A repeatable structure for screening a company before you commit real time to it — what to check, in what order, and when to stop.
Early-stage screening has a specific failure mode: spending two hours on a company that a ten-minute check would have ruled out. The fix isn't researching faster — it's researching in an order that lets you stop early when something disqualifies the company.
Before anything else, verify the company exists as described: founding date, headquarters, current status (active, acquired, shut down). This sounds trivial, but it's the fastest disqualifier — stale directory listings and outdated press are common enough that skipping this step wastes time downstream.
Check who's actually running the company and whether their background matches the pitch. Leadership red flags — undisclosed past ventures, unexplained gaps, a mismatch between claimed and actual experience — are cheap to check and expensive to discover later. This is also where you catch the most common early-stage issue: a technical founder running a go-to-market-heavy business, or vice versa.
Look at the funding history for a coherent story: does the round size and pacing match the stage they claim to be at? A company raising a large round on thin traction, or one that's been "about to close" a round for an unusually long time, is worth a second look before you go further.
Map who else is solving this problem and how the company differentiates. The goal isn't to build an exhaustive competitor list — it's to check whether the founder's stated differentiation actually holds up against what's publicly available about competitors.
You don't need a rigorous TAM model at the screening stage. You need a sanity check: is the addressable market big enough to matter, and is the company's own sizing claim in a believable range compared to public benchmarks in the space.
The step people skip: write a one-paragraph verdict immediately, while the context is fresh. "Pass — thin team, no differentiation" is more useful three weeks later than a folder of open tabs you no longer remember the reasoning behind.
Steps 1–5 are exactly what a synthesized report should hand you up front — founding facts, leadership background, funding history, competitive context, and market sizing, together, instead of six separate lookups. That's the gap a tool like Metis is built to close: turning this checklist into a single report you read once, so your actual analysis time goes into step 6 — the judgment call — instead of the data-gathering that precedes it.