Product and Architecture Roadmaps: Two Plans That Could Only Blur Together

CASE STUDY · UNNAMED ENGAGEMENT.

This page’s client is one of the biggest units at a global research and advisory company, where we ran an organisational change programme. Nine product and technology teams. Hundreds of millions of dollars of revenue a year. A goal of doubling it. To give the programme evidence, we assessed the organisation: twenty-five dimensions put to a scored question set, and a qualitative pass for everything a number can’t hold.

The analysis started with the map.

Start with the map

r ≥ 0.61 · Architecture roadmap
System resilienceArchitecture roadmapAutomated testingBusiness value clarityFeature definitionDelivery predictabilityTeam alignmentPortfolio agilityPortfolio visionPrioritisationProduct intakeProduct management rolesProduct roadmapPI planning rolesQuality metricsQuality confidenceRelease processRisk managementFlow of workStakeholder managementSustainable paceTeam-level planningTechnical debtTestable requirementsCross-team planningTalent clock-speed
size: mentions in the written observationscolour: mean sentiment, low to highedge: significant score correlation, thicker is stronger

Correlation, not causation: every edge is a measured statistical association (p<0.05) from this engagement, not an asserted cause. Edges below |r|=0.35 are omitted for legibility; the rest are drawn thicker and more opaque the stronger they are. The map stays focused on Architecture roadmap: hover any dimension to preview its own connections against it, and use the strength slider to keep only its strongest links.

The map opens centred on the architecture roadmap, one of the sorest, best-connected regions of the whole network, and the first thing to notice is who its close company is: the product roadmap. Two plans with different owners, different audiences and different time horizons, moving almost as a single measurement. This page is about why two roadmaps that are supposed to inform each other ended up unable to move apart.

What the numbers said

median 7 scale 1-10
012345678910

Architecture roadmap scores land on a median of 7 out of 10, with eleven respondents out of sixty-one at 4 or below. Not a crisis on paper. The written record is where the strain shows.

What the words said

What they scored
What they wrote: sentiment same axis
012345678910

The two dark blobs not lining up is the chart.

Just four coded observations on this dimension came from the team survey, and they are mixed; the weight of the record sits in the leaders’ retrospectives, quoted in the next section. What the team side does say pairs neatly across the two roadmaps:

“Our architectural planning is constrained by a lot of past decisions and external systems.”

“The product roadmap, when we get into sprints, is what is a little more murky.”

Constraint upstream, murk downstream. Hold that ordering; the leaders explain where it comes from.

Where it lived

Team 6 med 9
Team 1 med 7.5
Team 4 med 7.5
Team 2 med 7
Team 9 med 7
Team 3 med 6.5
Team 7 med 6
Team 5 med 5.5
Team 8 med 5
012345678910

Nine teams, anonymised, best to worst. Thin rows render wider and flatter; less data looks uncertain, not falsely precise.

Ranging from 9 at the highest down to 5 at the lowest, the nine team medians form a gradient rather than one rogue team. Which is what you’d expect when the constraint is a shared platform: every team plans on the same shifting ground, some just stand closer to the fault line.

What the architecture roadmap turned out to be entangled with

The measured neighbourhood:

  • The product roadmap (r = .74): the pair this page is named for
  • Program-planning role clarity (r = .75)
  • Automated testing (r = .66)

Why can’t the two roadmaps move apart? The leaders’ retrospectives answer with the state of the platform both roadmaps stand on:

“We have massive data integration issues across the company. Having a complete view of our customers and prospects is really challenging.”

“These major infrastructural hurdles are slowing down our ability to innovate.”

“The pace of innovation has been challenging, particularly as we’re now in the second year of a platform transition.”

Mid-transition, with unresolved data integration underneath, every product commitment inherits architectural uncertainty: you cannot schedule what you cannot scope, and you cannot scope on shifting ground. So the business plan and the technical plan stop being two plans; they become two views of the same constraint, and their scores blur together, which is a coherent reading of the r = .74 edge. Read as correlation only: nothing here proves a cause. The role edge (r = .75) adds the governance half: a roadmap nobody clearly owns cannot absorb uncertainty on the other roadmap’s behalf.

What happened next

The platform transition was already the organisation’s declared priority; what the assessment added was the shape of its cost, showing up not as an engineering line item but inside the product roadmap’s murkiness and the intake queue’s uncertainty. Investment continued into the transition, and into training for product and portfolio teams alike, while the organisation went on monitoring as conditions changed. The numbers that followed are not mine to claim. A check that travels: if your architecture roadmap is in flux, audit whether your product roadmap’s vagueness is actually its own problem before reorganising the product team.

A note on the data

A real engagement; an unnamed client, and it stays that way. Vendors go unnamed as well. Team names and identifying details are removed, and quotes are lightly edited for anonymity. Nothing on this page is an asserted figure: every number and chart is computed at render time against the underlying assessment data, almost 2,000 quantitative and qualitative data points feeding the calculation. One organisation, correlations not causes.

Morné Wiggins · Agility at Scale · Talk to me

Privacy Preference Center