From 20 person-days per topic to under an hour per venture

Client

Date

$10B+ global optics group, corporate venture unit

09/2026

The challenge

The corporate venture unit of a $10B+ global optics group runs an annual internal venture challenge. Its teams had to validate their ideas on market, viability and desirability before the next stage gate. Each validation topic was scoped at twenty person-days, and team research capacity was the binding constraint.

The solution

FifthRow deployed a venture validation engine that the unit operated itself. A team's own pitch material goes in as a dataset; a source-backed assessment of its market, its competition, its economics and its customer case comes out, one report per team, in the unit's own validation structure.

Under 1 hour

market assessment, from 20 person-days

11 of 11

ventures assessed in one day

Market assessment cycle timeUnder 1 hour, from 20 person-days

Cohort coverage11 ventures assessed to one standard

Validation topics systematised3, each run across the whole cohort

Analyst-hours equivalent167,754 hours saved in twelve weeks; equivalent to 346 FTEs

01 — VALIDATION METHOD ENCODING

The unit already had a validation method: a fixed set of objectives every team has to answer before it advances. What it did not have was a way to run that method at cohort scale. The first job was to turn the structure the unit teaches into systems that execute it.

Core Activities

Reverse-engineered the unit's own objective structure into three runnable report types
Made each team's pitch material a research input rather than a source of truth
Removed the manual hunt for each team's inputs before any research could start
Kept the unit's own wording and ordering, so outputs land in the programme's format
Ran the method first against ideas the unit had already worked through by hand

The Results

A validation method that exists as a system rather than as a set of instructions, so every team in the programme is researched against the same questions in the same order, whoever is running it.

20 person-days

per topic, replaced by one system

02 — COHORT REPORT BATCHES

Validation research had always been bounded by what a three-person unit could take on alongside everything else. The first cohort-wide batch changed the unit of work from one team to the whole programme.

Core Activities

Assessed the whole cohort on desirability and viability from a single batch
Covered willingness to pay, pricing, cost structure and the business case
Covered customer journeys, problem statements and the concept's own assumptions
Gave every team the same depth, not only the ones someone had time for
Returned one report per team, in the hands of the teams the next morning

The Results

Every team in the programme went into its next working session holding a source-backed read on whether anyone wants its idea and whether the economics hold, produced overnight rather than requested and queued.

20,222

analyst-hours equivalent, one evening's batch

03 — ASSUMPTION TESTING

The unit did not want research that necessarily agreed with its teams. A team's pitch deck, one-pager and pitch transcript were to be treated as claims to be tested, not as evidence to be quoted back.

Core Activities

Rebuilt each answer independently before comparing it with the team's own claim
Held the team's own material out of the evidence pool, so agreement carries weight
Flagged where segmentation, sizing or pricing did not survive outside evidence
Put the outside source behind every sizing, pricing and competitor claim
Reconciled the two readings into one view rather than picking a winner

The Results

Assessments that argue with a submission rather than restate it, so a review board can see which of a team's assumptions held up against the market and which did not, with the evidence attached to each.

37,470

research questions answered from outside sources

04 — COHORT MARKET ASSESSMENT

Market and competition was the topic the unit ranked first and the one it had scoped at twenty person-days. The finished assessment ran across the entire cohort in a single working day.

Core Activities

Assessed segmentation, scaling potential, market size, competitors, advantage and market forces
Launched all eleven ventures from one batch, inside five minutes
Finished every report inside the hour, on 13,620 verified sources
Tested top-down sizing against each team's own bottom-up numbers
Returned one report per venture, in the unit's own six-part order

The Results

Eleven ventures holding a market and competition assessment of the same depth and the same shape, produced inside one working day, on a topic the unit had budgeted in person-weeks.

42 minutes

average per venture market assessment

What it proved

The unit's own validation structure went in and came back out as three report systems. Eleven venture teams received desirability and viability assessments from a single evening's batch, then a six-part market and competition assessment on a single day, every report launched together and every one finished inside the hour. Across twelve weeks the platform ran 755 studies over 1,801 analysis steps.

The unit finished the programme holding its validation method as something that runs rather than something that is explained: three report systems, built and revised by its own staff, each executed across a full cohort in one launch. Teams with no venture experience reached their next stage carrying the same evidence base as experienced ones, and research capacity stopped deciding which ideas got looked at properly. Across those 755 runs the platform records 167,754 analyst-hours saved in twelve weeks, produced by three people serving eleven teams.

Results

Metric	Benchmark	Result	Improvement
Market and competition assessment	20 person-days per topic, the unit's own scoping	Under 1 hour per venture	20 person-days to under an hour
Cohort coverage	Team research capacity was the binding constraint	11 ventures assessed to one standard, same day	Every team served, none queued
Validation reporting	Inputs assembled by hand, one team at a time	One launch, whole cohort, one report each	Per-team effort replaced by a batch
Assumption handling	A team's own material taken as the starting evidence	Independent rebuild, then reconciliation	Claims tested in a non-biased way rather than repeated
Analyst-hours saved	Three people carrying eleven venture teams, working manually	167,754 hours across 755 runs	Equivalent of adding 346 full-time analysts across a twelve-week programme

ENCODED VALIDATION METHOD

A programme's own validation structure runs as a system, so every team is assessed against the same questions in the same order.

COHORT-WIDE BATCHES

One launch produces a report per team, so coverage stops depending on which idea somebody had time to research.

FALSIFICATION-FIRST ANALYSIS

Each answer is rebuilt independently before it meets the team's own claim, so a submitted assumption is tested rather than echoed.

SOURCED ASSUMPTIONS

Sizing, pricing and competitor claims arrive with the outside evidence behind them, so a review board interrogates the reasoning rather than the confidence.

LEVEL STAGE GATES

Teams without prior venture experience reach a decision gate holding the same evidence base as teams that have done it before.

OPERATOR-BUILT SYSTEMS

The programme's own staff assemble and revise the report systems, so a change to the method never waits on outside build effort.