IP Strategy for Deep Tech Founders · Part 3 of 3

What investors read in a portfolio, and what they ignore

Counting filings is the fastest way to look unserious to someone who has run this diligence before.

Claims

what a technical diligence reader opens first

Provenance

what makes a search believable

The count is not the asset

A slide reading 'fourteen patents pending' tells a careful reader almost nothing, and tells a careless one something false. What gets read is a small number of independent claims and whether they cover the thing the company actually sells.

A portfolio of two well-scoped families beats eleven speculative ones in every diligence conversation we have watched.

Show your work

Where a search was run, what it found, and what was decided in light of it is more persuasive than any assertion about strength. It is also the part most companies cannot produce, because nobody kept it.

Keeping it is a process choice made early, and it costs nothing at the time.

Key insight

In diligence, a documented decision to proceed despite a reference reads better than a portfolio with no references in sight.

What to do next

  1. 01Map each claim family to a product line in one page.
  2. 02Keep the search record with the filing record, not in someone's inbox.
  3. 03Retire families that no longer map to anything you sell.

Daniel Gould

General Partner, Castle Fund

Invests in deep-tech companies at seed and Series A, and has sat on both sides of the diligence table when a portfolio is the asset being bought.

ZeroToIP

Editorial

ZeroToIP builds one record that runs from a founder's first description of an invention to counsel's filing. This series is general information, not legal advice.