Every profitable business eventually needs a second growth curve. Almost every attempt to launch one damages the first. The question is not whether to design a new business model — it is how to install it without unwinding the one that pays the bills.
How do you design a new business model that adds growth infrastructure without destabilising the business you already have?
The reversal
The default framing of new-business-model design is a strategy question: pick the model. Cross-industry research on growth-curve transitions12 shows the failure mode is almost never model choice; it is model integration. Companies that fail to launch a second growth curve overwhelmingly report that the new model competed with the existing one for capital, attention, and talent — and that the existing model, being the profitable one, quietly starved the new one until it died. The design decision that matters is not what the new model is; it is how it will co-exist with the model you already have.
The insight stack
What actually moves the P&L
Decide the integration mode before the model
There are three viable integration modes: fully integrated (new model runs inside the existing P&L), ring-fenced (separate P&L, shared services), and structurally separate (independent operating unit, distinct capital base). Each has different failure modes. Fully integrated new models die from resource competition; ring-fenced models die from cultural drift; structurally separate models die from lack of parent-company patience. Choose the mode that matches the tension you can most credibly manage — not the mode that maximises theoretical synergy.
Protect the cash-generating model with explicit capital ring-fencing
New models consume disproportionate capital and executive time. The core business must be defended by a written capital allocation contract: the new model may draw no more than X% of forecast operating cash in year one, and cannot compete with the top three capital priorities of the core business. Without the contract, capital drift is the default and the core business quietly weakens.
Build the new model on a compounding unit, not a project
The compounding logic — the specific transaction, subscription, or engagement that produces a repeatable margin — is the artefact worth engineering. If the new model cannot be described as one repeatable unit with a repeatable margin, it is not yet a business model; it is a portfolio of one-offs. Delay launch until the compounding unit is designed and tested at small scale — twenty to fifty instances is usually enough evidence.
Instrument the new model against three tests, not one
Financial (does the unit economics work at scale?), operational (does the delivery architecture hold at scale?), and strategic (does it strengthen or dilute the core positioning?). A new model that passes financial and operational but fails strategic is a distraction that will eventually be sold or wound down. Kill it before it consumes another eighteen months of executive attention.
Sequence go-to-market against the client base you already have
The single highest-ROI move for a new business model is to launch it into the existing client base first — same buyers, adjacent problem, warm decision-support. Selling the new model to strangers is a strategic option; it is almost never the first quarter's tactic. Existing clients are the fastest, cheapest source of evidence that the model works, and their feedback is the only pre-scale signal that a stranger buyer will eventually confirm.
Case example
A £34M professional services firm launching a productised offer
The problem: the firm needed to add a repeatable, higher-margin revenue stream to complement its bespoke advisory work — faster to deliver, cheaper to sell, and better fitted to a client base that had begun to demand a lighter engagement option. An initial attempt to launch a productised offer as an inside-P&L extension of the advisory practice consumed twelve months, produced £180K of revenue, and drew senior consultants away from advisory work worth £2.4M. The programme was re-scoped as a ring-fenced business unit with a written capital contract (no more than 8% of core operating cash in year one), a dedicated P&L, and go-to-market restricted to the existing client base for the first two quarters. Twelve months later, the ring-fenced unit had produced £2.1M in revenue at 42% contribution margin, the core advisory practice had grown 14%, and the productised offer's customer feedback had become the primary input into the firm's next positioning refresh.
Mini-playbook
Six-move installation of a second business model
Choose the integration mode (integrated, ring-fenced, separate) before the model.
Write a capital-allocation contract protecting the core business.
Engineer the compounding unit; validate at twenty-to-fifty instances before launch.
Instrument against financial, operational, AND strategic tests — kill on strategic failure.
Launch into the existing client base first; open to strangers only after evidence.
Review the model quarterly against explicit thresholds — no funding by inertia.
How Strategy Labs installs this
Anchored to Corporate & Growth Strategy
Strategy Labs designs new-business-model transitions inside CAE, sequencing the integration-mode decision, capital-allocation contract, compounding-unit design, and go-to-market ramp across the seven-stage engagement lifecycle. Governance for the ring-fenced unit is built into the same operating cadence as the core business, not bolted on separately.
Customer-need validation, willingness-to-pay research, and integration-mode benchmarking run inside PDC, so the model design is anchored to primary evidence — not to internal enthusiasm or competitor imitation.
Frequently asked
Related questions executives ask
- How do we know whether to integrate or ring-fence the new model?
- Integrate only if the new model shares more than 70% of the delivery architecture with the core business, and its buying committee overlaps by more than half with the core buyer. Otherwise ring-fence — the operational friction of forcing integration usually costs more than the theoretical synergy delivers.
- How big should the initial capital commitment be?
- Small enough that the failure of the new model does not damage the core P&L, and large enough that the new model has a real chance to prove its compounding unit. In mid-market services, that band is usually 5–10% of forecast operating cash in year one.
- When do we exit the ring-fence and integrate?
- Only after the new model has produced repeatable unit economics at scale (typically 200+ instances) and its strategic contribution to the core positioning is proven. Premature integration collapses the new model back into the core P&L and its distinct discipline evaporates.
Over to you
If you designed a second growth curve as a ring-fenced unit with a written capital contract, which of your current commitments would fund it — and which would you pause to protect the core?
Continue reading
More Strategic Growth briefings
When is a growth business actually ready for growth capital?
Most scale-ups raise growth capital to prove product-market fit. The compounders raise after they have already proven it — and use the money to buy speed, not survival.
Read briefingWhy is pricing the highest-ROI decision most executive teams underinvest in?
A 1% improvement in captured price flows to operating profit at close to 100%. A 1% improvement in variable cost or volume never does. Yet pricing sits in a spreadsheet the CEO has not opened in nine months.
Read briefingHow should a growth business decide what to defund?
Every growth portfolio has a hidden liability: the initiatives that stopped compounding two years ago but still consume executive attention. Defunding is the highest-leverage capital-allocation decision most boards never make.
Read briefing
Discussion
(…)Comments are moderated before appearing. Your email is only used for moderation and is never shown publicly.
Loading discussion…