New business launches go fatally late when the sequence is wrong. The common failure modes are "start building product immediately" or "polish strategy for a year." Here's the four-phase order we actually use in our engagements, written down as a method.
Phase 1: Strategy — decide who, what, and why
The first risk of a new business isn't technology or funding. It's running off without clarity on whose problem you're solving.
Three things to do in this phase:
① Define the target segment (industry, size, function, role).
② Identify one strong inconvenience that customer has.
③ Articulate why "now, you, are solving this."
Don't try to "finish strategy" with a thick deck. Compress to one page that explains it. If it can't compress, it can't be validated either. In our engagements, the first 2–3 weeks are typically spent rewriting that single A4 page 5–10 times.
Phase 2: PoC — hit a working version, fast
With strategy in place, build working software. The point isn't "the perfect product" — it's the smallest thing that gets a real reaction from someone.
Using an AI-native development stack, PoCs come together in days or weeks. Showing customers something they can touch surfaces the corrections to strategy far faster than continuing the slide discussion.
Don't end the PoC at "technical validation." Production-readiness conditions (accuracy, operations, cost) should be in scope too. See Three conditions for taking PoC to production for the full unpack.
Phase 3: Early sales — go deep with the first 3–5 customers
Once the PoC is up, start selling immediately. The biggest mistake at this stage is to spray it through marketing. Why: strategy is still hypothesis, and so is product.
Pick 3–5 target customers, go deep, and run hypothesis validation through real sales conversations. Reflect what you hear into next week's product. That speed is how you reach the right answer before competition does.
The CEO should be in the room here. Decisions made on the spot, words heard directly — that's how 0→1 stays accurate.
Phase 4: Growth — formalize, then scale
When the pattern starts showing in the first customers, scale. Three lanes:
① Formalize sales operations (per-stage CVR targets and the operating model).
② Design customer success (retention and expansion after winning).
③ Select marketing channels (branded search, content, partner-led).
Critical: don't scale before the pattern exists. If you advertise without a pattern, lead cost goes down but conversion doesn't follow — and CAC (customer acquisition cost) deteriorates.
Wrap-up: stick to the sequence, and 0→1 doesn't run late
Strategy → PoC → early sales → growth looks simple, but it gets violated in the field constantly. The failure patterns are predictable: "skipping PoC and running ads," "skipping early sales and relying on marketing," "skipping formalization and hiring people."
PlusDivide carries these four phases in one team. Strategy, PoC build, early sales, growth design — without switching vendors. That's our defining strength.
FAQ
Frequently Asked Questions
What's the first thing to do when launching a 0→1 business?
Compress the strategy — who, what problem, and why your company — into a single A4 page. Bigger than technology risk or capital risk is the risk of running off without clarity on whose problem you're solving. Strategy that can't compress can't be validated, and you'll lose months to course correction later.
When should you build the PoC?
Immediately after the strategy fits on one page. Not a perfect product — the smallest thing (mockup is fine) that gets a real reaction from someone. With an AI-native development stack, PoCs come together in days to weeks. Watching customers react to working software exposes strategy gaps far faster than continuing to argue over slides.
How many customers should you focus on in early sales?
Three to five, going deep. The single biggest mistake at the 0→1 stage is spraying the offer through marketing. Strategy is still hypothesis, product is still draft — large lead volume buries the learning. A small set of customers in deep dialogue, with what you hear reflected into next week's product, is what wins.
Should the CEO be in early sales personally?
Strongly recommended. Three reasons: (1) you get customer language firsthand, (2) you can decide on the spot (spec changes, pricing, contract terms), and (3) the CEO's conviction reads in the room. Founders who delegate sales too early slow the strategy correction loop and lose direct contact with the market.
How do you know when to scale a new business?
When the pattern exists. Three checks: (1) Sales operations are reproducible (per-stage CVR), (2) Customer-success design (retention, expansion) is visible, (3) An acquisition channel is identified. If you scale before the pattern, lead cost drops but conversions don't follow, and CAC blows up.