Engine 2

Engine 2 banking — you'll also hear it called a Side Car bank, or sidecar strategy — is increasingly popular among banks that want to deliver a brand and offering accurately and expediently to a market segment, without the burden or restrictions of the main organisation.

Whichever name you use, the play is simple to describe and hard to fudge. You stand up a second, autonomous bank or division on a modern stack beside the main organisation. It serves its segment properly from day one. Then, as it proves itself, you migrate workloads and customers across — until the old engine can be switched off.

Temenos invited Moroku CEO Colin Weir to share his thoughts on what Engine 2 is, and on the innovation triad — the three fundamentals for breakout Engine 2 digital banking. As Colin explains in the video below, Engine 2 is not an innovation lab or a skunkworks. It's a bank.

Colin Weir on Engine 2 · In collaboration with Temenos

Origin

Where the Idea Comes From

The name is borrowed from Bain & Company's Engine 2 growth concept. Engine 1 is the core business: it pays the bills, holds the customers, and runs on assumptions that were true when it was built. Engine 2 is the next core business — built inside the same company, on today's assumptions, with real ownership and its own P&L.

Bain's central warning is the one most banks ignore. Your second engine is not a side project. It is not a percentage of someone's time, or a floor with beanbags. It's where the growth compounds — while Engine 1 funds it. Bain has applied the model to banking directly: the institutions that get it right treat Engine 2 as the future of the firm, resourced accordingly, and protected from Engine 1's antibodies.

In banking, the translation is concrete. Engine 1 is the legacy core, the product silos, the change queue measured in quarters. Engine 2 is a new bank on a modern stack — cloud-native core, API-first, digital origination — free to serve its segment the way the segment actually wants.

Market Proof

The Pattern in the Wild

This isn't theory. The Side Car playbook is documented across the last five years, at home and abroad.

86 400 → UBank (NAB)

NAB acquired the neobank 86 400 and then moved UBank's customers onto its platform. The lesson: you can buy the engine rather than build it. NAB ran old and new in parallel, and migrated only when the new stack was feature-complete and the team had migration muscle.

ANZ Plus

The biggest Australian version. ANZ built an entirely new retail platform and launched it beside the classic bank, progressively migrating existing customers across. Decommissioning the old core is the end of the journey, not the start — ANZ has been explicit that the classic bank keeps running until Plus has earned the workload.

Up + Bendigo

The closest analogue for mutuals. Up ran as a modern stack beside the Bendigo mothership and reached customers — young, digital, national — that the main brand couldn't. It worked well enough that Bendigo acquired Ferocia, the team behind it, to bring the engine fully in-house.

Unloan (CBA)

The single-product Side Car. CBA launched Unloan as one simple digital home loan on a brand-new stack. You don't have to rebuild the whole bank to start; one product, done properly on modern rails, is a legitimate Engine 2.

The Global Tier-1s

JPMorgan launched Chase UK as a new digital bank rather than exporting its US stack. Standard Chartered built Mox in Hong Kong on a greenfield core. NatWest launched Mettle for small business. Three of the world's most sophisticated banks looked at modernising in place and chose to build beside instead.

The Honest Bit

Why Engine 2s Fail

The graveyard is real, and it's instructive. RBS built Bó, a consumer digital bank, and closed it six months after launch. Xinja, one of Australia's most hyped neobanks, handed back its licence and returned every deposit. The wider neobank graveyard tells one story on repeat.

The common cause is not technology. The modern stacks worked. The common cause is that these were deposit-taking machines with no lending engine and therefore no path to revenue. Xinja paid market-leading interest on deposits while its loan product was still on the roadmap. That's not a business; that's a countdown.

A deposits-only Side Car burns capital acquiring customers it can't monetise, then asks the board for more. The Side Cars that survived — UBank, Up, Unloan, ANZ Plus — all had, or quickly acquired, a way to lend. Design the revenue engine in from day one, or don't start.

The New Angle

Engine 2 for Mutuals and Community Banks

Here's the part almost nobody writes about: you don't need to be NAB to run this play.

Picture a credit union on a legacy core. The vendor's roadmap serves the vendor. Costs rise every renewal. The contract locks you in for years, and every conversation about modernisation ends with a migration estimate that would terrify the board. Rip-and-replace is a bet-the-institution project, so nobody makes it, and the meter keeps running.

A Side Car at mutual scale breaks the deadlock. New-to-bank members onboard straight onto the modern stack from day one — they never touch the legacy core. Existing members migrate wave by wave, on your schedule, as each product on the new stack proves itself. And the legacy core is decommissioned in line with contract termination, not in a panic before it.

That contract end-date is the forcing function — and the negotiating leverage. A credit union with a live, proven Side Car walks into the renewal conversation with an alternative on the table. A credit union without one signs whatever is put in front of it. The best time to stand up Engine 2 is two to three years before the contract ends; the second-best time is now.

The Architecture

Engine 2 and Core+ Are the Same Journey

There's a false choice doing the rounds: either hollow out the core in place (the Core+ or "core plus" pattern) or build a Side Car bank. Put an orchestration layer in the right place and they're the same journey at different speeds.

With the Moroku Digital Services Layer sitting between your channels and your ledgers, every experience — app, web, broker, branch — talks to one API. Which core answers is a routing decision, not an architecture decision. Both cores sit behind the same API for the whole transition.

One API in front. Two engines behind. Migration becomes routing.

That changes what Engine 2 costs and what it risks. Hollowing out in place becomes Engine 2 the day you stand up the second ledger: workloads re-point, they don't rebuild. Channels don't notice. Members don't notice. The migration stops being a big bang and becomes a sequence of routing changes you can rehearse, stage and reverse. It's the manufacturing model from our Digital Banking Strategy, applied to the hardest problem in the shop.

Where to Start

Prove It Before You Commit the Bank

The worst way to start a Side Car is a slideware business case that asks the board to commit the institution on assumptions nobody has tested. The best way is an experiment: pick the two or three assumptions the whole case hangs on — will this segment onboard digitally, will the lending engine price and approve fast enough, will the routing hold — and prove them in a lab, in weeks, for a fraction of the cost of being wrong.

That's exactly what our Readiness Assessments are built for: a fast, honest read on where your institution stands and which assumptions to test first.