Repositioning Propeller: how to talk about AI without sounding like software

Propeller is a finance platform built for the AI era. It gives venture-backed and private equity-backed founders the finance team they'd otherwise have to build themselves, from day-to-day accounting to CFO-level strategy, and pairs those people with PropellerOS, its built-in intelligence layer. Since 2008, it's worked with more than 1,400 companies, including 24 unicorns.

The AI was already there. What Propeller needed was a brand confident enough to talk about it, without losing sight of the people its clients rely on.

But Propeller's clients don't stay for the technology alone. They stay for the CFOs and controllers who sit in their board meetings and help them through the hard decisions. If the brand leaned too hard into AI, it risked sounding like exactly the kind of software company its clients never hired.

We could have made the AI sound as impressive as possible. Instead, we realised Propeller's most convincing AI story was one it had been living for years.

How Propeller was built

Over 18 years, Propeller made choices most firms avoided. It built dedicated teams instead of relying on contractors, and it standardised ways of working instead of designing a bespoke engagement for every client. It also chose a single, unified data system over flexibility. Those choices suited clients who needed advanced capabilities, but they priced Propeller out of serving everyone.

Then AI came along, and those same choices started to pay off. As founder Chris Fenster puts it, "We didn't pivot to AI; we discovered we'd been preparing for it all along."

Why that matters so much now

In finance, AI is only useful if you can trust what it produces. It can do a lot of work quickly, but it can't yet check itself. In DualEntry's 2026 AI accounting benchmark, released in March this year, even the best model got about 1 in 4 accounting tasks wrong. A mistake like that can mean a failed audit or a fundraise that falls apart.

So founders need a different kind of reassurance now. They want to know that someone will check the work and catch what the AI gets wrong. Propeller calls this the Output Gap, meaning the lag between what AI can do and what a business can actually put safely to use. Closing it takes experienced people and years of data. Propeller has both. 

Telling the brand’s story with confidence

Propeller already had the line to carry all of this: "Not fractional. Foundational." Our job was to make the rest of the brand as sure of itself as that line.

That meant writing plainly and letting the facts do the work. Twenty-four unicorns says more than any adjective could, so we let the numbers lead.

It also meant talking to founders directly and starting with the problem they're facing, well before the copy gets anywhere near a service. The growth-stage page opens with "Growth can break finance, unless you build it right." The Embedded teams page makes its case in one line: "You don't hand sensitive financial data to strangers." Instead of simply listing services like controllership and FP&A, we described what Propeller's support looks like in practice. For an early-stage company, that means Propeller will "set up your books, build your first model, and close month one before you hire employee ten."

We also let the FAQs tackle uncomfortable questions head-on. They all sit under the heading "Questions your board will ask", and they include the ones a sceptical founder would actually raise, such as "Won't AI-native finance tools make an embedded team unnecessary?" Propeller is even upfront about who it isn't for: "We're your partner through multiple stages and exit, or we're probably not a fit."

And we were careful with AI language. We steered away from "automate", because it's easy to hear as "replace", and wrote about Propeller's people and its AI working together instead. The result is a brand that talks about AI with the confidence of a business that's been building towards it since 2008.

What this means for your brand

If you run a service business, don't start your AI story with the technology. Start with what your clients already trust you for. If that's your people and the way you work, you already have the most convincing part of your AI story.

Here's where we'd start:

• Take another look at the decisions you've been downplaying. 
  Make a list of the choices that made your business more selective or slower to scale. Propeller's dedicated teams and unified data once priced it out of serving everyone, and those same choices now sit at the heart of its AI story.
• Answer the question clients are really asking. 
  Once AI can do the work, clients want to know who's checking it. Show them by naming the people involved, explaining how outputs get reviewed and being honest about what the technology can't do yet.
• Write your words-to-avoid list first.
  Go through your current copy and flag any term that could make clients think people are being replaced, or that you're a supplier rather than a partner. Propeller's list included "automate", "vendor" and "contractor", with "augment" and "amplify" as the preferred way to talk about AI.
• Swap adjectives for evidence.
  Wherever your copy calls something impressive or innovative, look for the number that proves it. If you can't find one, it's worth asking whether the claim belongs there at all.
• Give the problem a name.
  Propeller calls the lag between what AI can do and what a business can actually put to use the Output Gap, and that name gives clients a simple way to see why its people matter. A clear name for your clients' problem can do the same job for you.

Contact us to discover more about what we can bring to your upcoming brand repositioning process. And sign up to the Series Eight newsletter to read the full Propeller case study as soon as it drops (link’s in our footer).