Claude Skills let a consultancy package a repeatable method, a review checklist, a document format, a domain-specific workflow, into something a client installs and runs themselves. Plugins extend this further by bundling Skills with the connectors and commands they depend on. For Australian consultancies whose value has historically been delivered by the hour, this is a genuine change in what can be sold, and it deserves a clearer-eyed assessment than the productisation enthusiasm it usually attracts.
What a Skill actually captures
A Skill is a structured set of instructions that Claude loads when a task matches its description. In consulting terms it captures method rather than judgement: the sequence of checks a good analyst runs, the format a deliverable takes, the questions worth asking about a particular class of problem. What it does not capture is the judgement to know when the method does not apply, which is the part clients are usually paying most for.
Good Skill candidates: repeatable review processes, standard document formats, checklist-driven analysis, onboarding workflows a junior would follow.
Poor Skill candidates: work whose value is situational judgement, stakeholder management, or anything where the right answer depends on knowing the client's politics.
Best candidates of all: the work you currently do at a discount because it is beneath your rate but clients keep asking for it.
The business model question underneath
Packaging a method as an installable Skill converts billable hours into a product, which sounds unambiguously good until you model the revenue. A consultancy billing $18,000 AUD for a recurring review engagement that a Skill could largely automate has to decide whether to sell the Skill for a fraction of that, keep the engagement and use the Skill internally to improve margin, or some combination. Firms that productise without answering this find they have competed with themselves.
The version that works for most Australian consultancies is the third option: use Skills internally to compress delivery time on the method-driven portion of an engagement, which improves margin on fixed-fee work, and package externally only the pieces that are genuinely commodity and act as a lead-in to higher-value work.
Distribution as a positioning asset
A publicly available Skill that solves a real problem for a specific Australian industry is a credibility artefact in a way a case study is not, because prospects can run it themselves before they ever speak to you. For a consultancy competing for attention against larger firms, a well-built Skill that a Sydney accounting practice or a Melbourne logistics operator installs and finds genuinely useful is a stronger introduction than any amount of content marketing.
This works best when the Skill solves a real but bounded problem and its limits are obvious in use. A Skill that handles the straightforward eighty percent and clearly signals when a situation needs a human is both more useful to the client and better positioning for the consultancy than one that overreaches and produces confident nonsense at the edges.
What building one actually takes
Less than most firms expect. A first Skill capturing a method a firm already has documented internally is typically a few days of work, mostly spent making implicit steps explicit and testing against real cases. The expensive part is not the authoring, it is the discovery that your documented method has gaps you have been covering with unwritten judgement, which is uncomfortable and also the most valuable output of the exercise.
Budget for iteration. A Skill tested only against the cases you had in mind will fail on the first genuinely messy client situation, and the difference between a Skill that gets used twice and one that gets used weekly is almost always the volume of real-world testing behind it.
Support is the part firms forget to price
A distributed Skill creates a support obligation that consultancies rarely budget for. Clients who install it will hit edge cases, ask why it produced a particular output, and expect an answer. If the Skill touches anything with compliance weight, an accounting treatment, a regulatory checklist, a contract review step, the questions get sharper and the stakes higher, and a wrong answer produced by your Skill on a client's system is your reputational problem regardless of the disclaimer.
Firms that distribute Skills successfully treat that support surface as part of the offer and price accordingly, whether that means a paid support tier, a named contact, or simply a conservative Skill that refers more situations to a human than strictly necessary. Firms that publish and walk away tend to get one round of goodwill and one round of complaints.
The honest recommendation
Not every Australian consultancy should productise. If your differentiation is genuinely relationship and judgement rather than method, Skills will compress your delivery cost without giving you a new product, and that is a perfectly good outcome to pursue on its own terms without the distraction of a distribution strategy.
If a meaningful share of your revenue comes from repeatable method work, though, the calculation is different, and the firms who move first in a given Australian vertical will define what clients expect the tooling to look like. That positional advantage is the real argument for building one sooner rather than waiting to see how the category settles.
