How Product Managers Can Mess Around With Open Datasets

Most Cities, States and Federal Agencies are working on some type of Open Data initiatives. The most common is an “Open Data Portal” that makes it easy to grab and use datasets:

https://data.cincinnati-oh.gov/
https://data.colorado.gov/
https://data.commerce.gov/
https://www.data.gov/

Some cities are using Open Data to publish performance metrics like the Seattle Police Department or Louisville’s LouieStat.

Civic Leaders working on these initiatives cite promoting transparency in Government, improving performance and providing data for innovation as reasons why Open Data is so important.

As a Product Manager, it’s helpful to be familiar with what’s out there and how you can play around with these datasets to better understand how your product may benefit.

Before you dive into querying APIs, checkout a few of these projects to see the end result of building something with Open Data.

USAFacts
CollegeScorecard
500 Cities Project
Data.gov

Ok, now let’s dig into some datasets you can play with.

Socrata’s Open Data Network
Socrata hosts over one hundred different data catalogs for governments, non-profits, and NGOs around the world. Checkout their Open Data Network where you can search for datasets.

For example, here’s a page about San Bernardino County Employment. Click “View API” to end up on a page giving you data and an API call you can paste into your browser or Postman.

Namara
Namara has organized a bunch of public datasets into a beautiful UI. Create a free account, sign in, create a new project then click Open Data in the left column to search and add datasets to your project. You can view the table data and manipulate it or call the data using their API.

https://api.namara.io/v0/data_sets/{DATA_SET_ID}/data/{VERSION_ID}?api_key={YOUR_API_KEY}

In your project settings, you can generate an API key. Then, in each dataset you can click “API Info” and get the data_set_id and version_id.

ProPublica
You can use ProPublica to request data about Congress such as a list of Recent Bills and Member Voting records.

https://propublica.github.io/congress-api-docs/#congress-api-documentation

You’ll need to request an API key by emailing apihelp@propublica.org then pass that in the X-API-KEY header.

For example, to query Rep. Jared Polis’s voting record:
https://api.propublica.org/congress/v1/members/P000598/votes.json

Open Data and the big IaaS Platforms

Another approach is to checkout Public Datasets baked into AWS, Google Cloud Platform and IBM Bluemix.

This is a great example of using Google BigQuery on NYC Public Datasets.

AWS hosts a bunch of Open Data in S3 buckets.

IBM, as part of the NOAA Big Data Project, has built an easy way to download tons of data.

Additional Reading

A few hashags to search around on are #govtech, #opendata, #opengovdataand #opengov. Follow people like @Josh_A_New, @JoshData, @DataInnovation, and the @SunFoundation.

Here are a few links related to Open Data policy and relevant news.

Some history on U.S. Federal Open Data Policy

DATA Act passed in 2014, America’s first open data law. It directs the federal government to transform all spending information into open data.

Conversation on the future of Open Data as Administrations change and the Preserving Government Data Act of 2017

The OPEN Government Data Act “directs all federal agencies to publish their information as machine-readable data, using searchable, open formats and requires every agency to maintain a centralized Enterprise Data Inventory that lists all data sets, and also mandates a centralized inventory for the whole government (data.gov)”.

Open Data 500 US is an interesting survey results showing what kinds of companies use which agencies’ data.

What I Want To See From Evernote in 2017

I have almost all of my strategic thinking, articles I’ve found useful and reference material in Evernote. Their browser extension works great as does their Mac, iPad and iPhone apps. I even have the WSJ integration enabled so I see relevant news with my Notes.

It’s time for Evernote to not only store my information, but really help me be smarter and better at everything I do.

I would happily opt-in to this feature giving them access to learn from my personal data as long as I had a mechanism to “mark Notes private” which would exclude them from Evernote’s Machine Learning activity.

Given a seed list of Notes or a Notebook, I want Evernote to help me:

  • Monitor important news and activity from Companies and People I’m interested in
  • Show me correlations and visualizations in my Note data so I can better connect the dots and broaden my context
  • Suggest actions I should take based on my Note data

Evernote knows the Companies, People and Topics I’m interested in. Their browser extension could contrast my browsing behavior and work style with what I save into Evernote to learn more about me. They know my travel habits based on where I save Notes and all of the travel data I store in Evernote. They know about my kid’s activities because of the receipts I save, they know the gift idea list I’m keep for my upcoming 15th wedding anniversary.

Example:
I want to build a list of Venture firms funding healthcare companies. I know that Mattermark and CB Insights have these by segment, but I want my own list and to apply my own logic. I want to understand all the people that work there and what they are talking about. I want to know about their investments. I want to know when people leave the companies. I want to dig deeper and see trends using visualizations, etc.

I have been to the Evernote conferences, think Phil Libin was a visionary leader (#selfie) there and continue to be a paid user of the product. I’m hoping 2017 is the year that the Evernote team blows my doors off.

The First 90 Days for a Product Manager New To Healthcare

I talk to Product Managers all the time that are considering building products or features that would move them into the Healthcare space. This post is for them.

As you venture into Healthcare, you are embarking on the most TLA filled space imaginable. Talk to other Product Managers that are veterans in the space and you’ll barely be able to follow along. They are not trying to sound smart, it’s simply that the TLAs get ingrained so deep that it takes real effort to not use them.

You’ll also come across a few big topics like:

  • How regulation impacts business models
  • Selling into large Healthcare organizations
  • HIPAA and GxP
  • Healthcare and Life Sciences are very different

I began my first 90 days as a PM in Healthcare on a beach in Florida with my wife and kids on Thanksgiving break. I lounged around reading/listening to these 3 books:

Doing this gave me an understanding of the ACA, Pharma and all of the problems especially in the US Healthcare system. I was beginning to understand some of the TLAs.

My next step was to talk with other Product Managers. I had to ask a lot of dumb questions in these conversations but everyone was super nice and helpful. #givefirst These conversations helped me build a list of news, people, podcasts, blogs, etc that now make up my daily healthcare and biotech information pipeline. On a daily basis I found myself reviewing the Healthcare Top 100 and thinking about the business models of these companies. I subscribed to daily healthcare industry news from Becker’s Hospital Review, STAT and Modern Healthcare. I followed Zdogg and David Chase on LinkedIn. I listened to podcasts from Catalyze.io and a16z.

At this point in my first 90 days I was back in the office, still under a nice “you are still new” grace period and learning from everyone I could. The next step was to talk to Techstars Alumni, Denver-based Founders and friends of mine that are running Healthcare Startups. I had 50 conversations in two weeks and summarized most of them in a shared Box folder the team uses. For a handful of the conversations, I summarized and sent an email to entire team to raise awareness. This quickly established me as the expert for our “Healthcare Startup Developer” persona.

In the first 90 days:

  • Tell stories to build empathy about the people the product is trying to help.
  • Establish yourself as the voice of the customer on your team.
  • Show you are passionate and curious about the problems to be solved and the people you are trying to help.

It’s been one year since I began my journey as a Product Manager in Healthcare. I am once again sitting on the beach over Thanksgiving break pondering our Product Roadmap, Industry trends and reading everything I can about AI in Healthcare, Big Data Analytics use cases, Serverless Compute, how the ACA may change in the next Administration, CRISPR and more.

If you want help with your first 90 days, please don’t hesitate to reach out….and good luck!

Photo Credit: https://unsplash.com/@joaosilas

Thinking about Enterprise Software Startups

I ended up down a rabbit hole of research on Sapphire Ventures thanks to the Origins (Notation Capital) podcast on my flight home from Boston early this morning. Sapphire invests in Enterprise software companies. I got to thinking, if I were to analyze companies in that space with my Product Manager hat on, what would I look for?

These 5 areas came to mind.

1. How will the Product interoperate?

Ex: Zapier, BI fabric, Hybrid Cloud

Is the team thinking about how to move pieces of data around from their app to other apps, from their app to the Enterprise systems, between their Public Cloud and the Enterprise’s on-prem and Dedicated Clouds?

How will insights and raw data from their product be accessible to the Enterprise’s BI Fabric, Data Scientists, etc? How does that strengthen the value of the offering?

2. Where are Users interacting with the Product?

Ex: Mobile, APIs, Slackbots, Echo

Is their product enabling all types of Users to be engaged anywhere? Is extension of the product easy by a Customer Developer via APIs? Is there potential for an Ecosystem to organically grow around the product? Does it feel like a Platform? Is the pretty Mobile app for the on-the-go Sales person just as well thought out as the Developer API?

3. What are the Combinatorial Effects?

Ex: Exongenous Datasets, Data Network Effects

Is the team thinking about combining datasets together to create something new? Does the product have inherent data network efforts? As more people use this, will the value increase?

What two features used together accomlishing something really powerful?

4. What role does analytics play?

Ex: NLP, Computer Vision, Salesforce Einstein

Enterprise data is flowing through the product. How is that data being mined into features? How are signals being extracted using NLP, Computer Vision, Machine Learning, etc? Does the business analysis get smarter the more people use it? Does AI feel like a foundational part of their approach or do they think of it as gimmicky and a nice-to-have?

5. Talk about the Tech Stack

Ex: Microservices, Serverless

Is the team using technologies like AWS Lambda? Do they talk about Reference Architectures and Blueprints? Are they taking a Microservices-first approach?

A few more articles I came across while writing this post:

It’s fun to think this stuff through. I remember the days of meeting with investor after investor while in Techstars and how many of those conversations led to strategy and product improvements.

There’s a lot going on in Healthcare Tech

In 2015, venture funding of digital health companies surpassed $4.3B and accounted for 7% of total VC funding in the US. Deal sizes are growing and the percentage of later stage deals is increasing signaling a maturing in healthcare investments.

In my first six months as part of the Watson Health team, I’ve observed a few trends such as:

Google’s Investment in India will impact Healthcare
Google announced they will train 2M Indians on Android OS and promote internet use among rural women by 2019. In India, 5% of the population has health insurance (“cash for care”) and over 70% live in rural areas without access to quality healthcare (source). No doubt Entrepreneurs and Engineers will be creating major innovations in this space. Related: IBM and Manipal announcement

a16z is talking about Digital Therapeutics
Behavioral change is an area Startups/Developers/Apps have and will continue to embrace. A positive signal is a16z’s movement into this space asVijay Pande talks about in this interview. In this class of app, Email/SMS/Push Notifications/Phone calls are the engagement mechanisms.

VC Investing has increased == Startup activity is very hot
A few months ago Rock Health published their 2015 Healthcare Funding report, a must read. Combine this funding data with a review of new Healthcare products on Product Hunt,new Healthcare companies raising capital on AngelList and scanning Dan Primack’s Term Sheet or any other funding source and you will have a good grasp of the pace.

Regulation continues to provide opportunities
Today, major Health IT spend is in certified electronic health record (CEHRT) technology needed to comply with the federal meaningful use (MU) program, better security systems, and ICD-10 conversion software. Coming soon, additional legislation from the Protecting Access to Medicare Act kicks in mandating “that starting January 1, 2017, physicians ordering advanced diagnostic imaging exams (CT, MRI, nuclear medicine and PET) must consult government- approved, evidence-based appropriate-use criteria, namely through a CDS system.” (source)

Another helpful way to look at the Healthcare VC space is to think about the trends and contrast with VC investments.

Macro-Trends

  • The Consumerization of Healthcare
  • Consolidation of and competition between Hospitals and Integrated delivery systems
  • Strategic Investing (ex: Mayo investing in Helix)
  • Monitoring and Prevention
  • “Obamacare” disruption

Healthcare Funding Categories

  • Health IT Software
  • Digital Health
  • Medical Devices
  • Payer Disruption
  • Biotech

For further reading, I recommend checking out the various portfolio companies from Rock Health, Kapor Capital, SafeGuard, Arsenal and GV. Also checkout the Accelerator programs like Techstars Cedars-Sinai andMore Disruption Please to get a feel for the early stage. CB Insights is always publishing great insights such as this Healthcare IoT market map.

Here’s a brief sampling of some investments I’ve seen recently in these areas:

  • Patient Engagement
  • Prescription Management
  • Healthcare Analytics
  • Genetic Testing
  • Elder Care
  • Life Sciences
  • Medical Devices
  • Nanotechnology
  • Biotech
  • Insurance
  • Gene Therapy
  • Health and Wellness
  • Digital Health
  • Health IT

In Health IT software:

Care Coordination: Patientping, HealthLoop
Payer Management: Oration
Data Analytics: Medivo, BeneStream
EMRs: Elation

In Digital Health we see:

“Communities” like:

  • Health coaches (eating, personal trainers, etc)
  • Community for X
  • Competition
  • Reviews and Ratings
  • Connecting Providers (Patientping), Caregivers to Seniors (Honor)
  • Crowdsourcing data (Human Dx)

“Monitoring” is a huge category including:

  • “smart devices”, watch, smart phones
  • insights, behaviorial analytics
  • personal health and nutrition assistants
  • DNA and other self testing (23andme, Helix, uBiome)

Health and Wellness Platforms like ShareCare, Welltok and Omada Health.

Products like monthly food (Birchbox) and care packages (Citrus Lane).

Content like articles, daily emails and health guides (HealthSherpa)

Dev Tools like HIPPA data stores (TrueVault, Aptible, Catalyze.io) and IoT data streaming (Sense360).

In Biotech we see…

Biorepository, Genetic Analysis, Cellular Models, Regenerative Medicine, Bioinformatic Analytics.

Startups are using new techniques to harden defensibility into their business models such as creating Developer Ecosystems and baking in data network effects. A good example of data network effects at work is Recombine, a genetic testing company. They have built a network of partner clinics that administer its tests; with each new test, Recombine gathers more DNA data which (with appropriate consent) it can run machine learning on to improve its tests and nimbly develop new ones (therefore gathering more data). Recombine uses Machine Learning tools to find and learn patterns in historical data and uses these patterns to generate predictions. Recombine is 4 years old and has raised $3.3M.

I hope this post gets your brain spinning on all of the opportunity and innovation that’s happening in healthcare.

Using Trigger Lists in Product Management

I’m a big fan of “trigger lists”. The exercise of building them and the value they bring to a Mind Mapping or Design process have proved beneficial to me over the years. One of my favorites is David Allen’s GTD Incompletion Trigger List.

Recently, I transitioned from obsessing over providing Developers with APIs that would help them build amazing things with AI to obsessing about Healthcare and how AI can provide better care while lowering costs.

I pounded a Doppio and spent an hour brainstorming this trigger list to help me empathize with Users and better understand Actors in the crazy ecosystem that is today’s Healthcare tech.

I am a…

Healthy person
Cancer survivor
Farmer
Factory Floor Worker
CRO Administrator
CIO
CFO
CEO
Developer
Product Manager
Auditor
Patient
Physician
Nurse
RN
PA
Administrator
Researcher
Daughter
Son
Parent
Community Oncology Clinic
Hospital CEO
CMS Employee
FDA Committee Member

And I have…

Outcome data
Clinical trials
Drug databases
Medical journals
App Store Reviews
Medical Devices
Demographics
Avatars
Full Contact API data
Clinical Trial Participants
Patient data
Lab results
Population data
Reimbursement data
Patent filings
Hunches
Students
Research and Health kit data
Hospital trends
Emails
Tweets
Blog posts
Survey results
Internet searches
Essays
Product reviews
X-rays
Photos
Instagram searches
A list of questions

And I want to…

Find Patterns
Organize my data
Filter my data
Search my data
Understand social media
Build an Android app
Surface correlations
Have access to information

So I can…

Comply with regulations
Stay up-to-date
Collaborate with a Physician
Track my progress
Get credit for a course
Be reminded of an appointment
Find cost savings
Sell an app
Make people healthier
Prove a point
Get reimbursed
Understand health trends
Track my Clinical Trial
Find a Hospital
Research and buy my medication
Predict outcomes
Make more money
Connect data together
Build a treatment plan
Find a Clinical Trial
Predict the Future
Support Meaningful Use
Make evidence-based clinical decisions
Analyze adverse events
Provide better treatment “in the field”

For those familiar with Agile, you’ll recognize the “As a User I want” format of this trigger list.

We all have so much stuck in our heads, try creating one of these trigger lists for something in your world and you’ll be surprised at how it can help.

Why a Product Manager Needs To Wander

Inbox zero, one-on-ones, daily sync calls, daily standups, weekly meetings, roadmap planning, sprint planning, quarterly planning, exec offsites…oh, and…screencasts, wireframes, designs, qa testing, a/b testing, analytics, customer interviews, customer feedback, support tickets, user stories…and on and on it goes.

Whew.

I love to travel for work and always have. I love to hike and ski by myself too. I like doing things like walking my dog, running, riding my bike and riding the bus alone.

I like to wander.

As you walk through the airport, take a moment and observe people and think about their day. Where are they from, how did they start their day, did they drive to the airport or did someone drop them off, what is stressing them out in their life right now? See that annoying person over there? See that pissed off looking person waiting in the security line? How about the friendly looking older couple walking slowly to their gate? What do you think could make their life better? What business are they in? What experience or product do you think significantly impacted their career ten years ago?

Here’s what Brad Feld was thinking about this morning in the San Jose airport.

As a Product Manager, this is one of the most important things I do even though most people don’t get it and ROI can’t be tied to it. Being empathic is often cited as an important trait of a Product Manager. Wandering helps build this muscle. When I wander, I start to see patterns, feature ideas flow and I meet people and have serendipitous interactions about new ideas that aren’t possible in most contexts.

Establishing cadences, rhythms and process is key to succeeding as a PM, especially as your team scales. Just remember to break yourself out of those molds from time to time and go get lost.

Next week I’ll be in NYC visiting the IBM Watson team at Astor Place. My early mornings and late nights won’t be totally packed with meetings and dinners, I will protect that time and wander around. Who knows what I’ll think of.