1 of 16

Advocacy Through Prototypes

Princeton JuST

NICK PUNT

2021/01/20

2 of 16

About me

I’m a product creator interested in elevating the human condition.

I’ve spent several years at the intersection of tech and social good, in education and mental health.

I’ve also worked in a variety of media (news, social media, video games) and am interested in how mediums shape our thoughts & actions.

  • Built teletherapy products at Mindstrong
  • VP product/design at OneSignal, push notification delivery service
  • Co-founded EdSurge, the leading edtech news source
  • VP product at Uversity, building social networks for incoming college students

nickpunt.com - nickpunt@gmail.com - @nickpunt

EDUCATION

Stanford MBA, MA Education

UC Irvine BA Psychology

3 of 16

Overview

What to expect and what to focus on when doing advocacy through prototypes

4 of 16

Your goal is to influence & advance the conversation around a particular problem.

It’s not to launch a real product. (though that may be nice)

It’s not to find the perfect solution. (though that helps)

It’s not to win a debate. (though that may happen)

5 of 16

Your audience is anyone with the influence to bring products to life.

They’re PMs, Engineers, and Designers at companies in the space.

They know how decisions get made and things get built.

You want to show what is possible, and why it’s worth pursuing.

6 of 16

The more compelling the story you tell, the more likely it influences your audience.

The problem should resonate.

Your insight and solution must be plausible. (the more the better)

Your analysis must overcome their doubts. (they have many)

7 of 16

Your audience should walk away with an awareness of broader possibilities.

You want to push the boundaries of their thinking.

Ideally you plant a seed in their minds that they keep returning to.

You want to get them to interpret the problem themselves.

(it helps if you can get them to think the idea is theirs)

8 of 16

Twitter Mea Culpa

Walking through an example of advocacy through prototypes

(moving over to the article)

9 of 16

Where to Start

Some tips on approaching problems

10 of 16

Consider Underappreciated Problem Areas

Within any big problem there are tons of small facets that people aren’t very aware of. Your goal is to raise awareness, so try picking one of those.

Consider things that:

  • Happen infrequently
  • Only affect certain users
  • Are below the surface like algorithms and psychology

SUBJECTS

Health & wellness

Politics

Current events

Etc

EFFECTS

Poor/dangerous decisions

Loss of resources

Incorrect beliefs

Loss of connection to reality

Rewards extreme content

PSYCHOLOGICAL DRIVERS

Addiction to news & gossip

Inattentiveness

Overconfidence

Lack of understanding

Automaticity of reading

Ex: Misinformation

HOW IT MANIFESTS

Clickbait headlines�Ease of sharing

Virality before verification

Lost nuances

Misleading images

Bad sources

Lack of corrections

Trending topics

Matters of opinion vs fact

Echo chambers

Desire for validation

Confirmation of beliefs

Hustle & grift

Desire to help

11 of 16

Chase Hunches

Read & learn until you develop a hunch.

Chase the hunch through design until you reach a stopping point.

Then return to reading about the problem. Repeat.

Chasing a hunch is about relying on your own problem solving skills to motivate you.

It’s often helpful to incorporate constraints progressively like this vs. all at once.

You want to avoid analysis paralysis and getting disheartened by complexity.

OBSERVATION

‘Why can’t people admit mistakes and avoid all this nonsense?’

HUNCH

‘What if we could admit mistakes?’ → Mea Culpa flag with follow-ups

RESEARCH & CONSIDERATION

Cultural and psychological difficulties in apologies and mistakes, history of poorly written copy in social media dialogs & tags, user emotional state

UPDATE

Remove follow-ups design, tweak copy “@user has indicated they made a mistake in this tweet”

ETC...

Ex: Mea Culpa

12 of 16

Start with the obvious

Early on try to just solve a problem in an obvious or literal way.

Use these solutions as dummies to beat up and guide deeper questions.

Your goal is to have answers for all relevant second- and third-order effects.

In the process you may need to update your obvious solution.

On the other hand, you may also find obvious solutions are easier for users to understand, despite their imperfections.

PROBLEM

Users share a story from a sketchy source.

OBVIOUS SOLUTION

Include ‘this is a sketchy source’ tag.

QUESTION

How is this tag added?

OBVIOUS SOLUTION

Other users flag it.

QUESTION

How is sketchiness determined?

LESS OBVIOUS SOLUTION

The sum of user tags on a domain creates a ‘sketch score’. At some threshold a tag is added.

Ex: Sketchy source

13 of 16

Start with the obvious

Early on try to just solve a problem in an obvious or literal way.

Use these solutions as dummies to beat up and guide deeper questions.

Your goal is to have answers for all relevant second- and third-order effects.

In the process you may need to update your obvious solution.

On the other hand, you may also find obvious solutions are easier for users to understand, despite their imperfections.

QUESTION

What if users brigade a particular domain to try to take it down?

CORRECTIVE SOLUTIONS

  • Include human intervention prior to assigning sketch tag to make sure tag is justified.
  • Weight taggers by history of accurate tags.

FOLLOW-ON QUESTION

Who is intervening? How do they decide?

POSSIBLE SOLUTIONS

  • Write a set of mod guidelines to clarify ‘sketch’, don’t flag if in doubt (reduce false positives).
  • Outsource to third-party group.
  • Forget user tagging solution, just have mods look for popular domains that are sketchy.

Ex: Sketchy source II

14 of 16

Clear Next Step

One way to influence is to show a clear next step, something that could be implemented tomorrow.

HOW TO

Make your case in the form of an experiment that can be run, ideally as a product requirements doc.

Be clear about metrics, what success looks like, and abuse potential.

Your case should be relatively airtight.

Another way to influence is to present a pot of gold that highlights a destination worth pursuing.

HOW TO

Make your case in the form of “we get X if we can just overcome Y.”

Pick a single problem to leap over, and be clear what overcoming it means.

Show how getting to the destination creates new possibilities to resolve some long-standing issues, and include a few solutions to demonstrate that.

Pot of Gold

vs.

15 of 16

Don’t limit your story to just UI

UI is just the tip of the iceberg of a product. Most ideas require changes under the hood, so consider how you might change APIs, algorithms, and business processes.

If the plausibility of your idea relies on assumptions of how users will use it and important second-order effects (e.g. it’s not just a straightforward UI), you’ll probably want to model those. Consider prototyping interactive widgets with sliders to adjust the key variables. Remember, your goal is show something plausible.

16 of 16

Q&A

NICK PUNT