on the page(s) connected to their focus area.
- Categorizing someone’s voice: We planned on having the user select the category or group they think their concern belongs to. How many options do we give them and what steps do we take them through to ensure the proper match?
- Instead of putting the categorization burden on the user, our system will make the decision itself. This also removes the friction of the Yes-No-Try Again process that would push some users to exit. This of course, relies on a “smart” system to tag words appropriately.
- We have two main types of visitors to the site: those who have a specific concern to voice, and those who want to explore and maybe come across an interest.
- We want to make sure people can visit the site and take action right away, whether by contributing their voice or by viewing what others have said. To reduce friction, login intervention needs to happen at different, distinct moments in the journey for our two main types of users. It’s really only when they are taking an action that needs to be linked to a human (themselves), that they should be prompted to log in or create an account.
One of our riskiest assumptions is that any software or data algorithm we develop will be able to effectively tag and categorize people’s voices in a way that they feel properly represented. Not being experts in this field, Kaley and I plan to spend this next week researching just what it will take to create this capability and where we may have limitations.
Additionally, we haven’t had much opportunity to run a simulation of this platform by users. We need to measure responses to the part that comes after submitting their voice. Will they be interested in reading what other people say? Will they feel satisfied if given programs or shown ways to participate? This is particularly critical at this beginning stage, because, as of now, the Pulse of Austin has not matured to the level of making an impact. How do we encourage Austinites to engage with a platform that only has the promise, and not yet the proof, of making an impact?