The network behind the portfolio

A strategy for your external IP network is the deliberate design of which external patent firms, or outside counsel, build which parts of your patent portfolio, under what kind of relationship, and why. Most IP departments do not have one. They have a list of firms, a set of rates, and a habit of sending work to whoever handled the last case. The result is a network that treats a foundational invention and a routine continuation the same way, and quietly caps the quality of the entire IP strategy.

In our recent white paper, The CIPO advantage: what great looks like in 2026, we argued that leading IP organisations treat intellectual property not as a legal right but as a means to control business choke points: the few places where speed, margin, and bargaining power concentrate. They deliberately mix IP value models. They coordinate patents, trade secrets, data rights, and contracts to secure the points that matter.

Now follow that logic one step further. If a handful of patent families carry a disproportionate share of business value, then the question of who drafts them is a strategic question. A choke-point patent drafted by whichever firm happened to have capacity that week is a strategic decision made by accident.

Yet that is how most networks operate. Firms accumulate over decades: inherited from acquisitions, added for a jurisdiction, kept out of loyalty. Every firm is engaged the same way, with the same instructions, the same review process, and the same annual rate discussion. Nobody can say which relationships are strategic and which are transactional, because the network has never been designed. It has only grown.

This matters more than it looks. External firms typically execute the majority of a company's patent work: drafting, prosecution, filings across dozens of jurisdictions. The strategy may be set in-house, but it is realized, or not, in the claims an external attorney writes. A brilliant portfolio strategy executed by an undifferentiated network is a brilliant strategy on paper only.

The most effective IP functions have drawn the obvious conclusion: they handle their external network in the same way that they manage the portfolio itself; segmented by strategic value, governed differently by segment, and actively developed and pruned over time. In practice, this rests on four principles.

The starting point is a make-buy logic that mirrors the portfolio strategy, case by case.

Benchmark companies increasingly apply a tiered approach. Cases of high strategic importance are handled with internal resources as far as possible, not only for quality control but to build and retain the internal expertise that strategic work depends on. Cases of lower importance are outsourced end-to-end through a streamlined, standardized process, with minimal internal involvement and minimal variation from one attorney to the next.

Observe the change. The conventional approach is based on activity: "we outsource drafting, we keep prosecution." The modern approach is based on case importance: the same activity is handled entirely differently depending on what the case is worth to the business. That is the portfolio strategy, mirrored in the sourcing decision.

This has an uncomfortable implication for many departments: it requires knowing, case by case, which parts of the portfolio are strategically critical. Departments that cannot make that call cannot differentiate their sourcing. That is often the first sign that the portfolio strategy itself needs sharpening.

Once the decision has been made regarding whether to make or buy, the external network can then be designed in accordance with that decision. This involves organization design just as much as it does procurement. As we argued in Organizing for business-driven IP management, a modern IP operating model is more than lines and boxes on an internal org chart; the external network sits in the same layer as the internal structure, the organization and its interfaces, and it deserves the same design attention. In practice, leading IP functions segment their firms into tiers, each with its own selection logic, price logic, and depth of relationship. The exact segmentation model can vary, but typically includes some variation of the following:

Tier 1: Strategic Partners. A small number of firms entrusted with high-complexity, foundational inventions, where the goal is legal bulletproofing and cost is secondary. Typically top-tier boutiques with deep technical alignment to the company's core R&D. These firms should know the product roadmap, understand which families anchor which business plays, and be treated as co-creators of the portfolio, not as contractors executing instructions.

Tier 2: Volume Partners. Firms handling the high-volume core of the portfolio, where the goal is standardization, speed, and cost efficiency at consistently good quality. Often firms with a unified technology stack and global reach, able to run cross-border filing and prosecution with low-touch efficiency. Volume concentrated on few firms buys leverage, priority, and predictable pricing. It also buys flexibility: a committed Volume Partner is the shock absorber that lets a lean internal organization handle workload peaks without panic hiring.

Tier 3: Specialists. Small technical boutiques engaged for frontier technologies such as quantum or bio-convergence, where generalist firms lack depth. Used selectively, at a justified premium, for the cases that demand them.

Most companies review their firms the way they review invoices: backward-looking, cost-focused, and only when something goes wrong. Leading IP functions run structured relationship management instead, and three practices stand out.

• Named ownership. Every strategic relationship has a designated relationship owner on the client side: a single person accountable for the health of the partnership, distinct from the sourcing role that owns the commercial negotiation. Separating the two matters. It keeps the relationship conversation strategic and the price conversation honest.
• Structured business reviews. Strategic Partners meet with the client on a fixed cadence, typically semi-annually, for a two-way, future-focused session. Performance data is on the table, but so is the firm’s view of how the client could make it easier to deliver good work, and a joint look at the volume pipeline for the coming year.
• Performance measured on more than cost and deadlines. Benchmark companies score their key firms across five dimensions: quality of work product, commercial performance, speed and delivery, service and communication, and, tellingly, strategic contribution: the firm's proactivity in offering unsolicited advice and improving joint ways of working. That fifth dimension is the clearest marker that a company has moved beyond procurement thinking. Some go further on quality, running blind reviews in which senior experts assess anonymized work samples against a defined quality standard, separating the reputation of a firm from the evidence of its output.

Patent portfolios are pruned. Underperforming assets are abandoned, and budgets are redirected to what matters. Almost no IP department applies the same discipline to its firms. Relationships begin with scrutiny and then simply persist; the network's exit door is unmarked and rarely used. Leading functions treat every firm relationship as a lifecycle with four stages, each managed with intent and purpose.

• Entry is earned. New firms are qualified centrally, against explicit criteria such as technical depth, jurisdictional coverage, language capability, and commercial terms. Increasingly, qualification ends not with a signature but with a pilot: a controlled ramp-up on live cases before the firm earns full volume.
• Onboarding is an investment, scaled by tier. A new Strategic Partner gets a high-touch start, including joint workshops to transfer technical context and align on strategy. A new volume firm gets something leaner and just as deliberate: systems training, process alignment, and unambiguous performance expectations from day one. In both cases the message is the same: this is how we work, and we will help you succeed at it.
  Development runs through the review cycle. The scorecards and business reviews are not only feedback mechanisms; they are the evidence on which a firm's place in the network is periodically re-confirmed rather than assumed.
• Exit is planned, not suffered. The true reason why firms are seldom off-boarded is that no one wants to transfer an ongoing file, so indefinite tolerance of underperformance becomes the less risky option. In order to eliminate this excuse, leading IP functions introduce a structured transition procedure, shifting the active cases to other firms over a specific period of a few months. In this manner, off-boarding ceases to be a disruption and turns into a routine.

None of this should be taken as a justification for churn. The strategic partnerships in question should be long-term and stability should be seen as one of their benefits. On the contrary, it is about intent: each firm in the network is there as a result of a decision that is continually reviewed, and all parties on both sides are aware of this.

Two developments are raising the ceiling on what a well-designed network can deliver.

The first is joint planning. Companies with a clear view of their R&D pipeline are turning it into rolling demand forecasts by technology area and jurisdiction, and sharing that forward visibility with their Strategic and Volume Partners. A company that can see next year's workload can staff for it, commit to it, and price it. Forecasting turns the client from an unpredictable source of urgent work into a client worth investing in.

The second is AI, on both sides of the relationship. As we explored recently in The IP Catalyst and AI-powered portfolio engineering, leading IP functions are weaving AI into their own operations in a disciplined, use-case-driven way. The frontier is doing the same across the network: actively enabling firms to use AI tools safely on the client's work, and agreeing up front how the resulting efficiency gains are measured and shared. Companies that leave this unaddressed will pay yesterday's prices for tomorrow's productivity. Companies that structure it will not.

Both practices depend on the foundations above. You cannot plan jointly with a network you have not segmented, and you cannot share AI gains with firms you only meet at invoice time.