Twelve decisions.
Told the way they were made.

The challenge

Doubling group revenue meant growth from developing markets, but two-thirds of projected growth sat in a few hundred cities. Syndicated providers had little city-level coverage and the board question was due at a global general-manager conference eight weeks from kick-off.

The work

The model measured relevant demand rather than headline GDP. Income distributions isolated consumers crossing the threshold into branded consumption; where category data stopped, airline seats, internet performance and other proxies rebuilt demand city by city.

The result

A demand map covering 6,400 metropolitan areas was adopted to steer sales and marketing allocation. Cities in deprioritised countries - including Manila, Lima and Monterrey - outranked whole markets on the former list.

1,260 indicators tested
Eight weeks to answer
General-manager conference

The challenge

The country’s grey online-gambling market had little robust public data. The chief executive of a major listed operator needed an independent answer before a licensing window closed for 18 months; the surface story supported entry.

The work

Fieldwork was run personally in the local capital city. Two commissioned surveys and 40 interviews rebuilt the market bottom-up and found the regulator’s card-payment data had failed to remove player winnings. Effective tax, demographic decline and impending TV competition were put into an entry P&L.

The result

The market sized at c.€330m gross gaming revenue - the bottom of published estimates - and the entry case showed two years of losses and single-digit margins thereafter. Findings went to the CEO and executive team a week before the deadline; the decision was not to enter.

c.€330m gross gaming revenue
1,517 consumers surveyed
40 industry interviews

The challenge

National statistics used incompatible definitions. Some markets were saturated while others were barely banked, yet every country team could produce a persuasive investment case. The centre needed a comparable view of where to commit development resource - and where to stop.

The work

Rather than extrapolate card growth, consumer spend was split by payment value and behaviour, each band assigned a structural ceiling for card penetration. All 30 markets were then rebuilt on a consistent basis across debit, credit and commercial programmes.

The result

The model sized c.€1.7tn of ten-year headroom, and the ranking contradicted the highest-profile internal cases: the largest pools sat in everyday mid-value spend and cheque displacement. The scheme used one fact base to direct resource toward them.

30 markets rebuilt
€600bn debit pool
€360bn credit pool

The challenge

A trading desk held exposure through physical assets and forward freight agreements. Before buying or time-chartering more ships, it wanted its bullish thesis tested against fundamentals rather than broker sentiment.

The work

Demand was rebuilt as tonne-miles across every major exporter-importer pair, using real sea distances. Supply was forecast by tanker class from the confirmed order book, with fleet productivity modelled explicitly. The two sides met in one route-level charter-rate forecast.

The result

The fundamentals contradicted the thesis: supply growth outpaced tonne-mile demand. The finding was given straight and the client exited rather than doubled down; the disposal later returned nine figures, with the model informing valuation, timing and negotiation.

c.720m tonnes modelled
Supply c.10% p.a.
Demand c.8% p.a.

The challenge

A UK sponsor was taking four digital marketplaces to market. The largest US platform contained the biggest value pool and weakest evidence; management’s growth case required independent validation to carry the process.

The work

The work centred on the network effects linking sellers, buyers and listings. Sixty interviews, surveys at scale and a bottom-up rebuild of the conversion funnel from more than a million platform rows surfaced and quantified an unstated marketplace flywheel.

The result

An independently validated case was presented personally to six mid-market PE houses, alongside an explicit read of US execution risk. The network-effect thesis and cohort analysis differentiated the data room; the business sold. The company has since gone on to list publicly.

60 structured interviews
1m+ data rows
Four marketplaces

The challenge

The investor needed to know whether UK vitamins and supplements merited entry, which model could win and which available companies deserved a closer look - without commissioning a full diligence cycle.

The work

The market was resegmented by consumer need rather than product type. Interviews, store and site visits, web traffic and bottom-up P&Ls showed a flat category concealing fast-growth niches - and that advertising and differentiation, not inherited trust, explained brand trajectory.

The result

The review supported a premium, vertically integrated online model focused on needs-based niches, rejected the personalised route and ranked ten targets. Amazon depth and subscription economics remained explicit open risks; the work informed entry, not a transaction.

c.£880m market
Needs-based resegmentation
16 expert interviews

The challenge

Successive socio-political shocks had hit footfall, an import restriction had squeezed the highest-value category and one store had lost a third of sales. The question was whether an investable business remained under the disruption.

The work

Five years of quarterly performance separated external shocks from operating economics, stripping out new-store drag and a one-period subsidy. One store’s underperformance was analysed openly rather than explained away.

The result

The prize was sized from affluent-household growth, while improving overheads, fit-out capital and store maturity demonstrated the underlying economics. The raise was supported on that evidence - not a generic recovery story. The arc since has proven the case: the company went on to list at a c.$60m valuation and now trades at a nine-figure market cap.

Overheads down c.6 points
Affluent segment ~17% p.a.
Five years rebuilt

The challenge

The proposal combined discretionary spend, a recession and a small operator competing with majors. The founder needed a case built to survive credit scrutiny.

The work

Office for National Statistics data back to 1978 inverted the consensus that gambling was simply recession-proof: over-the-counter betting was falling while machine gaming grew. Every proposed site was scored, growth rebased to 3% and economics benchmarked against the majors.

The result

The lending case named two underperforming shops, restated one result on a conservative basis and showed a viable 15-shop business on under half the funding. The company raised £3.2m when banks had stopped lending.

159 customers surveyed
47 competitor customers
15-shop defensive floor

The challenge

The consortium had the design and engineering for an estuary airport. It lacked the economic argument for why the decision mattered, with outline submissions due within weeks.

The work

A proprietary model of 2,000 metropolitan economies computed the share of global GDP within reach by flight time, while ten years of route data compared six hubs. The frame moved from where Britain could tolerate a runway to where Britain could reach the world economy.

The result

The client’s published submission reproduced and credited the 2,000-city model, with the 27% / 94% connectivity finding at the centre of its case. The proposal was not shortlisted; the Commission cited cost and environment.

2,000 metros modelled
Six hubs compared
Credited in published submission

The challenge

The presenting choice was 25,000 or 40,000 seats. The real question was how to deploy close to a billion dollars across stadium, offices, hotel, retail, residential and parking - in a city where the published statistics stopped short of the question.

The work

The city’s population was rebuilt by nationality, gender and age from four independent sources. Even combined high-growth cases left a 40,000-seat bowl filling after 20 years. Every asset class was then tested against demand, yield and land opportunity cost.

The result

The recommended plan anchored on an expandable 25,000-seat stadium, with a nine-figure reallocation of capital from empty seats into productive yielding assets including a luxury hotel, retail mall, and mixed-use office and residential development.

Four population sources
Demand, yield, opportunity cost
25,000 not 40,000 seats

The challenge

The industry’s debate was defensive: how to survive European competition. The ministry needed a strategy credible to government and thousands of mostly small firms - and specific enough to act on.

The work

The same agreement opening the home market also opened Europe. Competitiveness was benchmarked in both directions. A bottom-up sizing and 80-professional survey showed the gap was quality and market access, while the existing tariffs also taxed imported machinery and inputs.

The result

Twelve programmes assigned action across government, the chamber and individual firms. The roadmap was adopted and promoted to more than 4,000 companies. The stated limit remained: a 15-year strategy depends on execution across political cycles the adviser cannot control.

4,000+ firms reached
80 industry interviews
15-year strategy

The challenge

The plan called for 20m m² of development and up to half a million residents. Current sales velocity put the nominal 20-year plan at 120 years. The real question was what would make 20km² worth more than the land it sits on.

The work

Residential and commercial listings were regressed, residents surveyed and employers interviewed. Metro walking distance explained more value than distance to the centre; families traded commute for quality and green space; employers followed talent. The site was treated as an economy, not a housing product.

The result

The board received a demand-grounded route map: zone by walking minutes, move the mix toward the knowledge economy and sequence absorption over 20 years. The model stated its macro assumptions so the board could revisit them; execution rested with the client.

c.139,000 listings
c.3,400 residents
50+ employer interviews