A complete innovation sprint, from problem framing to ranked concepts

Client

Date

$4B+ global water and plumbing products manufacturer

11/2024

The challenge

A $4B+ global water and plumbing products manufacturer runs product innovation as a sequence: frame the problem, generate options, build the survivors into concepts, then rank them. Every stage was a separate piece of work with its own tooling and its own wait, so the sequence rarely ran end to end on one product question.

The solution

FifthRow ran the team's own innovation sequence as one connected sprint. A product area goes in; framed problem statements, a generated option set, developed concept boards and a ranked prioritisation come out - the same sequence, applied to a second product experience without rebuilding it.

4 stages

of one sprint connected end to end

2 product experiences

run on the same sequence

Sprint stages connected4, from problem framing to ranked prioritisation

Product experiences run2, on the same sequence

Distinct analyses run40 across the engagement

Competitive set breadthBeyond the team's previous hand-assembled limit

01 — PROBLEM FRAMING

An innovation sprint is only as good as the question it starts from. The first stage turned a product area into framed problem statements the rest of the sequence could actually run against.

Core Activities

Turned each product area into framed how-might-we problem statements
Read the industry challenges and external factors bearing on the area
Identified customer pain points, including among people who do not buy today
Framed the innovation question inside the resources actually available to act on it
Ran the framing with more than one member of the team, not a single author

The Results

Problem statements specific enough to generate against, grounded in external evidence rather than internal instinct, and owned by more than one person on the team.

2 people

framing the same problem set together

02 — OPTION GENERATION

With the question framed, the point was volume before judgement: generate widely, then let the ranking do the narrowing rather than filtering by opinion at the start.

Core Activities

Ran high-volume idea generation against each framed problem statement
Explored untapped spaces in the category as a distinct question from existing demand
Segmented the customer base four ways, so options could be aimed rather than generic
Kept generation separate from evaluation, so nothing was filtered too early
Held the output in a form the next stage could take directly

The Results

A wide option set aimed at defined segments rather than at the category in general, and no early filtering by whoever happened to be reviewing.

4 segmentation cuts

behind the options generated

03 — COMPETITIVE & SPECIFICATION GROUNDING

Concepts get argued down in review by whoever knows the competitive set best. Grounding came before concept development, not after it, and it went wider than the team could previously assemble.

Core Activities

Ran competitive product analysis across the category, repeatedly as the set changed
Collected and compared product specifications at component level
Built competitive sets larger than the team could previously assemble by hand
Mapped the consumer purchase journey for the product area
Read technological and regulatory movements bearing on the concepts

The Results

A competitive and specification base under the concepts before they were developed, wider than hand assembly allowed, so a concept review argued about the concept rather than about what the market contains.

7 reads

competitive product analysis, the most-run analysis on the account

04 — CONCEPT BOARDS & RANKED PRIORITISATION

The last two stages are the ones that produce something a funder can act on: options built into concept boards, then ranked on a consistent scale.

Core Activities

Developed the surviving options into concept boards, across two build passes
Produced a full product requirements document where a concept needed one
Ranked the concepts on a consistent three-tier prioritisation scale
Applied the same ranking to both product experiences, so the two were comparable
Left the sequence configured, so a third product area would not need rebuilding

The Results

Ranked concept boards for two separate product experiences, built and scored the same way, and a sprint sequence the team could point at a third area without starting again.

1 scale

both product experiences ranked comparably

What it proved

The team's own innovation sequence was run as one connected sprint: problem framing into high-volume option generation, then competitive and specification grounding, then concept boards and a three-tier ranked prioritisation. Forty distinct analyses ran across six weeks - four segmentation cuts, seven separate competitive product reads, component-level specification comparison, consumer journey mapping. The sequence then ran again on a different product area without being rebuilt.

Two things changed about how the work could be done. The competitive set stopped being bounded by what one person could assemble - the client's own prior limit with other tools was ten to fifteen - so concepts arrived at review already grounded in the category rather than getting argued down in it. And because the sequence was configured rather than assembled, the second product experience got the same treatment as the first, which is what makes two runs comparable instead of two separate exercises.

Results

Metric	Benchmark	Result	Improvement
Sprint continuity	Each stage a separate piece of work with its own tooling	4 stages connected end to end	One sequence rather than four handoffs
Competitive set breadth	"10-15 has been my limit with other tools", client's own words	Sets built well beyond that limit, run 7 times	Grounding no longer capped by hand assembly
Concept grounding	Competitive evidence gathered after concepts were argued	Grounding completed before concept development	Concepts reviewed on merit, not on market gaps
Prioritisation	Options ranked by the judgement in the room	One three-tier scale applied to both product areas	Two product experiences directly comparable
Repeatability	Every product area starting the sequence from scratch	Sequence configured and re-run on a second area	The second run started from a method

FRAME BEFORE YOU GENERATE

Idea generation against a vague question produces volume nobody can rank. Turning a product area into explicit problem statements first is what makes the rest of the sprint sortable.

GENERATE WIDE, FILTER LATE

Filtering options at the point of generation means filtering by whoever is in the room. Keeping generation and evaluation apart lets the ranking do the narrowing on a stated scale.

GROUND BEFORE THE REVIEW, NOT DURING IT

Concepts die in review because someone knows the competitive set better than the deck does. Doing the grounding before concept development moves the argument onto the concept itself.

A HAND-ASSEMBLED COMPETITIVE SET HAS A CEILING

Ten to fifteen competitors is roughly what one person can assemble and keep current. That ceiling silently shapes every concept judged against it.

COMPARE SPECIFICATIONS AT COMPONENT LEVEL

Product-level comparison tells you what exists. Component-level comparison tells you what it would take to beat it, which is the question a concept actually has to answer.

THE SECOND RUN IS THE PROOF

A sprint that works once may have worked because of the people. A sprint that runs a second time on a different product area, to the same standard, is a method.

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

An entire investment team building its own research agents in one day

Client

Corporate venture arm of a $10B+ global electronics manufacturer

Date

08/2026

11,805

analyst-hours equivalent, 8 FTEs

Matter More

Let's Talk