Build test
Build pattern 1 from scratch for a new imaginary business in under a day, with an AI assistant but no tutorial.
Rung 0 → 1 · The Gap, in depth
Get one stranger to pay you.
This page goes one level deeper than The Gap, on its first rung only: the minimum you need before you start, and how to set up the first real loop so you can run it, measure it and improve it every week.
Rung 0 → 1At a glance
- Milestone
- A stranger pays you
- Hard part
- Trust, no record
- Rough time
- 2–4 months
- Prereqs
- 4 groups · 4–6 wks
- First loop
- Set up, run weekly
Tags as on The Gap. reasoning follows from first principles. pattern recurs in shared freelancer stories but isn't measured. assumption is a starting number to replace with your own.
Chapter 01 · Part I
The milestone, exactly
Rung 1 = someone with no personal obligation to you pays an agreed price for work, because of the work.
Money from a friend or relative doesn't prove that strangers will trust you, because they're paying partly out of the relationship. The milestone is about proving that the work earns trust. reasoning
| Situation | Counts as rung 1? | Why |
|---|---|---|
| Your uncle pays you for his shop's system | No | Paid partly because of the relationship. It still counts as proof (Chapter 04) |
| Your uncle's supplier, introduced by him, pays you | Yes | The introduction got you a hearing. The supplier paid for the work |
| A shop you walked into pays you | Yes | No relationship at all |
| A classmate's startup pays you a token amount | Borderline | Count it only if the price was set by the work, not by the friendship |
| A free project for a stranger | No | Useful proof, but no one has yet decided the work is worth money |
So friends and relatives are still useful. They're your practice project and your first introductions. They just aren't the finish line.
Chapter 02
Trust without a track record
A small business owner says yes when the value they expect is clearly bigger than the price plus the risk they feel. reasoning
Yes when expected value > price + perceived risk
A track record mainly lowers perceived risk. Without one, you lower risk directly, by removing each specific fear the owner has. This is the trust stack: use all of it on your first stranger.
| The owner's fear | What removes it | What it looks like in practice |
|---|---|---|
| "It won't work for us" | Show before they pay | A working demo of their situation, built after the first conversation, shown on your phone |
| "I'll pay and lose the money" | Pay after it works | For the first stranger: full payment after it runs on their real data for a week |
| "You'll disappear" | Ownership and handover | Everything runs in their accounts. A written handover and a short video. A 14-day fix window |
| "It'll turn into a big mess" | Small, written scope | One problem, one outcome, one week, one page |
| "You don't understand my business" | Specificity | Repeat their problem back in their words, with the cost they told you, better than they said it |
| "Who are you?" | Borrowed trust and proximity | An introduction from someone they know, meeting in person, a local number, their language |
| "Why so cheap? What's the catch?" | Honesty | "I'm early in my career. That's why I take small fixed projects and you pay after it works." No invented experience |
Every device above is available to a student on day one. None needs a track record. Together they make saying yes feel safer to the owner than saying no. reasoning
Chapter 03 · Part II
A delivery kit
The capability prerequisite isn't "learn web development". It's a small set of solution patterns you can deploy for a business in about a day each. Most small-business problems you'll hear in the first loop fit one of these. reasoning
| Pattern | Problem it fixes | Built with (examples) |
|---|---|---|
| 1 · Enquiry capture | "Enquiries come by call, WhatsApp and Instagram, and some get forgotten" | A form or WhatsApp click-to-chat link → Google Sheet → instant reply to the customer + alert to the owner (Apps Script or n8n) |
| 2 · Follow-ups and reminders | "We forget to call back, chase fees, confirm appointments, remind about renewals" | Sheet of due dates → scheduled email, SMS or WhatsApp message (through a WhatsApp Business Platform provider, paid by the client) |
| 3 · Daily or weekly report | "I don't know today's numbers until the month is over" | Sheets or Tally exports → a simple dashboard (e.g. Looker Studio) or a summary message every morning |
| 4 · Document extraction | "Someone types bills, invoices or forms into Excel by hand" | Photo or PDF → LLM API extracts the fields → rows in a sheet, with a human checking each row |
A simple website with a working form is also sellable, but it's commodity work now. Use it as the front end for pattern 1 rather than as the offer.
The skill floor: three tests
Break-and-fix test
Break something on purpose (renamed column, expired token, wrong phone format). Find and fix it from logs within an hour.
Explain test
Draw the whole flow on paper and explain it to a non-technical person, who can then explain it back.
The reliability floor
These separate a professional from someone who built a demo. You need them before a stranger's business depends on your work:
- Their accounts: it runs in the client's Google or WhatsApp accounts, not yours.
- Alerts: failures send a message to your phone.
- Data export: their data can be exported, and a copy exists.
- No passwords in shared sheets or code.
- A handover: one page plus a 3-minute video.
Learning route: build each pattern twice, once for a practice business and once from scratch for a different business type. Aim for patterns 1 and 2 first, then 3 or 4. That's roughly 3–6 weeks at full time. assumption
Chapter 04
Proof and presence
One practice project, run like a paid one
Find one real business through family: a relative's, a neighbour's, or one they know. Treat it exactly like a paid project: a written scope, delivery, a measured before and after, and a handover. Ask for a written quote and permission to share the numbers. It doesn't count as rung 1, but it gives you one real case study before you meet a single stranger. reasoning
Three demo videos
One video per pattern, 2–3 minutes each, recorded on your screen and phone, in the language your buyers speak:
Demo video
- 0:00 · The problemIn the buyer's words: "Enquiries come in on WhatsApp and some never get a reply."
- 0:20 · What it costsMissed admissions, lost orders, hours of typing, late payments.
- 0:40 · The demoA customer enquires, the reply goes out instantly, the owner gets an alert, the row appears.
- 2:00 · The resultWhat changes, as a number where possible (e.g. "no missed enquiries in the practice project's first week").
- 2:20 · Next step"If this sounds like your business, message me. I'll show you a version for your case."
A place to point people: one day's work
A one-page site
"I help [buyer type] stop [problem]." The three videos. How it works in three steps. A "starting at" price. An honest line about you. A WhatsApp button.
WhatsApp Business profile
Business description, a catalog with your three offers, a greeting message, and quick replies for the questions you get most. This is where Indian small-business owners will actually contact you.
Chapter 05
The offer and money rails
Write your first offer on one page before you need it, so you're not making up scope and price in front of an owner.
First offer
- The problemIn their words, with the cost they told you.
- The outcomeOne sentence: "Every enquiry gets a reply within a minute and lands in one sheet."
- What's includedThe specific pieces, and 30 minutes of training in person.
- Not includedAnything you discussed but aren't doing. Third-party fees are paid by them, directly.
- TimelineLive within 7 days of getting access.
- Price and paymentA fixed price, paid after it has run on real data for 7 days.
- After delivery14 days of free fixes. Optional monthly care after that.
- What I need from youAccount access, sample data and 30 minutes for testing.
Pricing the first stranger project
About ₹5k–15k for one pattern for one business, plus optional monthly care at ₹1k–3k. assumption The rule that matters more than the number: when they push back on price, shrink the scope, not the price. A smaller project at a fair price protects the proof. A discount teaches you and them that the work is worth less. reasoning
Payment terms, and why "pay after it works" for the first one
At rung 0 the barrier is trust, not your cash flow. Paying on completion removes the owner's biggest fear. The small, written scope caps your risk to about a week of work. From the second or third client, move to 50% upfront. reasoning
Money rails for India
- Payments: UPI for local clients, plus a simple numbered invoice (Google Docs or Zoho Invoice is enough).
- Agreement: the one-page offer, accepted on WhatsApp or email ("Yes, go ahead"), is your written agreement at this size.
- Tax: keep a sheet of every payment from day one. Book one CA consult before your first international client or before income becomes meaningful. GST registration generally isn't required below the ₹20 lakh threshold, but confirm it with the CA.
Chapter 06
Your tracker, and what to skip
One Google Sheet with four tabs is the whole operating system for rung 0. It's what turns "I'm trying freelancing" into a loop you can measure and improve.
1 · Targets
Every named business: name, area, contact person, phone or WhatsApp, source, how you'll reach them, status, next step and date
2 · Problem log
Each conversation: date, business, the problem in their words, how often, what it costs, current workaround, would they pay, which pattern fits
3 · Pipeline
Each opportunity: demo shown? offer sent? yes/no/maybe? price, due date, paid?
4 · Weekly numbers
Contacted, conversations, problems logged, demos shown, offers made, yeses, paid, hours spent building vs. talking
Not needed yet
Each of these feels productive and delays the first conversation. Skip them until after rung 1. reasoning
Registering a company
Work as an individual. Revisit it when income and clients justify it.
A logo, brand or multi-page site
Buyers decide on the demo and the conversation, not the logo.
More courses and certificates
Learn what the next real problem needs, when it needs it.
DSA and interview prep
Useful for jobs. It has almost no effect on whether a shop owner pays you.
Upwork and Fiverr profiles
Useful later, with a niche and proof. At rung 0 you compete with everyone, globally, on price.
A content plan or personal brand
An audience comes mostly after results. Post your case studies later.
Picking a niche up front
The niche comes out of the problem log (The Gap, Chapter 06).
A perfect, general-purpose product
Build what one business needs this week. Reuse comes from patterns, not from a platform.
Chapter 07 · Part III
Choose where to start
The first loop runs on one buyer type, chosen by access and the density of pain, not by passion. Score 2–3 candidates from 1 (poor) to 5 (strong). Access and pain count double, because they decide whether the loop runs at all. reasoning
Pick the top one as your main buyer type and the second as your backup. You switch to the backup only through the rule in Chapter 12, not because a week felt slow.
Chapter 08
Build the target list
Before the first conversation, put 60–100 named businesses of your buyer type into the Targets tab. A list turns "I should find clients" into "I'm contacting the next five rows today". assumption
Where to find them
Google Maps ("coaching institute near [area]"), JustDial, IndiaMART for traders and manufacturers, local association member lists, Instagram location tags, notice boards around colleges and markets.
Who to start with
Businesses within travel distance first, then those reachable by introduction, then online-only. Medium-sized ones (a few staff, visibly busy) usually feel repeated-work pain most.
Ways to reach them, best first
| Channel | Why it works at rung 0 | How |
|---|---|---|
| Introduction | Borrowed trust: the strongest device you have | Ask your practice-project owner, relatives, professors: "Who do you know who runs a [type]?" Send them a two-line message they can forward |
| Walk-in | Proximity. Hard to ignore, easy to trust a face | Quiet hours only (mid-afternoon for most shops). Ask for 10 minutes, or when to come back |
| Associations and events | Many owners of one type, all in one place | Trader associations, coaching-owner groups, local business meetups |
| Personal message | Scales, but weakest trust | WhatsApp or Instagram. One specific line about their business plus a 60-second video. Never a template blast |
Ordering is reasoning from the trust stack in Chapter 02. Your own numbers will show which works best in your city.
Chapter 09
The conversation
The first conversation isn't a sales call. Its only job is to find a specific, recent, costly, repeated problem and earn a second meeting. Adapt the wording to your own language.
Conversation guide
- Open"I'm ___, I study ___ and build small automations for [type] businesses. I'm not selling anything today. I'm trying to understand how [type] businesses handle enquiries and follow-ups. Can I ask five quick questions?"
- Their routine"Walk me through a normal day. What happens when a new customer contacts you?"
- Repetition"What do you or your staff do again and again, on the phone, WhatsApp, Excel or paper?"
- Leaks"Where do things fall through: missed calls, forgotten follow-ups, late payments, typing errors?"
- A real instance"When did that last happen? What did it cost you?" Ask for a recent real case, not a hypothetical.
- Workaround"How do you handle it now? Who does it? Have you tried any software for it?"
- Value"If that were fixed, what would change for you?"
- Close"Can I come back next week and show you a small working version for your case? Seeing it costs nothing." Then: "Who else in your line of work has this problem?"
- Listen about 80% of the time. Write down their exact words in the problem log the same day. Those words become your offer and your next opening line.
- Don't pitch and don't quote a price in the first conversation unless they ask. If they ask, give a range and move on to the demo.
- A good conversation ends with a problem that's specific (a real, recent case), costly (money, customers or hours) and repeated (weekly or more), plus a date to come back.
Chapter 10
Demo first, then offer
Within 7 days, build a demo of their situation from your delivery kit, using sample data that looks like theirs: their business name, their kind of enquiries. Show it on your phone or laptop. This single step replaces most of the track record you don't have: they've now seen it working before being asked to trust you. reasoning
Replay the problem
"Last time you said about 5 enquiries a week get missed, and each lost admission is around ₹20,000. Here's what I built for that."
Show it working
Let them send a test enquiry from their own phone and watch the reply and the alert arrive.
Make the offer
The one-page offer from Chapter 05, filled in with their words. One price. Pay after it runs for 7 days.
Ask for a decision
"Shall I set it up this week?" If they need time, agree on a specific day to hear back. Don't leave it at "I'll think about it."
What owners usually say
They say
"Too expensive"
Shrink the scope, keep the price per piece: "Let's start with just the enquiry replies. Follow-ups can come later."
They say
"Let me think about it"
"Of course. Can I call you on Thursday?" Put the date in your pipeline. Follow up twice, politely, then move on.
They say
"What if you disappear?"
"It runs in your own account, you get a handover video, and you only pay after it has worked for a week."
They say
"We tried software before"
Ask what went wrong. Show how this is smaller, fits their existing WhatsApp and sheets, and needs no new habits.
They say
"My nephew can do this"
"Great. If he gets stuck, I'm here." Stay friendly. Ask who else might need it.
They say
"Do it free, then we'll see"
Your free project was the practice one. Offer the pay-after-it-works terms instead. That's already risk-free for them.
Chapter 11
Deliver, prove, ask
The first paid project is also where the next loop's proof and reach come from, so the delivery includes capturing both.
- KickoffAccess to their accounts, sample data, the "before" number written downDay 1
- Build and testBuilt from your kit and tested with their real data, including one deliberate break-and-fixDays 2–5
- Go live and train30 minutes in person with whoever will use it. Handover page and videoDays 5–7
- Run for 7 daysYou watch the alerts and fix issues the same day. Payment is due at the endDays 7–14
- Measure and capture proofThe "after" number, a testimonial, and permission to shareDay 14
- Ask and offer careA referral request with a forwardable message, and an optional monthly care planDay 14
Three testimonial questions
What was the problem before? What changed after? Who else would you recommend this to? A 30-second WhatsApp voice note or video is more convincing than typed text.
The referral ask
"Who are two people in your line of work with the same headache?" Then hand them a two-line message they can forward, with your demo video link.
Chapter 12
Weekly cadence and iteration
The loop improves only if you run it on a fixed rhythm and change one thing a week based on the numbers. About 40 protected hours a week: assumption
| Day | Morning | Afternoon |
|---|---|---|
| Monday | Weekly plan. Add targets to the list | Messages and introduction requests |
| Tue – Thu | Build: kit drills, demos, delivery | Conversations: walk-ins, calls, visits |
| Friday | Show demos and make offers | Delivery and follow-ups |
| Saturday | Skill drill on the pattern you're weakest at | Weekly review (below). Update the numbers tab |
Weekly targets while the loop is running
15–25 contacted → 5–8 conversations → 5–10 problems logged → 1–2 demos → 1–2 offers
Starting assumption numbers. After four weeks, replace them with your own averages and use the loop calculator to project your first client.
Iteration rules
Rules decided in advance stop you from changing everything after one bad week, or nothing after ten. reasoning
| When this happens | Do this |
|---|---|
| Contacted many, few conversations | Change channel, not buyer type: more introductions and walk-ins, fewer messages |
| 20 conversations, no problem heard 3+ times | Switch to your backup buyer type |
| A problem heard 3+ times | Make it your lead offer. Build its demo in advance and mention it in your opening line |
| 5 demos, no yes | Fix the offer: smaller scope, pay-after terms, and check the problem is actually costly to them |
| Yes, then silence or no payment | Tighten the written scope and acceptance, and get the "go ahead" in writing |
| Delivery takes over 2 weeks | Offer only problems your kit already covers. Add a pattern to the kit before selling it |
| Building over 60% of your hours after week 6 | Move hours from building to conversations. The loop is starved of input |
The weekly review: four questions, 30 minutes
- What did I learn about my buyers this week, in their words?
- Which problem came up again?
- Which step of the loop leaked most (see the numbers tab)?
- What's the one change I'll make next week?
Chapter 13 · Part IV
Launch checklist
The minimum before the loop is fully running. Start conversations once pattern 1 is built, the practice project is scoped and your buyer type is chosen, even if the rest isn't done. Your ticks are saved in this browser.
0 of 0 ready
Capability Delivery kit
Proof Something to show
Commercial Offer and money
Loop Ready to run
Chapter 14
The first 16 weeks
This is the 2–4 month rung laid out week by week. Build and talk in parallel from week 3. The dates are a guide; the exit points matter more. assumption
- Set upPatterns 1–2 started. Tracker built. Buyer types scored. Practice project scoped. Target list of 60+Weeks 1–2
- ProofPatterns 3–4. Practice project delivered and measured. Three demo videos. Presence live. First 10 conversationsWeeks 3–4
- Loop at full speed5–8 conversations a week. Problem log filling. First demos shownWeeks 5–6
- First offersDemos turn into one-page offers. Iteration rules applied weeklyWeeks 7–8
- First deliveryA first yes delivered within 7 days and run for 7 daysWeeks 8–10
- Rung 1First payment from a stranger. Proof captured. Referral asked for. Care plan offeredWeeks 9–12
- BufferIf there's no payment yet, find the leaking step (Chapter 12) and fix one thing a week. This is still inside the 2–4 month rangeWeeks 13–16
After rung 1, the next rung (1 → 3) is mostly the same loop again, now with a real case study and referrals. See The Gap's 6-month plan for how it continues.