Establish what is held where
A clear picture of which entity owns which asset, which is frequently not the picture anybody expected.
Commercial and corporate
Structure is invisible until something goes wrong, and then it decides how much of the business is exposed. Most companies do not choose a structure. They accumulate one, an entity at a time, and end up with value sitting in the company carrying the most risk. Fixing that is a transaction rather than an administrative task, which is why the timing matters.
The usual position is a single company doing everything. It owns the brand, holds the customer contracts, employs the staff, owns the software and takes every commercial risk in the business. Nothing about that is wrong while things go well. It simply means anything that reaches the trading business reaches everything the business is worth at the same time.
The opposite failure is a structure that grew by accident. An entity created for a market, another for an idea that was abandoned, one that exists because a bank asked for it, several dormant, and nothing written between any of them. Nobody in the group can say with confidence which company owns which asset, which is a problem the moment anybody outside needs to be told.
Then there are rights sitting in the wrong place. Registrations in a founder's personal name, a trade mark owned by a company that stopped trading, software licensed to an entity nobody uses, and contracts signed by whichever company happened to be on the letterhead that year. Each is fixable in isolation and expensive to fix all at once.
The costly version is a reorganisation attempted at the worst possible moment. Moving assets during a fundraising or a sale means asking for consents, triggering change of control provisions in contracts nobody has reread, and answering questions from the other side about why any of it is being done now. Structuring done calmly is administration. Structuring done under a deadline set by somebody else is leverage handed away.
The first question is what you are separating and why. Protecting the valuable assets from trading risk, isolating an investment, keeping territories apart, preparing part of the business for sale and satisfying a lender are different objectives, and a structure built well for one of them can obstruct another. A group assembled without an answer to that question tends to produce cost and administration without producing protection.
Whether the structure is documented decides whether it does anything. Group companies are separate legal persons. If one uses another's brand, software, staff or customer relationships, there needs to be an agreement saying so and describing the terms. Where that use is undocumented, the holding company's ownership is asserted rather than demonstrated, and a buyer, a lender or an opponent will treat it accordingly. Intragroup arrangements are the part everybody skips and the part that is looked at first.
Moving an asset between entities is a transaction, not a bookkeeping entry. It may require consents, it may trigger change of control clauses in customer and supplier agreements, it affects arrangements with employees, and it has tax consequences that need to be worked out alongside by somebody advising on tax. The sequence in which the steps happen matters and cannot be reconstructed afterwards, which is why this work goes badly when it is done quickly.
Finally there is what the outside world can see. Who holds the shares, who the directors are, who exercises control, and what has actually been filed. Buyers, lenders and banks look at that before they look at anything you tell them, and inconsistencies between the register and the reality raise questions out of proportion to their importance. Incorporating in another country because the structure sounds efficient produces the same problem in a less convenient place, because substance tends to follow where the people and the decisions actually are.
A clear picture of which entity owns which asset, which is frequently not the picture anybody expected.
Valuable assets separated from trading risk, in a shape that matches what you are trying to protect.
The licences and arrangements between group companies written down, so ownership can be shown rather than asserted.
Consents, change of control provisions and sequencing handled so a transfer does not create new problems.
Subsidiaries and branches set up for a reason, with the ongoing obligations understood before they are created.
A structure and a set of records that survive being examined by somebody looking for reasons to pay less.
A small trading company with one product and one market should not build a group. Extra entities bring filings, accounts, bank arrangements and administration, and a holding company with nothing properly documented beneath it provides cost without protection. Structuring earns its keep when there is genuinely something to separate, and until then the honest advice is to keep the registers clean and wait.
The same applies to overseas entities set up because a market seems promising. A company in another country is an ongoing obligation, not a marketing decision, and it is usually possible to trade into a market for some time before one is needed. The right moment is when local requirements, customers or people make it necessary, rather than when it sounds established.
Commercial and corporate
Positions harden the moment the other side takes advice, and the quiet routes stop being available once a demand has gone out. While nothing has been sent, everything is still open.