Escrow and holdbacks for licensing risk

Escrow and holdbacks for licensing risk are the buyer mechanisms that keep part of the purchase price within reach after close, so an inherited publisher audit demand is met from the seller proceeds rather than from the return. Inherited software licensing exposure is usually latent and unquantified in standard due diligence, and it tends to surface as an audit months after the deal completes. An escrow or a holdback turns a contractual promise into available cash, which is what matters when a true up demand actually lands.

Escrow and holdbacks for licensing risk, defined

An escrow is a sum carved out of the consideration and placed with a neutral third party, released to the seller only after a defined period or once defined conditions are met. A holdback is similar in effect but simpler in form, where the buyer retains part of the price directly and pays it over later if no claim has arisen. Both serve the same purpose for software risk: they ensure money is there to cover a loss the buyer can foresee but cannot yet quantify, instead of leaving the buyer to chase an unsecured indemnity claim against a seller who has already been paid in full and may have distributed the proceeds.

The reason these mechanisms matter so much in the software dimension is timing. A change of control can itself prompt a publisher to open a review, and audit clauses commonly allow the publisher to look back across several years of deployment. So the very event that completes the deal can trigger the demand, and the demand can arrive long after the seller has banked the money. An escrow sized and timed to the audit window keeps the seller funds in play for exactly as long as the exposure remains live.

How escrow and holdbacks differ from an indemnity

An indemnity is a promise to pay. An escrow or holdback is money already set aside to honour that promise. The distinction is the whole point. A buyer can hold a perfectly drafted indemnity and still recover nothing if the seller has dissolved, distributed the proceeds, or simply refuses to pay and forces the buyer into a costly dispute. The escrow removes that collection risk for the amount it holds. This is why, for a known or reasonably foreseeable licensing exposure, buyers pair the indemnity with secured funds rather than relying on the covenant alone. The mechanics of the underlying promise are covered in software licensing indemnities explained.

Sizing is where the commercial work happens. The escrow should reflect the quantified exposure, not a round number pulled from precedent. That means the cost to cure has to be modelled before the negotiation, including any settlement, back maintenance, additional licence fees, and associated costs, as set out in quantifying cost to cure for the deal model. An escrow set too low leaves the buyer exposed for the excess; one set too high is capital the seller will fight to release early. A defensible number, anchored to a quantified estimate, is far easier to hold in negotiation.

Sizing, timing, and release conditions

Three terms decide whether an escrow actually protects the buyer. The amount must match the quantified exposure. The duration must outlast the realistic audit window, which for the major publishers can run well beyond twelve months given how long a change of control review can take to surface. And the release conditions must be specific enough that the buyer can draw the funds when a demand arrives, not only after a final adjudication. A buyer who agrees that funds release only on a court judgment has recreated the weak indemnity trigger inside the escrow, because most licensing exposure settles through negotiation with the publisher rather than litigation.

Buyers should also watch the interaction with other protections. Where an exposure is large or genuinely uncertain, an escrow can sit alongside warranty and indemnity insurance, which is examined in warranty and indemnity insurance and software risk. The escrow covers the known, identified risk that an insurer will usually exclude, while the policy can address the unknown breach. The overall split of who carries what is then settled in the agreement, as covered in negotiating software risk allocation in the SPA.

Which publishers drive the need for an escrow

Not every line of the software estate justifies ringfenced cash. The case for an escrow is strongest where the inherited agreements come from publishers with both a history of post deal audits and pricing models that punish over deployment. The major post deal audit risks come from Oracle, SAP, Microsoft, IBM, and increasingly Broadcom following its acquisition of VMware, Salesforce, and ServiceNow, as of June 2026. Where the target runs material estates from those publishers, an escrow sized to the worst credible true up is a proportionate response. Where the estate is small or low risk, a holdback or a simple price adjustment may be enough, and an escrow only adds cost and friction.

The buyer should map the escrow to the specific publishers and metrics that carry the exposure, rather than apply a generic percentage of the consideration. A processor based metric, a virtualisation rule, or an indirect access question can each turn a modest deployment into a large demand, and the escrow has to be sized to that reality. The same deployment and entitlement work that quantifies the exposure also tells the buyer which agreements deserve secured cover and which do not, so the protection is concentrated where the real risk sits.

Escrow and holdbacks for licensing risk sit within software in deal valuation, alongside indemnities, warranties, and risk allocation in the agreement. The quantified findings that size these mechanisms come from software spend diligence. Engage your own counsel for legal interpretation of any escrow agreement or clause.