Six connected experiments · 2023—now

My evolution as an entrepreneur

Noah Tanna

A record of six connected experiments, from a tutoring plan to a new information product. Each one taught me something specific, and that lesson shaped the next project.

Idea → action → reality → next experiment

The method

01
Idea
Something I believe is true.
02
Build
The smallest real version of it.
03
Reality
Who actually shows up, pays, stays.
04
Revise
The belief changes, or the project ends.

Every chapter shows what I did, what happened, what I learned, and why it led to the next experiment.

Chapter 01

Tutor Tanna

A tutoring company I planned completely before I ever spoke to a parent.

What I wrote down

K–5: Narrowed from education, to tutoring, to group tutoring for elementary students
$30: Per session, phased in after three to five months of building proof
2×: Sessions a week, at home, at school or online, whichever removed the most effort
0: Paying parents

Testimonials and case studies were meant to make the result feel certain. I had neither, because I had no students.

What changed my mind

I had built a business.I had built a plan, but no one had asked for it.

What I learned

“Before deciding what to build, I needed to understand the people I wanted to serve.”

Failed — no paying customers

What came next

Plan first, assume demand

Research the problem
Talk to the people it affects
Then decide what to build

What I learned · Ch 01

“Before deciding what to build, I needed to understand the people I wanted to serve.”

Next · Chapter 02 ↓Bridging the Digital Divide

5 owner interviews

Chapter 02

Bridging the Digital Divide

I researched why older small-business owners were being left behind—and what useful support would look like.

Market research + five semi-structured owner interviews

What the research revealed

5/5: Said an online presence matters to them
3/5: Depend on a younger relative or employee to run it
4/5: Named cost as the blocker
4/5: Said the tools they were sold were too generic to fit

The wider research set the context. Five owner interviews showed how the problem felt up close.

What the owners told me

“They sold me clicks. I never saw a single customer walk through the door.”Owner, interview 04

What I learned

“I didn't want to sell businesses more clicks. I wanted to help them solve the problems holding them back.”

What came next

Another agency selling clicks and packages

Work directly with each owner
Websites first
Then marketing, automation and AI where they help

What I learned · Ch 02

“I didn't want to sell businesses more clicks. I wanted to help them solve the problems holding them back.”

Next · Chapter 03 ↓N&B Business Uplift

Chapter 03

N&B Business Uplift

Not an agency selling clicks. Months inside each business, learning what the owner actually needed and building around it.

Six client partnerships

Six businesses. Six different definitions of useful.

Elliot's Dry Cleaners homepage — Elliot's Dry Cleaners

Gonzalez Framing homepage — Gonzalez Framing

Back Alley Thrifts storefront homepage — Back Alley Thrifts

Daniel Armstrong homepage — Daniel Armstrong

Generation to Generation homepage — Generation to Generation

Not clicks. Partnership.

We spent months in each client’s shoes—then built the proof, path, and follow-up that could turn attention into a real customer.

Elliot's Dry Cleaners reviews section with testimonials, hours and contact details — Reviews, hours, address, phone — the page that closes the sale

What changedDrag to move from what I sold to what I ended up selling.

Year one — websitesNow — systems

A pipeline that books the job
Marketing as follow-up
Automation that does the chasing
AI systems the owner can run alone

A site to be found: → A pipeline that books the job
Marketing as posts: → Marketing as follow-up
Manual admin: → Automation that does the chasing
Design decisions: → AI systems the owner can run alone

Still runningThe company that came out of research—and stayed close to the owner.

Live site

nbbusinessuplift.comopen ↗

nbbusinessuplift.com — loads when you reach it

The lesson

The service I first imagined selling was not always the most valuable one. I started with websites and expanded into marketing, automation and AI because clients needed them.

What I learned

“I began looking beyond what technology could do to what it could make possible for people.”

What came next

Improving one business's operations

Extending one person's knowledge
To people they can't reach directly

What I learned · Ch 03

“I began looking beyond what technology could do to what it could make possible for people.”

Next · Chapter 04 ↓Find a Tree / Mentor AI

Chapter 04

Find a Tree

Daniel Armstrong’s philosophy is “find a tree and start.” I joined as a strategic partner and AI specialist, not as founder.

What I madeA mentorship platform: his method, delivered without him in the room.

The partnership

What it became

Mentor AI — Daniel’s knowledge, made available beyond the room.

Live site

ai.danielarmstrong.comopen ↗

ai.danielarmstrong.com — loads when you reach it

What I recognized

Mentor AI extended Daniel’s knowledge beyond the people he could meet in person. I worked on it as a strategic partner and AI specialist.

What I learned

“Could the same idea work for creators, letting their audiences engage with their knowledge beyond what they publish?”

Find a Tree remains unfinished

What came next

One mentor, one audience

Many creators
Each with their own content
Each with an audience asking questions

What I learned · Ch 04

“Could the same idea work for creators, letting their audiences engage with their knowledge beyond what they publish?”

Next · Chapter 05 ↓Creator AI

Chapter 05

Creator AI

An AI trained on one creator's own content, so their audience could ask it questions and get advice in that creator's voice.

The ideaConcept illustration — no product was built

Everything a creator has already said, answerable for every follower.

01
Creator
One person whose advice people already follow.
02
Content
Videos, posts, captions, replies: their real advice, not generic advice.
03
AI
Trained only on that content, answering in the creator's voice.
04
Audience
Each follower asks their own question and gets advice that points back to the creator's work.

A follower asks: “I have 400 followers and no idea what to post. What would you do?”

The answer draws on the creator’s own advice about early growth, with links back to the videos it came from.

Why I did not pursue it

Audience interest meant there was a business.The audience wanted it, but the creator would have to pay, and creators could build it themselves.

What I learned

“A compelling idea wasn't enough. I needed to understand the problem deeply and decide whether my solution offered distinct value.”

Rejected — not pursued

What came next

Creator AI: an appealing concept

A different problem: finding relevant, trustworthy financial information
Customer discovery before product decisions

What I learned · Ch 05

“A compelling idea wasn't enough. I needed to understand the problem deeply and decide whether my solution offered distinct value.”

Next · Chapter 06 ↓Noded

8 customer interviews

Chapter 06

Noded

There is no middle ground between a long mainstream article and a fast, unverified social feed. Noded restructures events into nodes outlining causes and consequences. I joined the team early, brought on for my previous entrepreneurial experiences.

Customer discovery — eight interviews I ledRecorded in our interview database, then coded for what repeated and what pushed back.

Eight interviews supported the need. They did not prove people will adopt it or pay.

What repeated

Less volume and more relevance to their industry and market.
Short sessions with context, causes and consequences made clear.
AI can help, but every claim needs a visible source.

Not yet established

Whether people will switch to it from their current sources.
Whether anyone will pay for it, and how much.
Whether a short format can still feel complete and trustworthy.

Where it stands

Eight interviews supported a new way to consume news: less volume, more relevance, and events restructured into nodes that make causes and consequences clear, with visible sources. They have not shown adoption, product-market fit or willingness to pay. Our next step is testing a focused product with real users.

What I learned

“The interviews defined how news should be restructured before we built it.”

In development — not yet validated

What came next

Six separate experiments

One approach
And the people it helped

What I learned · Ch 06

“The interviews defined how news should be restructured before we built it.”

Next · Chapter — ↓What it adds up to

How I evolved

Six experiments, one approach.

Each one changed how I approach problems, customers, technology and whether a business can work.

01 · Tutor TannaMove beyond planning and test assumptions.
02 · The researchListen to the people I want to serve.
03 · N&B Business UpliftTurn those needs into practical solutions.
04 · Mentor AIUse technology to extend human expertise.
05 · Creator AIAsk whether an appealing concept can support a business.
06 · NodedLet evidence guide what the product becomes.

Who it helped, and what I'll pass on

Small-business owners

N&B gave small-business owners websites, marketing and automation built around what their businesses actually needed.

A mentor's reach

Mentor AI explored how Daniel's mentorship could reach people well beyond the conversations he has directly.

People following the news

Noded restructures events into nodes that highlight causes and consequences — our redesign of how news is consumed, still being tested.

I want to carry these lessons into future ventures and pass them on to the people I work with, mentor and help. What I learn matters most when it helps someone else make progress.

Start with people. Test with evidence.Build what creates value beyond me.

nbbusinessuplift.com ai.danielarmstrong.com