Unlocking Cross-Portfolio Synergies

Every holding company deck promises the same thing: a portfolio worth more than the sum of its parts. The word gets used so often it has stopped meaning much. The honest version is narrower and worth being precise about, because the difference between cross-portfolio gains that compound and ones that exist only on a slide is the difference between a healthy group and a rollup that quietly bleeds out.

What actually compounds across a portfolio

A few things genuinely transfer between commerce companies that share a sector. The clearest is shared infrastructure: plumbing every brand needs but none should be rebuilding in isolation. Constellation Software, often held up as the template, runs hundreds of vertical software businesses on a model where head office handles capital allocation, finance, tax, and legal, while operating groups disseminate playbooks and unit heads run their own shops. The lesson is not "centralise everything." It is "centralise what has real economies of scale, and leave the rest alone."

For a digital commerce group, the list of things worth running once at the group level tends to be short and specific:

• Payments orchestration and fraud tooling, where volume genuinely lowers cost and improves approval rates
• A shared data warehouse and BI layer, so each brand can answer questions about itself without rebuilding the stack
• Group-level finance, accounting, and tax, especially across European jurisdictions
• Vendor contracts on cloud, observability, and core SaaS, where pooled spend earns real discounts
• A talent pool for senior specialists (performance marketing, lifecycle, supply chain) that no single small brand could afford full time

The other thing that compounds is knowledge. A founder who has solved a margin problem in one category can save another operator six months of guessing. That transfer happens through people, not memos. It works when operators trust each other and have a forum to compare notes. It does not work when "best practices" are pushed down from a corporate function that has never run the business.

What doesn't (and why operators keep trying)

Most cross-portfolio promises fail in the same predictable ways. Forced cross-selling between brands whose customers do not actually overlap produces noise, not revenue. Rebranding acquired companies onto a single platform usually destroys the trust that made them worth buying. Centralised marketing teams tend to flatten what was distinct about each brand into a house style nobody asked for.

The Thrasio story is the cautionary tale operators cite, and rightly so. The company acquired roughly 200 Amazon brands in a couple of years, scaled headcount past 1,600, and tried to integrate at a pace the businesses could not absorb. When the capital environment turned, the integration debt was larger than the operating gains. The thesis was not wrong in principle; the execution assumed integration was free. It is not.

Tiny, the holding group run by Andrew Wilkinson, sits at the opposite extreme: hire good operators, pay them well, leave them alone. That model under-captures some real shared-services gains, but rarely destroys value, which is more than most aggregators can say. The truth for an operator-led group sits between the two poles, and getting the line right is most of the job.

How we think about it

Our default at Rocketlink is to assume integration is expensive and the burden of proof sits with whoever proposes it. A shared service has to clear a bar: cheaper or better at the group level than at the brand level, and the brand has to want it. Cross-sell runs only where customer overlap is real and measurable. Playbooks are offered, not imposed. Brand, pricing, and product decisions stay with the operators who know their category.

In practice this is mundane: books that close faster, a data layer every brand can query, contracts negotiated once instead of seven times, and operators who pick up the phone for each other. The discipline is saying no to integrations that sound good in a deck and cost more than they return. That is the version of cross-portfolio gains we think compounds, and the only version we are willing to claim.