Learn Blitzscaling

Blitzscaling is the strategy of spending heavily and moving fast to seize a whole category before anyone else can — and then defending it. BlitzLab grades how well companies pull that off, on a single 100-point ruler. This guide explains the ruler, how to read a score, and what a score cannot tell you.

What blitzscaling actually is

Most companies grow step by step: win a customer, earn a profit, fund the next step. A blitzscaler does the opposite. It raises a large amount of capital up front, accepts losses for a period, and races to become the default in its category while the market is still forming. The bet is that being first at scale is worth more than being efficient early — because once customers, suppliers, and partners standardize on one winner, switching becomes expensive for everyone.

That is why BlitzLab calls itself the credit rater for blitzscaling. A credit rater does not tell you whether a bond will make you money. It judges one thing: how strong the borrower's position is. BlitzLab does the same for strategy. It judges how strong a company's blitzscaling position is — its speed, its moat, and whether it is genuinely replacing an old order — and it leaves everything else to the reader.

Two gates come before any score

Not every fast-growing company gets scored. Two gates must both pass first, and a company that fails either one is not scored at all. Instead, BlitzLab publishes a judgment file explaining exactly why it failed, so the reasoning is public even when there is no number.

Gate 1 asks whether the business is real. The evidence is product-market fit: customers actually using the product, or, for a mature company, sustained growth or a second growth curve. A company with no real product revenue cannot pass this gate, no matter how exciting its story is.

Gate 2 asks whether there is room. BlitzLab estimates the market's ceiling over ten years and requires it to be at least ten times the company's current revenue. A company can be excellent and still fail here, if the market it serves is simply too small to support a blitz at its current size.

Both gates matter because blitzscaling only makes sense when a real business meets a large prize. Speed without a real business burns capital. A real business without room has nothing to seize.

The 100 points, in plain language

A company that passes both gates is judged on 100 points, split into four sections. The split itself tells you what BlitzLab thinks matters most.

Section A is Speed, worth 25 points. It covers revenue growth, how fast the company raises capital, hiring and productivity, and whether its market share is moving. Speed is the blitz itself: how fast the company is taking ground.

Section B is Moat, worth 60 points — the largest section by far. It covers network effects, whether distribution scales, market size, and gross margin. The reason it dominates the score is simple: speed without a moat is just a race anyone can join. The moat is what makes the lead last after the blitz ends.

Section C is Creative destruction, worth 15 points. It asks what old order is being replaced, whether customer behavior is changing irreversibly, and whether the cost structure of the industry is being disrupted. A company can grow fast inside an existing order without destroying anything. This section separates the two.

Section D is BlitzFailing demerits, worth up to minus 30 points. It subtracts for debt and interest burden, revenue quality problems, governance risks, and regulatory or social costs. A company can score well on speed and moat and still lose points here if the foundations are fragile.

The total is the sum: A plus B plus C, minus D. Grades follow the total: S at 90 and above, A at 80 and above, B from 60 to 79, C from 40 to 59, and D below 40.

How to read a scorecard

Every scored company has a public scorecard. Read it in this order.

First, the total and the grade. The total is the headline, but it is not the whole story. Two companies can share a total with very different shapes — one fast with a thin moat, another slower with a deep one — and the shape tells you more about the risk than the number alone.

Second, the section breakdown. Compare the four sections against their maximums. A high B with a low A usually means a strong position that is no longer expanding quickly. A high A with a low B usually means a fast company whose lead is not yet defended. A large D deduction deserves its own look: read which demerit items drove it, because debt, governance, and revenue quality are different kinds of risk.

Third, the 65 line. BlitzLab draws a success line at 65, set by a blind backtest of twenty companies — the first draft's line of 80 was lowered after measurement, and the model has been frozen at version 1.0 since. Crossing 65 means the company's blitzscaling position clears the bar the backtest set. It does not mean the company will succeed, and it does not mean anything about its share price. It means the strategy position, as judged, is strong.

Fourth, the flags and the history. A total demerit of minus 15 or more flags a company for BlitzFailing watch — a warning that the foundations need attention, not a verdict on the business. And every scorecard shows its history: scores are revisited after each quarterly earnings report, and past scores are never silently rewritten. If a number changed, the history shows when, and any arithmetic correction carries a public note.

Fifth, the confidence labels. Every figure on a scorecard is labeled confirmed, estimated, press-reported, or unconfirmed. Confirmed means it comes from a filing. Estimated means BlitzLab calculated it and says so. Press-reported means it comes from reporting, not a filing. Unconfirmed means it could not be verified and is not used as evidence. If a score rests heavily on estimated or press-reported figures, read it with that in mind.

Unlisted companies work slightly differently. Because they file less, they are scored as ranges — for example 51 to 68 — with lower confidence shown separately. The range is not a hedge on the score; it is an honest statement that the evidence supports a band, not a point.

What a score cannot tell you

This is the most important section of this guide, and BlitzLab publishes it openly. The team pre-registered three hypotheses testing whether the score predicts stock returns: crossing the 65 line, profitability turns, and score slope. All three were rejected. The published conclusion is plain: the score measures business quality; entry price governs returns.

That means a high score is not a reason to buy anything, and a low score is not a reason to sell anything. BlitzLab publishes no buy, sell, or hold recommendations, no price targets, and no implied returns. Expectations and concerns are stated for every company, so the reader can see both sides of each judgment — but the decision about any security is always the reader's own, made with information beyond this site.

The model itself has limits worth knowing. It is frozen at version 1.0, so weights and thresholds do not drift with fashion — but freezing also means the model cannot quietly adapt when the world changes. It judges strategy from public evidence, so a company that discloses little can only be judged coarsely, as a range. And it judges the business, not the stock: a brilliantly run company can be expensively priced, and a poorly scored company can be cheap. The scorecard never claims otherwise.

Where to go next

Start with the company index, which lists every scored company with its score and grade. Open any scorecard and read the sections before the total. To see two companies side by side, use the Compare tool, which puts their breakdowns next to each other. To see how the same ruler applies to a fund, visit ARK Lens, which grades the holdings of ARK's published funds and shows where their scores overlap. To see the reasoning behind the model — including the validation work that rejected its own hypotheses — read the Research index. And for the formal rules, the Methodology page states the gates, the weights, and the evidence standards in full.

BlitzLab scores businesses against a scholar's standard — not investment advice. No price targets, no buy/sell recommendations.