The Un-SaaS Ch. 14 — The 10-Day Deployment Sprint

Chapter 14 of 26

This is a chapter from The Un-SaaS: A Toll Stack Engineer's Handbook. Each Friday, we're publishing a chapter as a bonus for our readers.

In the Apollo program, the engineers at NASA had a phrase: “All-up testing.” Instead of testing each stage of the Saturn V rocket separately — first stage, second stage, third stage, each in isolation over months — they decided to test the whole thing at once. Stack it up. Light it. See what happens.

The conventional wisdom said this was reckless. The engineers argued it was the opposite: testing each stage separately meant you learned about integration problems last, when they were most expensive to fix. Testing all-up meant you learned about them first.

The same principle applies to your first toll position. You don’t spend three months perfecting a landing page, then three months crafting an email sequence, then three months wiring the analytics. You build the whole thing in ten days — minimum viable, not perfect — and light it.

Why? Because the experiment log can’t start until the thing is live. And the experiment log is the thing that makes it good.

A Saturn V rocket on the launch pad. Engineers watch from a control room — testing the whole stack at once, not in pieces.

Before Day 1: The Immersion Week

The 10-day sprint assumes you already understand the partner’s niche. If you don’t, the sprint fails — not on the technical build, but on the content. A landing page that uses the wrong vocabulary. An email sequence that recommends the wrong product first. A pre-sell angle that sounds like a marketer instead of an insider.

Before the sprint clock starts, spend a week — five to ten hours — immersing:

Watch or read the partner’s last twenty pieces of content. Not skimming. Absorbing. Note their recurring phrases, their favorite analogies, the questions their audience asks in the comments.

If the partner sells a course or membership, buy it. Take it. Not because you need the information — because you need to understand how the partner teaches, what they emphasize, and what their paying customers value most. This is the intelligence that makes your pre-sell content specific rather than generic.

Read the comments. The audience tells you exactly what they’re confused about, excited about, and skeptical of — in their own language. That language goes directly into your headlines and email subject lines.

Build a voice file: twenty phrases the creator uses repeatedly, five topics they return to, three opinions they hold strongly. AI will use this file to draft copy. You’ll use it to review what the AI produces.

This immersion isn’t optional and it isn’t one-time. You’ll deepen it as the position runs. But the initial week is the difference between a landing page that converts at 28% and one that converts at 14%. Same infrastructure. Half the results. Because the words were wrong.

The Sprint

Ten days. One toll position. From zero to live traffic flowing through your infrastructure.

This is not a ten-day plan for perfection. This is a ten-day plan for live. Every element is minimum viable. Every piece will be improved by the experiment log over the following months. The goal is to get real data flowing as fast as possible.

Here’s the math worth internalizing before you start: ten days at three to four hours per day is thirty to forty hours of total work. That’s a single work week spread across two calendar weeks. At the end, you have a live toll position that can produce revenue for months or years — on infrastructure that cost $15 a month to run.

No SaaS MVP ships in forty hours. No consulting pipeline fills in forty hours. But a toll position goes live in forty hours because you’re not building a product. You’re building a bridge.

Day 1: Infrastructure Setup.

Set up your $15 stack.

Email provider account (Buttondown, Mailerlite, or Brevo free tier)
Landing page hosting (Carrd annual plan, or a static HTML template on Cloudflare Pages)
Redirect layer (Cloudflare Worker or free redirect tool)
Analytics (Plausible or self-hosted Umami)
Experiment log spreadsheet (Notion or Google Sheets — create the template with columns: date, hypothesis, variant, metric, sample size, result, learning)

These four layers — landing page, email, redirect, analytics — are the entire toll position infrastructure. Not five layers. Not twelve. Four. Everything you build for every future partner runs on the same stack. You’re setting this up once.

If you write code: deploy on Cloudflare Pages, Vercel, or Netlify free tier. You control every pixel, the page loads in milliseconds, and there’s no platform branding polluting your partner’s voice match.

If you don’t write code: Carrd at $9 per year is the fastest path to a single-page capture site. The trade-off is less control for faster build time. Either works. Pick and move.

Done means: You can log into all four tools. The email provider sends a test email to your personal address. The landing page host shows a blank page at a URL you control. The redirect layer resolves a test link. The analytics dashboard loads. The experiment log has its template columns. Total time: two to three hours.

Day 2: Partner Landing Page (Draft).

Build the landing page for your first partner’s audience.

Voice match the partner’s content (watch three of their recent videos or read their last five posts)
Write the bridge headline
Create three to five pre-sell slides (AI drafts the first pass, you polish)
Add the email capture form (one field: email)
Set up the redirect to the partner’s checkout page
Mobile-test on your phone

The landing page has seven elements in a fixed order: voice match, bridge headline, pre-sell content (context, differentiation, objection — three paragraphs), email capture form, value exchange, and handoff. Miss one and conversion drops. Add extras and you add friction.

The capture form goes after the pre-sell content, not before it. Placing the form above the content is like asking for someone’s phone number before introducing yourself. Single field — email only. You can ask for their name in the welcome email.

Total time: three to four hours. The AI agent handles 60-70% of the copywriting. Your job is framing and voice matching.

Done means: You have a live URL. The landing page loads in under two seconds on mobile. The headline references the specific product or topic the visitor expects. The pre-sell copy sounds like the partner wrote it, not like a marketer wrote it. The email capture form submits successfully and redirects to the partner’s checkout.

Day 3: Welcome Email Sequence (Draft).

Write the six-email welcome sequence from Chapter 11.

Email 1: Value exchange delivery (the thing the landing page promised)
Email 2: Story + light product mention
Email 3: Case study + warmer recommendation
Email 4: Direct recommendation with transparency
Email 5: Objection handling
Email 6: Last touch

Here’s the thing most operators get wrong: they write the welcome sequence like a hospitality brochure. “Welcome! Here’s a little about us. We’re so glad you’re here.” Nobody asked for that. The subscriber traded their email for something specific. Email 1 delivers it. Email 2 earns the right to keep showing up. Email 3 earns the first recommendation. The sequence is an escalation of earned trust, not a series of introductions.

AI drafts all six. You review, adjust voice, and load them into your email provider with the timing triggers (Day 0, Day 2, Day 4, Day 7, Day 10, Day 14).

Front-load the sequence. Three emails in the first four days, when attention is highest. By day fourteen, half the engagement is gone. That’s not a failure of your content — it’s the natural psychology of a new subscription. The first week is your monetization window. Don’t waste it on “about me” content nobody requested.

Total time: three to four hours.

Done means: All six emails are loaded in your email provider. The automation trigger fires on form submission. The timing delays are set. Each email has a single clear CTA. Subject lines reference specific value, not generic welcomes. You’ve sent yourself through the full sequence and read every email on your phone.

Day 4: Redirect Layer + Analytics Wiring.

Create redirect links for every outbound product link in your email sequence
Wire analytics events: page view, email capture, email open, link click, purchase (if trackable via affiliate dashboard)
Test the full flow: click the partner’s link → arrive at your landing page → enter a test email → receive the first email → click the product link → arrive at the partner’s checkout

Total time: two to three hours.

Done means: Every outbound link in every email routes through your redirect layer. Every redirect resolves to the correct merchant page with your affiliate tracking tag intact. Analytics events fire on page view, email capture, email open, and link click. You have walked the full flow yourself — start to finish, on your phone, not just your laptop. If any step breaks, you know exactly where.

Day 5: Deal Memo + Partner Conversation.

Send the partner the demo. If you already have a verbal agreement, formalize it with a one-page deal memo covering:

Revenue split structure (Shape 1, 2, or 3 from Chapter 8)
List ownership
Test duration (fourteen days minimum)
Exit terms (30-day notice, infrastructure reverts to operator)

Get it signed or agreed in writing (email is fine). Then send the partner the single URL — the one link that replaces their current description/bio link.

Total time: one to two hours for the memo, variable for the conversation.

Done means: The deal memo is signed or agreed in writing. The partner has the single URL. Both parties understand the test duration, the revenue split, and the exit terms. Don’t overcomplicate this. An email thread with “yes, agreed” is a valid agreement. You’re not negotiating a Series A — you’re confirming that both sides understand the arrangement.

Days 6-7: Buffer.

Something will break. The email automation won’t fire correctly. The redirect will loop. The analytics event won’t register. The landing page will render badly on one specific Android phone.

Two buffer days are not optional. They are structural. Use them to test, fix, and re-test the entire flow from the subscriber’s perspective.

Common Day 6-7 fires: the email automation triggers twice (duplicate sends). The redirect loops because the destination URL has a trailing slash mismatch. The landing page loads in four seconds on mobile because an unoptimized image snuck in. The analytics script blocks the email capture form on Safari. The affiliate tracking tag gets stripped by the redirect layer.

Every one of these has happened to operators on their first deployment. Every one is a fifteen-minute fix if you find it during the buffer. Every one is a week of lost data if you find it after go-live.

Days 8-9: A/B Test Setup.

This is the step that turns your toll position from “I think this works” into “I can prove this works.” The A/B test is not optional. It’s the evidence you’ll use in the Day 15 partner check-in, and it’s the evidence that earns partner two.

Configure the A/B test:

Control: the partner’s existing direct-to-checkout link
Variant B: your landing page → email capture → redirect to checkout

The partner’s single URL now routes 50% of traffic to each path. Set up the tracking to compare: conversion rate (purchases), email captures (Variant B only), and total revenue per visitor.

In most cases, Variant B wins — not by a heroic margin, but by 15-30% in conversion rate, plus email captures that the Control doesn’t produce at all. The captures are the sleeper metric. Even if the conversion rates are identical, the email list Variant B builds is a compounding asset the Control can never produce.

Day 10: Go Live.

The partner’s audience starts flowing through the split test. Real visitors. Real data. Real experiment log entries.

Check the analytics once in the morning and once in the evening. Resist the urge to change anything for the first seventy-two hours — you need baseline data before you can optimize.

The first twenty-four hours deserve a closer watch. Confirm: captures are registering in your email provider. The welcome email fires within minutes of signup. The redirect links resolve correctly. The analytics dashboard shows page views that match the email provider’s capture count (if you’re seeing 50 page views and 0 captures, the form is broken). If everything checks out after twenty-four hours, pull back to the twice-daily cadence and let the data accumulate.

Done means: Real visitors are arriving. Real emails are being captured. Real welcome sequences are firing. The experiment log has its first entry: “Hypothesis: A pre-sell landing page will outperform a direct-to-checkout link. Status: running.” You have a live toll position. It’s imperfect. It’s operational. The experiment log will make it good.

What to Intentionally Skip

The sprint’s purpose is a working system, not a perfect one. Operators who try to build a masterpiece in ten days don’t ship in ten days.

Do not build advanced segmentation. You don’t have enough subscribers to segment yet. Do not create multiple landing pages. Prove the model on one page for one partner before you diversify. Do not automate your dashboard. The manual spreadsheet tells you which metrics actually matter — automate later, after you know what to watch. Do not add the Sideways Survey — that’s a month-two improvement. Do not build grouped offers or VIP tiers — that’s month four.

Every feature you skip now is a feature you can add later with the benefit of real data. Every feature you add now is time stolen from going live.

Day 10 is not the finish line. It is the starting line.

What “Done” Looks Like After Day 10

You have:

A live landing page capturing emails from your partner’s traffic
A six-email welcome sequence firing automatically for every new subscriber
A redirect layer tracking every outbound click
An A/B test running against the partner’s existing setup
An experiment log with its first entry: “Hypothesis: A pre-sell landing page will outperform a direct-to-checkout link. Metric: conversion rate + email capture rate. Status: running.”
A Day 15 partner check-in on the calendar
A Day 20 minimum viable metrics check scheduled

You do not have:

A perfect landing page (the experiment log will improve it)
Perfect email copy (the experiment log will improve it)
Revenue (probably — it takes five to fourteen days for the first purchases to flow through)
Confidence (that comes in week three, when the data starts showing a clear winner)

The sprint is the beginning, not the end. The experiment log starts on Day 10 and runs for months. Every week, you’ll test one thing, log the result, and make the position slightly better.

At Day 20, pull your minimum viable metrics. Captures: 5+ per day on a partner with 100K+ views. Capture rate: above 15%. Email 1 open rate: above 55%. Email 1 click-through rate: above 8%. First sale: within 14 days. If you’re hitting these numbers, the system works. Now optimize. If you’re below these numbers, diagnose: is the landing page underperforming (capture rate low)? Is the email sequence underperforming (click-through low)? Is the product wrong (clicks but no conversions)?

The diagnosis tells you what to fix. The experiment log tells you how to fix it. One experiment per week for the next 90 days.

The Checklist

Print this. Or screenshot it. Or stick it in Notion. This is the minimum viable deployment checklist.

Twelve items. Ten days. One toll position.

Most engineers look at this list and think it’s too simple. That’s the point. A SaaS MVP has hundreds of decisions: database schema, API architecture, authentication, payment processing, onboarding flows, notification systems. A toll position has twelve decisions. The simplicity is the leverage.

The difference between the operator who ships in ten days and the one who’s still polishing at day thirty isn’t skill. It’s the willingness to let the experiment log close the gap between “live” and “good.” Perfection is a form of procrastination with better branding.

Thirty to forty hours. Twelve checklist items. One working system that has the shape of an asset on day eleven. The next ninety days prove whether it behaves like one.

Everything after this is optimization. And optimization is what engineers were built to do.

The First 500 Subscribers

The hardest subscribers to acquire are the first 500. Not because the mechanics are complex — the sprint above proves they aren’t. But because operators overthink the page, underspend on testing, and abandon campaigns before the data matures. The psychology kills the attempt before the economics have a chance to work.

The Wright Brothers didn’t build a better airplane first. They built a wind tunnel to test wing shapes. The first 500 subscribers is your wind tunnel — you’re testing the economics, not building the business.

The first 500 subscribers can come from your partner’s traffic — that’s the standard path and the one most aligned with the model’s premise that you monetize existing demand, not create new demand. But if your partner’s audience is small or if you want to validate the economics faster, a small paid traffic test can accelerate the data. This is validation spend, not the operating model. Once the position proves itself on partner traffic, paid traffic becomes optional.

Traffic quality tiers. Not all traffic is created equal, and the tier you start with determines how fast you learn.

Tier 1: Sponsored newsletter drops. Cheapest per click ($0.40-1.20), fastest to deploy, lowest complexity. You’re placing a sponsored send in a newsletter that serves the same niche. The quality varies — some newsletter operators deliver engaged subscribers, others deliver low-intent clicks. Start here because the data arrives in days, not weeks. Even low-quality data teaches you what your landing page does under real conditions.

Tier 2: Social ads. Moderate cost ($1.00-3.00 per click), requires creative assets, benefits enormously from Tier 1 data. Once you know which headline converts on a sponsored newsletter audience, that headline becomes your ad copy. The Tier 1 data subsidizes the Tier 2 creative.

Tier 3: Intent-based ads. YouTube pre-roll, Google search ads, podcast sponsorships. Expensive per lead ($3.00-8.00 per click) but highest downstream value because the subscriber arrived with active intent. This is month-three traffic, not week-one traffic.

Start at Tier 1. Graduate to Tier 2 after you have conversion data. Consider Tier 3 only after the Lead Economics Formula from Chapter 12 shows positive margin at the higher cost.

The 14-day sprint budget. The math is mechanical. If sponsored newsletter drops cost $0.60 per click and your landing page converts at 35%, each subscriber costs $1.71. Five hundred subscribers at $1.71 = $857. At 100 clicks per day, that’s fifteen days of data.

Under a thousand dollars to prove the economics of a position that can produce revenue for years. Most operators spend more on a weekend conference.

The iteration rule: test until it works. Scale until it doesn’t. A profitable campaign is the most valuable asset an operator owns — more valuable than the landing page, more valuable than the email sequence, because the campaign is the thing that feeds both. Don’t stop a winner to try something new. Scale the winner. Test the new thing separately, with separate budget, on separate tracking. Two campaigns running simultaneously: one producing, one learning.

Ninety percent of the subscribers on your list won’t buy. That’s normal. That’s the architecture working as designed. The list economics work because the ones who do buy are worth multiples of the acquisition cost — the Lead Economics Formula proves this. And the 90% who don’t buy today are an asset: they receive content, they build trust, they recommend the resource to others, and some percentage of them buy later.

The first 500 subscribers changes something psychological, too. Below 500, the experiment log feels theoretical — small sample sizes, high variance, hard to draw conclusions. Above 500, the data stabilizes. A/B tests produce meaningful results. Revenue patterns emerge. The operator stops asking “will this work?” and starts asking “how do I make this work better?”

That shift — from existential doubt to operational optimization — is worth more than the $857. It’s the moment the toll position stops being an experiment and starts being a business.

Next Friday: Twelve mistakes that kill toll positions before they compound.