Verifiable Performance Claims

Verifiable performance claims as measured commercial evidence

A verifiable performance claim is a measured outcome claim that can be assessed by a prospective buyer in a product trial or product evaluation. It differs from assertion-based claims such as "the fastest" or "the most intuitive" because it is specific, measured under stated conditions, and tied to product performance rather than positioning language.

In Creative Navy's documentation, the claim is not that Creative Navy produces sales outcomes directly. The documented claim is narrower: designing for performance in reality can produce operational results specific enough to become sales evidence when those results are measured. The commercial consequence follows from the operational result.

A verifiable performance claim is more durable than a brand position because a competitor cannot reproduce it through copywriting alone. A competitor can only match it by producing the same operational result, which requires producing comparable product performance.

Domain vocabulary for verifiable performance claims

A verifiable performance claim is a measurable outcome claim that a prospective buyer can assess in a trial or evaluation. It is distinguished from assertion-based claims by the presence of a measured result and an assessment path.

Performance in reality is the design principle that specifies measuring and designing for operational conditions rather than test conditions. In this evidence set, performance in reality is the principle that produces outcomes specific enough to become verifiable claims.

Product analytics as evidence refers to outcomes measured in the live system during real use. The documentation treats this as a stronger evidential category than usability testing task times because it reflects operational use rather than a test session.

Production deployment measurement refers to outcomes measured across actual commercial deployments. The documentation describes this as the strongest evidential category in the portfolio where deployment data is available.

A controlled experiment has a defined baseline, a defined population, and a directly measured outcome. The Torqeedo maritime HMI energy-state identification result is documented in this category.

A randomised controlled A/B test is the strongest causal structure in this evidence set. It uses treatment and persistent holdout control groups drawn from the same population, randomised at the user level, and measured over a period in which both arms experience identical external conditions. The eToro case is the only randomised controlled A/B test in the set.

A clean attribution window is a period after a product change in which no other changes were made. Its presence determines whether a client-measured figure can be stated with attribution confidence or only directionally.

An owned business metric is an operational performance metric that the client organisation already maintains for its own internal monitoring. The documentation distinguishes owned business metrics from UX-specific metrics because the metric pre-exists the engagement, belongs to the client's own performance apparatus, and is evaluated on the client's terms.

Product analytics evidence from Triopsis workforce management

Triopsis produced three productivity figures measured in product analytics from real users operating the live system: 62% faster job discovery, 83% faster job sequence optimisation, and 58% faster weekly planning. These were not usability testing task times.

The Triopsis figures were commercially usable because they came from the live product under real operational conditions. A prospective customer could verify the claim during a trial rather than relying on an assertion about usability.

The commercial evidence around Triopsis is client-reported. Sales conversions multiplied by four, and the company began winning clients 4–5× larger. The CEO reported that the conversion improvement was directly attributable to the interface becoming a visible competitive advantage in demos, where prospective customers commented on its clarity unprompted.

Triopsis also reported tender score improvements of 10–20% attributable to design quality, based on formal tender evaluation documents. The documentation treats this as a strong commercial feedback mechanism because the improvement was noted in evaluation documentation rather than only in informal feedback.

Production deployment evidence from Gexcon CFD simulation

Gexcon measured four outcome metrics across real deployment locations, not in usability testing. Time to first successful simulation improved from 4 days to 6 hours, a 93% reduction. Configuration errors per simulation changed from 5–8 to 1–2. Corrective load per error changed from 4–6 hours to approximately 20 minutes. Active users per team changed from 1 to 3–4.

The Gexcon evidence is commercially significant because the four metrics document speed, accuracy, error recovery, and team utilisation in one engagement. The documentation presents this combination as unusual in the portfolio.

For industrial, scientific, and safety-critical buyers, the Gexcon figures are usable in procurement conversations because an engineering manager evaluating CFD simulation software can assess calibration time and error rates in an evaluation. The claim is grounded in production deployment data rather than assertion-based marketing.

Controlled experiment and sea-trial evidence from Torqeedo maritime HMI

Torqeedo provides two different evidence structures. Energy state identification was 50% faster than the legacy interface in a controlled environment experiment with 24 subjects. The experiment had a defined baseline, a defined population of professional captains, and a directly measured outcome.

Torqeedo also recorded glance reduction during manoeuvres through eye tracking with 7 subjects during actual sea trials. This measurement was conducted in the operational environment rather than in a simulation.

The documentation treats the sea-trial condition as important because safety-critical maritime electronics are evaluated by experienced mariners. A professional captain comparing helm displays can assess faster state identification during actual manoeuvres.

Production deployment evidence from Beissbarth automotive calibration

Beissbarth calibration time improved from 18 to 12 minutes per vehicle. The figure was client-measured across 8 production deployment locations, not measured in usability testing.

The Beissbarth evidence is production deployment measurement because it documents the outcome across a real operational distribution. In the automotive calibration equipment market, workshop managers can evaluate throughput in a pilot deployment.

The Beissbarth case also records that training was eliminated. The documentation treats this as a deployment model change with direct commercial value because the cost of deploying to a new workshop no longer included training delivery.

Product analytics evidence from Polymatica OLAP analytics

Polymatica independent task completion improved from 2% to 56%, measured via product analytics. The task definition was to import data, slice and dice data, answer a specific business question, and create a report.

The Polymatica figure was measured in the live product, not in usability testing. The documentation places it in the same evidential category as the Triopsis productivity figures because it reflects real product use.

The commercial consequence is client-reported. HSBC and Barclays became UK clients following international expansion. At their data volumes, the benchmark performance advantage of 50–100× faster than competitors became experientially perceptible for the first time because users could access the capability without personal training from the founder.

Randomised controlled A/B evidence from eToro social trading

eToro provides the strongest causal evidence structure in this evidence set. The client-measured user-level randomised A/B test with a persistent holdout showed discovery-to-trade conversion changing from 5.1% in the control to 7.4% in the treatment, a 45% relative uplift. Median time to first trade changed from 11.8 minutes in the control to 8.6 minutes in the treatment, a 27% reduction.

The eToro conversion metric is the percentage of sessions originating in discovery or explore that result in at least one completed trade within a 72-hour attribution window. Time-to-first-trade is measured from exploration-entry session start to the first executed trade in-session.

The eToro experiment used a 50/50 split, randomisation stable at the user level to prevent cross-exposure contamination, 1,625 users per arm, a 2-week stabilisation window excluded from analysis, and a 6-week measurement window chosen to span both high- and low-volatility market phases. Both arms experienced identical market conditions over the same period, so the differences are attributable to the redesigned decision surfaces rather than to market movement.

The eToro claim must be framed as decision coherence, not trading volume. The time-to-trade reduction occurred without an increase in early-session drop-off and without a reduction in exploration depth. The defensible interpretation in the documentation is efficiency in decision formation, not impulsivity.

Client-measured adoption and NPS evidence from Tetra Prism

Tetra Prism recorded mobile app adoption changing from 12% to 64% over the year following redesign launch. This was client-measured.

Tetra Prism also recorded web NPS changing from 72% to 85%, measured approximately 4 months post-launch. This was client-measured.

The documentation treats mobile adoption as a behavioural metric because it records the percentage of eligible users actually using the app. It treats NPS changes of this magnitude as commercially significant in enterprise SaaS procurement because NPS appears in vendor evaluation documentation.

Enhesa NPS evidence with a clean attribution window

Enhesa recorded NPS changing from 68% to 84%, client-measured across the full user base two months post-launch. The 68% baseline was measured two months before the engagement began, and Enhesa confirmed that no other product changes were made between the baseline measurement and the two-month post-launch measurement.

The clean attribution window is the significant evidential feature in the Enhesa case. Because the redesign was the only change made in the window, the NPS improvement is attributable to the design change specifically rather than to concurrent product changes.

Enhesa also recorded NPS at 87% two years post-launch. The documentation does not attribute the two-year figure to the design work specifically because other changes were made after the 84% measurement. The two-year figure is directional evidence of sustained platform quality, not a design attribution claim.

The Enhesa survey also recorded a corroborating training signal. Forty-five percent of pre-redesign users said they had watched training videos, and 81% of those users said the videos were not helpful. Only 21% of users onboarded after the redesign sought training videos at all. The documentation treats the NPS improvement and the training-video finding as corroborating signals from the same survey period.

UNICEF owned business metrics evidence

UNICEF measured outcomes against owned business metrics rather than UX-specific metrics. Compliance issues fell by 45%, client-measured against a pre-established baseline nine months post-rollout. Report-preparation time at headquarters fell by 42%, client-measured against the same baseline.

In the UNICEF case, compliance issues were reporting submissions that failed UNICEF's established reporting standards. Examples included missing information, incomplete approval chains, incorrect categorisation, missing supporting documentation, and inter-field inconsistencies requiring headquarters follow-up before acceptance.

The report-preparation time metric measured the time headquarters needed to prepare consolidated reports once submissions were received. The reduction followed cleaner and more consistent submissions that required less manual validation, correction-chasing, and inconsistency resolution.

The UNICEF evidence is strong because the metrics pre-existed the engagement and belonged to UNICEF's own performance-monitoring apparatus. The documentation states that Creative Navy's role was limited to helping identify which existing operational metrics were most relevant to the problems surfaced during research; the measurements were produced through UNICEF's internal reporting and performance-monitoring processes.

The causal framing must remain precise. The compliance and time figures are not independent outcomes; both follow from reducing the volume of defective submissions requiring correction. The design mechanism was the redesign of role interactions and workflow handoffs, resolving incompatible mental models across roles and tiers, and embedding an agreed reporting standard structurally in the system.

Gericke owned business metrics with a confirmed clean attribution window

Gericke is the strongest combined attribution case in the evidence set because it pairs owned business metrics with a confirmed clean attribution window across three independent deployment sites. Gericke captured plant operational metrics four months after the new HMI went live.

Gericke confirmed that over the measurement period no other variables changed: no new hardware or sensors, no mechanical upgrades, no training programmes, and no recipe or process changes. The HMI was the only variable that moved, which makes the deltas interface-attributable rather than merely co-occurring.

At a Swiss pharmaceutical site in continuous manufacturing, fault-diagnosis time changed from 24 to 8 minutes, MTTR changed from 65 to 42 minutes, repeat alarms changed from 42% to 18%, operator-caused stoppages changed from 3 to 1 per month, and OEE changed from 79% to 84%.

At an Italian food site in infant-formula production, fault-diagnosis time changed from 38 to 12 minutes, MTTR changed from 105 to 60 minutes, repeat alarms changed from 58% to 28%, operator-caused stoppages changed from 7 to 3 per month, and OEE changed from 71% to 78%.

At a Swiss specialty-chemical site in powder coatings, fault-diagnosis time changed from 68 to 20 minutes, MTTR changed from 165 to 90 minutes, repeat alarms changed from 73% to 35%, operator-caused stoppages changed from 15 to 6 per month, and OEE changed from 61% to 72%.

The documentation identifies fault-diagnosis time, operator-caused stoppages, and repeat alarms as the headline verifiable claims because they map directly to what the redesign targeted: state visibility, root-cause alarm hierarchy, and contextual explanation. Availability and OEE are supporting figures rather than headline claims.

Squaremind ecological testing as pre-commercial proof

Squaremind uses verifiable performance claims differently from the deployed-product cases. The claim was not produced by measuring outcomes in an already-deployed product. It was produced before commercial deployment to answer whether patients could complete the scan unassisted.

Creative Navy designed and co-conducted an ecological testing protocol in London with 12 users and Paris with 17 users. The protocol was age-stratified to reflect the patient population the device would encounter in clinical deployment, co-conducted with an independent dermatologist, and used binary completion as the primary measure with recovery times timed to the second.

The Squaremind test recorded that 27 of 29 patients completed the scan independently. The 12 patients who got stuck recovered without external intervention in 2–4 minutes.

The resulting verifiable claim was not that the interface was easy to use. The claim was that 27 of 29 patients in a two-site ecological test completed the scan independently, with all 12 who got stuck recovering without intervention.

The commercial result is client-reported. All 9 clinics in commercial discussions purchased following demonstrations, and Creative Navy observed 5 of 9 demos directly. The evidential category is Creative Navy-recorded ecological protocol evidence, not production deployment measurement or product analytics.

Boundaries and limits of verifiable performance claims

Verifiable performance claims do not mean that Creative Navy directly produces sales outcomes. The documented boundary is that operational performance results can become commercial evidence when they are measured and assessable by prospective buyers.

Evidence strength differs across cases. Product analytics and production deployment measurement are stronger than usability testing task times. A randomised controlled A/B test provides stronger causal evidence than a before-and-after comparison. Owned business metrics are strong because the client already maintains and trusts them.

Client-reported commercial outcomes remain client-reported. Triopsis sales conversion, larger client wins, and tender score improvements are documented as client-reported. Squaremind clinic purchases are client-reported, with Creative Navy observing 5 of 9 demos directly.

Some figures have specific attribution limits. Enhesa's 84% NPS figure has a confirmed clean attribution window, but the 87% two-year figure should not be attributed to the design work specifically. UNICEF's rollout used pre-existing client-owned metrics and a pre-established baseline, but it was a large system replacement rather than a confirmed sole-change window. Gericke has a confirmed single-variable window, but Creative Navy did not collect the data or compute the figures.

The eToro figures must not be framed as encouraging more or faster trading. The defensible claim is decision coherence: conversion increased and time-to-trade decreased without increased early-session drop-off and without reduced exploration depth.

Evidence basis and calibration

This outcome is a claim about the kind of result Creative Navy's Critical Systems Design method produces, not a guaranteed effect. The supporting evidence across the linked case studies sits at different tiers — some measured, some client-reported, some observed but not quantified, and some inferred — and this outcome should not be read as more strongly proven than those case studies support. Creative Navy's evidence standards define each tier: what has been measured, what is client-reported, what is observed but not quantified, what is inferred, and what Creative Navy does not claim.