We’re back
Y’all - I can’t begin to share how excited we are to be posting here again.
A few months ago, we paused our regularly scheduled experimenting to embrace a period of more open, free-flowing exploration - with continued, spontaneous, experimenting sprinkled in, of course. We’ve tried a lot, learned a lot, and want to briefly recap what we’ve done and where we’re going.
Recap
Following a
- Research (talking to people)
- Discovery (building an MVP for those people)
- Launch (sharing that MVP with a larger audience)
- Test (feedback and $$$ value)
process, we’ve cycled through about a dozen micro-products, including:
-
billy.health (Automated & Free Medical Billing for Private Practices)
- TLDR: Leverage the low / negligible marginal cost of AI Medical Billing to offer it to private practices for free. Make $$$ by selling their anonymized & aggregated live billing data to Fair Market Value & Medical PE firms
- Wins
- Landed a partnership with HCA (one of six major FMV firms in the US) to provide live billing data
- Built MVP AI Medical Coding Assistant with ~85% accuracy (compared to ~95% human accuracy)
- Tested with ~20 private orthopedic practices & narrowed in on the need for lower-cost, higher-simplicity RCM management tooling
- Why’d we move on?
- Data not valuable enough (clients expect to pay ~$1000 / data segment).
- Low demand for high volume / live data, ~100 data points is all that is needed (billing rates)
- Not in our area of expertise and interest
-
dumplinks (PDF & print your reading backlog)
- TLDR: Short project, super simple: dump your reading backlog as links & get them nicely formatted, bound, and printed with table of contents + summary. Spurred by this tweet, we had IRL and online friends who said they’d be willing to pay for something like this
- Wins
- Super quick turnaround (3 days spent building)
- Made ~4 books @ ~$20 a book
- Why’d we move on?
- No willingness to pay (i.e., when actually sent a payment link, little conversion - kind of an “ehhhh I shouldn’t” response)
-
nutty.club (anupam’s profile) (Track calories + exercise while going nutty for nutrition; linked page is my profile, log in to your own here: nutty.club)
- TLDR: App that hooks into smart scales, GCM, etc.. & uses your exercise level to calculate ideal calorie intake for weight loss
- Wins
- Daily active users (Four, not including Anupam); high retention amongst people who care & were already doing this manually (Google Sheets, Excel, etc…)
- Shareability (anecdotal but friend group checks/sends profiles every day)
- Why’d we move on?
- Tried moving this into a wider audience but nutrition is a very niche and fragmented group. Everyone tracks different metrics differently
- High saturation of “apps” that already exist
Each was a rapid research / discovery / launch / test of a tiny product for a small but passionate audience - the result of a friendly conversation or user interview that turned into our first customers. For those that we’ve endlessly bothered for feedback, thank you.
And while each of these micro-launches has plenty of room for refinement and growth, none of their initial users reflected a level of passion that exceeded the bar for us to continue pursuing. Each was conceived, born, launched, and paused within a week. Except for one -
Email: The Angry and Habitual
What started as an experiment to scan your email inbox for previous orders & issue instant cash offers turned into folks ranting at us about their inbox. From confusion to boredom to anger, I have never seen individuals express so much disdain for something they use for at least 5+ minutes (personally - non-work) a day.
Folks told us:
Email is the most boring thing I do every day
I hate clearing my inbox
I am currently doing a [Civilization V style] infrastructure project to fix my [email]
I get random notifications all day that distract me
But also said:
I use my email to keep track of stuff I’ve bought” & “I mark [orders] unread so I can track when they’re arriving
I know I have free stuff in my email I can just never remember
I really love getting updates from the LA Philharmonic… and half my time is spent sorting those out from the junk
It’s the only place where I can get through all of my news in the morning [and forward them to my daughter]
…I set reminders in Superhuman for tasks I have to do later… [like] emailing my leasing agent or tracking a check I’m expecting in the mail
This is conflicting: people hate their email inbox, but still
- Open it every day
- Use it for tasks and info they can do / get elsewhere
A few more examples from sales / business development conversations:
Email is for to-dos. We have a to-do app [Wrike] and Slack but not everyone uses it so we all just put [tasks] in emails to each other
We add subject lines like Client Off to sort email chains
I use a lot of external tools [like Copy AI] to write outbound emails since they’re up to date but it doesn’t have context
TL;DR: People use their email daily for things they can do without email. And they hate it. Why?
What’s Exciting to Us
Both Personal and Professional Email:
- Consistent daily ritual
- High willingness to explore alternate clients / apps (we consider Gmail / Outlook as ‘defaults’)
- Frequent hacks to hijack default inbox behavior (plugins, external tools, subject lines, marking as unread, reminders, labels, etc.)
- GenZ friends building custom tools on top of their email
- I.e. friend built a web app that aggregated and formatted their newsletters
- Consultants adding “Client-Off” to designate breakaway internal threads
- Using email as a task list
- Using email for file storage
- GenZ friends building custom tools on top of their email
Personal Email
- Few senders, high volume. I.e., a few senders do most of the sending volume in an individual’s personal inbox.
- Little overlap between two inboxes. In the top 15,000 emails in 6 inboxes tested, there was only a max overlap of ~100 senders between them.
- Clustered by category. Like transactions on a credit cards, most email in an inbox could be sorted into one of five categories: orders, promos, newsletters, notifications, and comms
- Your last 500 emails say a lot about you. In searching through our design partners’ inboxes, you can capture a pretty clear snapshot of a person - their demographics, interests, personality, etc - from their last 500 emails.
Professional Email
- Fragmented (i.e., lots of niche and particular use cases depending on segment) with high $$$$$$ value problems
- Average email ROI is approx. 37:1 in outbound sales
- High competition in email tooling but no dominant email client (outside of standard Gmail)
There are compelling problems in both consumer (personal) and enterprise (professional); we found personal email more exciting and started there.
Wrap Up & Methodology
There’s a lot more to share here but this post is already getting really long!
As we move forward, we want to focus more on publicizing our methodology. In other words:
Sharing publicly (reddit, hackernews, twitter, discord, etc.) Higher craft and quality (thorough, clear, polished) Posting conclusions (take a stance, see things through)
We focus on a cycle of research / discovery / launch / test. This post is strictly in the research phase. Next week will be focused on translating that research into a discovery with early prototyping of Soar V1.
Til then!
(Wouldn’t be an experiment post without a teaser for next week 😉. See below! )