From a quarterly scan to a standing foresight function

Client

Date

Global specialty chemicals leader

08/2026

The challenge

A global specialty chemicals leader ran corporate foresight, technology scouting and policy monitoring out of its new ventures and sustainability function. Keeping up with three converging streams of change consumed the equivalent of several full-time analyst years each year, and the work was reactive, siloed across teams and reset every quarter.

The solution

FifthRow set up a standing intelligence function on its platform: a defined set of priority areas in, ranked and structured signals out, on a fixed cadence and into the surface the team already works in. It runs continuously rather than on request, and the group's own new ventures team operates it alongside FifthRow.

236,200

analyst-hours equivalent, three months

20+

areas under continuous watch

Coverage20+ technology and policy areas

Cadencedaily brief, fortnightly radar and sweep

Structure per areathree levels, value chain to individual entrant

Evidence base, three months~223,500 sources analysed

Analyst time equivalent~236,200 hours in three months

01 — STANDING TECHNOLOGY SCOUTING

The scouting layer maps a landscape once and then keeps it current. Every fortnight it returns what has appeared since the last cycle rather than rebuilding the whole picture.

Core Activities

Maps 20+ priority areas in a single pass, in hours not a quarter
Surfaces new technologies, vendors, partnerships, patents and funding events every fortnight
Structures every area at three levels: value chain, technology segment, individual entrant
Distinguishes seven signal types, so a funding round and a policy move are not one event
Reaches sources general-purpose tools miss: specialist databases, public-affairs filings, lobbying records, industry-body publications

The Results

The team stopped re-running a landscape study to find out what had changed. A new entrant or a new technology arrives already placed inside a segment the team tracks, with sources, partnerships and benchmarking criteria attached, so fit can be judged on arrival rather than after a follow-up commission.

20+

areas kept current, not rebuilt

02 — SCHEDULED POLICY & MARKET MONITORING

Policy and market monitoring moved off the list of things somebody would get to and onto a fixed schedule, covering Europe and the US on the same cadence.

Core Activities

Sweeps the policy pipelines bearing on the group's products and customers every fortnight
Picks up upstream public-affairs and lobbying signals that precede a rule
Draws on regulatory filings, industry bodies, trade press and specialist databases
Covers Europe and the US continuously, with targeted scouting elsewhere
Returns implications for the business, not a list of links

The Results

Coverage became a specification rather than a function of who had a free afternoon. Nothing in scope waits longer than a cycle to be seen, and the signals that arrive earliest, in filings and public-affairs activity, are read at the same time as the trade press that reports them later.

2 a month

sweeps nobody has to start

03 — DAILY SIGNAL SYNTHESIS

Six concurrent scans run each night, one per top-level strategic priority, and land as a single ranked brief in the morning instead of a feed nobody had time to read.

Core Activities

Runs one scan per strategic priority, so each is watched on its own terms
Synthesises and ranks what it finds rather than aggregating it
Delivers one morning brief in place of about fifty daily articles
Makes the whole body of intelligence answerable through one conversational interface
Exposes it inside the collaboration tool the team already works in

The Results

The day starts with implications rather than links. Three streams that innovation, sustainability and corporate strategy each tracked separately now converge into one ranked read, and the team spends its attention on what the signals mean rather than on finding them.

~50

daily articles, one ranked brief

04 — DEEP DIVES & WHITE-SPACE EXPLORATION

When a signal is worth more than a mention, the same function takes it further, on demand, without a new brief or an external study.

Core Activities

Explores where incumbents are weak and new entrants forming, and returns where the room is
Takes a chosen white space through to a staffable project plan
Runs deep dives on an individual company the radar surfaces
Classifies and ranks potential partners for investment and acquisition shortlisting
Returns market sizing, business model, build-buy-partner path, regulatory, IP, capital and adoption risks, and a next step

The Results

Exploration ends in something the business can act on rather than in a reading list. A white space that looks promising on the radar can be sized, framed as a problem statement, tested against the risks and written up as a business case on the same function that found it.

~35,500

questions answered without an analyst

What it proved

In the first three months the group stood up one continuous function across 20+ technology and policy areas: a fortnightly radar returning only what is new, fortnightly policy and market sweeps across Europe and the US, six daily scans synthesised into one ranked morning brief, deep dives on demand. Across all platform use the team analysed about 223,500 sources, had about 35,500 questions answered and saved an estimated 236,200 analyst hours.

The group now holds a foresight function that does not reset. Coverage is specified rather than dependent on who had time, and every area is structured the same way, value chain to individual entrant, so a new signal arrives already placed in a tracked segment. The three streams innovation, sustainability and corporate strategy followed separately converge into one ranked morning brief, and a promising white space becomes a staffable project plan without commissioning anything.

Results

MetricBenchmarkResultImprovement

Analyst time keeping up

Benchmark: several full-time analyst years, three teamsResult: ~236,200 hours saved in three months

Capacity returned to assessment

Continuity of the work

Benchmark: reactive, siloed, reset quarterlyResult: one standing function carrying context between cycles

Quarterly restart replaced by continuity

Daily reading load

Benchmark: about fifty unranked articles dailyResult: one ranked brief each morning

Reading replaced by ranked signal

Landscape coverage

Benchmark: as far as an analyst reachedResult: 20+ areas re-read every cycle

Specified rather than sampled

Structure per scouted area

Benchmark: signals collected flatResult: three levels, sources and benchmarking attached

Assessable on arrival

LAYERED SIGNAL CLASSIFICATION

Every area is structured from value chain down to individual entrant, so a new signal arrives already placed in the segment it belongs to.

DELTA-ONLY SCOUTING

The radar returns what is new since the last cycle, so a landscape stays current without being rebuilt each time it is read.

SCHEDULED COVERAGE

Policy and market sweeps run on a fixed cadence, so what gets watched no longer depends on who had time this quarter.

RANKED MORNING SIGNAL

A synthesised brief arrives each morning in place of a raw feed, so the day starts with implications rather than with links to read.

HARD-TO-REACH SOURCES

Specialist databases, public-affairs filings, lobbying records and industry-body publications are read alongside commercial data, so weak signals surface before mainstream coverage reaches them.

WHITE SPACE TO PROJECT PLAN

A promising white space is sized, risked and written up as a project plan on the same function that surfaced it, so exploration ends in something staffable.

Recent case studies

See more proof

The intelligence function six business areas never had to staff

Client

$30B+ global electrification and automation group

Date

08/2026

596,505

analyst-hours equivalent, 150 FTEs

From one-off research requests to a standing new-business intelligence practice

Client

$500M+ global speciality chemicals manufacturer

Date

08/2026

382,406

analyst-hours equivalent, 155 FTEs

Matter More

Let's Talk