The homepage says 15x faster model validation and 64% faster dashboard setup; the data readiness page says 30x and 85%. The about page shows "46.1%" over a source line for "61.1%", and the server-rendered pages read "5-5 hours" and "5–5% of project hours", probably counters that animate in the browser. A VDC director checking claims, and any crawler or AI assistant, sees the wrong numbers.
First action
List every proof number on the homepage, about, pricing and data readiness pages with its source.
How to measure it
Proof numbers with one value and a source
- Now
- –
- Target
- all of them
- By
- 2 weeks
Steps
Inventory numbers
Every stat on the main pages and its source.
Product Marketing Lead 1 day
Pick the canonical value
One number per claim, the most defensible.
Product Marketing Lead 1 day
Fix rendering
Server-render final values; animate from the real number.
Engineering lead 1 day
Add sources
Footnote or tooltip naming the customer or survey.
Product Marketing Lead 1 day
Update decks
Same numbers in sales decks and one-pagers.
Product Marketing Lead 1 day
Check monthly
Any new page uses the canonical list.
Product Marketing Lead monthly
Why it works
Pick one sourced number per claim (with the customer or survey behind it), use it everywhere, add a "source" tooltip or footnote, and render the final values in the HTML so the count-up animation starts from the true number.
The Product Marketing Lead owns messaging and the ROI story. Enterprise buyers in a skeptical industry check numbers, and conflicting ones undo the "trusted data" message itself.
Risks and how to handle them
Lower numbers look less impressive
Name the customer behind each; specific beats large.
Evidence
- Your websiteHomepage: "Unlock design data to reduce rework" (a year ago: "The better way to work together in 3D"); sections for architecture, engineering, construction and owners; Speckle Intelligence, "AI that queries the model, not a PDF"; a model checker that scores each package. Proof numbers: "15x Faster model validation", "40x Faster turnaround on 3D model client presentations", "64% Faster custom analytics dashboard setup" and, in the server-rendered page, "5-5 hours Less weekly rework per team".Source
- Your websiteThe State of Data Readiness in AEC page (2026 AEC data survey, gated whitepaper) gives different proof numbers for the same claims: "5-10 hours of rework eliminated per team, every week", "30x faster model validation", "85% faster custom analytics setup", "98% reduction in reporting cycles".Source
- Your websiteAbout page: teams lose "28–46 hours a week per project to manual validation, QA/QC, data conditioning, and reporting"; "30–38% of repetitive VDC/BIM work can be automated"; "46.1% of AEC leaders say their data isn't clean enough to power AI reliably"; "5–5% of project hours are lost to data issues", while the source line beneath reads "Source for 61.1% and 5–10%". Open source since 2016; founder and CEO Dimitrie Stefanescu.Source
- HiringProduct Marketing Lead (Remote USA, Eastern time, posted July 2026, Growth team): own "how we package, position, and sell a category-defining product into the enterprise", turning a platform "currently positioned as the design intelligence layer for AEC" into packaging, messaging and enablement "that helps our team win six-figure enterprise deals" with General Contractors and Engineering Service Providers. Four jobs: packaging and positioning (name, bundle, price-frame, differentiate per segment), messaging, a sales enablement engine (decks, one-pagers, case studies, competitor comparisons, ROI tools), and thought leadership and webinars on the AI transformation. "Build-from-zero"; 7+ years of product marketing.Source