Continuation patents: how to build a wider moat from a single filing

A continuation patent uses the same specification and priority date as its parent but presents different claims. The mechanism is simple. The strategy behind it is where the value lives.

Rock Innovation filed a parent patent in November 2016 with thirteen independent claim elements, including a flagging mechanism for exception handling. From that single specification, we built two continuation patents: one with ten elements adding criteria-based retrieval, and one with seven elements and no flagging requirement at all. All three share the same priority date. All three are in force.

The result is a family with different limitation sets that can be evaluated separately against a documented system.

Why you start narrow and go broad

Applicants may seek broad claims at the outset, but broader language can raise additional prior-art and support questions during examination. The appropriate strategy depends on the disclosure, the art, and counsel's judgment.

The '073 parent patent has thirteen limitations in the independent claim discussed here, including mechanisms for flagging discrepancies, replacing data, and recording corrections. The public prosecution history, rather than speculation about examiner preference, shows how those claims reached issuance.

With the parent on file, we had a granted patent and an established specification. The first continuation, the '029, trimmed to ten elements and added criteria-based retrieval. The specification already supported this claim scope; we just had not emphasized it in the parent. The second continuation, the '711, went to seven elements. No flagging. No mandatory device criteria. The broadest possible expression of the reconciliation architecture that the specification supports.

Each continuation was examined and issued. Issuance of the parent did not automatically validate the continuation claims; each has its own prosecution record.

What the progression looks like

The '073 includes limitations concerning discrepancies, flags, correction, and documentation. Financial settlement, compliance, or audit systems may offer comparison points, but a chart is persuasive only to the extent each element is supported by reliable evidence.

The '029 catches systems that use filtered queries against blockchain data sources. It does not require flagging. It requires that the system retrieve data based on specified criteria. A DeFi protocol that monitors positions below a certain health factor, or a supply chain system that pulls sensor readings from a specific region, or a bridge that watches for events on specific contract addresses: all of these perform criteria-based retrieval. Ten elements. Broader than the '073. Still specific enough to hold up.

The '711 uses seven limitations and does not include the same flagging or device-criteria language. Reading, processing, and writing blockchain data are only part of the inquiry; the remaining limitations and claim construction still matter.

The licensing advantage

Different claim sets provide separate avenues for analysis in a licensing conversation. The absence of flagging or criteria-based retrieval may distinguish one claim while leaving different questions under another; it does not mean another patent automatically applies.

The shared priority date means all three patents benefit from the same thin prior art field. November 2016 was early. The blockchain infrastructure that would later fill the prior art pool did not exist yet. Each continuation inherited that timing advantage from the parent.

Counsel evaluating the family should review written-description support, priority entitlement, amendments, claim construction, and each prosecution history. This article does not offer an opinion on those legal questions.

That is how you build a moat from a single filing. Start specific. Go broad. Keep the priority date. Details are at rockinnovationip.com.