Pricing is the commercialization layer. Fix the scoping architecture before you chase another client. AI does the grunt work, humans hold the judgment.

Shirtcliff's route in began at university, where a geology and zoology degree convinced her that life on an oil rig was not for her. She cross-credited to a master's in marketing and commerce, spent much of her career client side, then started and ran her own content agency before moving into software.

Her first software business was Sonar, whose product Traffic became Traffic Live, one of the early platforms to move from server-based delivery, where she used to ship Mac Minis to offices, to software as a service when it launched in 2007. It was a project management and workflow tool used by agencies across WPP and other networks. She sold it to Deltek in 2015 with, by her account, over 1,000 agencies, three offices in London, Sydney and New York, and nearly 80 people.

When she exited Traffic Live it became clear the workflow and ERP tools did not do scoping and pricing. Before a dedicated tool existed, people scoped in Excel, a staff plan built on hours, roles and rates plus a statement of work in words. Everyone does it differently, so there is no methodology, and she thinks there should be. The real problem is the intersection of sales, who want the win, creative, who want award-winning work, and finance and operations, who protect delivery and margins.

It breaks in how you bring the parts together and put structure, or architecture, around them. Without it you get scope creep, sales overselling, creative over-delivering and operations overwhelmed by runaway margins. When work comes out of sales it is normally undersold, because salespeople want the sale, and creative businesses just are not good at this discipline.

On the figures Shirtcliff cites, an agency returns roughly 150K per head while a technology business like Google returns around 2 million and Apple around 3 million. The difference is structure and alignment, and the gap only widens with AI. If every head is only putting on 150K, she says, that is not a sustainable business, and scope creep on top makes it worse.

By her count, 60 to 70 per cent of agencies still record and bill for time, which she believes AI should have ended. People move to fixed fees but still build them from timers. Time and materials persisted, and margin became a problem, because changing it sat in the too hard box.

WPP moved their model about 12 months ago and now talk about outputs: charging for a deliverable, packaging up what you deliver, who does it and how, and putting a methodology behind it for a fee. It is not broken down by roles, hours and rates. The route out is deliverables, outputs and eventually outcomes, and charging clients for the AI and technology agencies already buy rather than absorbing it.

The tools are not as connected as they should be because each asks different questions and holds different data: CRM handles who to follow up and when, scoping handles what the work is, workflow handles who does it and when, and ERP handles billing and recording. Scoping and pricing are their own discipline, which is why Shirtcliff describes SCOPE Better as a configure price quote system, distinct from a CRM or a workflow tool. In an AI world she expects agents to run across all of it, and says platforms must be ready for that.

Shirtcliff expects agents to assemble all pricing within a three-year horizon, but insists humans stay the commercialization layer, reading the client and the human conditions that shape a price. Today AI does the grunt work and people check, tweak and move on, and pricing will be the same. The new gain is the efficiency loop: learning and comparison across scopes that never existed before.

Having built tools for people who dislike software, she has learnt it has to look good, be drag and drop, and absorb feedback quickly, while flexing to the many different ways creative businesses work. Agencies, she notes, are creative inside their business as well as outside it, which is the challenge she has embraced.

The instinct is to solve every problem by finding more clients, when it is harder to win a new client than keep an old one. Shirtcliff argues the commercialization of the work matters more than the next win, yet agencies chase recognisable CMOs while cashflow and lumpy billing quietly stall growth. The cobbler's shoes problem is winning more business while letting it seep out the back end.

Some agencies differentiate brilliantly and many do not. She sees a proliferation of transformation agencies and a plethora of vanilla names, and calls it a copycat industry. The ones who stand out have thought about their architecture and price for their differentiation, leading with strategy and methodology rather than a list of deliverables.

She points to Canva, brilliantly easy, generous and slick, and Pinterest, which nearly became legacy and had a renaissance by connecting to AI early. A manifesting seminar stuck with her too: the idea that what you think about comes into your consciousness, which she believes matters more as human connection and flow gain value in an AI world.

Her advice for anyone starting out is to be hungry and learn lots, but above all to bring grit and determination. Say you will do something and stick at it. As she puts it, mama didn't raise a quitter. The industry has lost longevity as people move to the tech platforms, but she still calls it challenging and fun.