What Companies Had to Fix Before AI Reached Production

Bar chart of the largest gap to close before AI reached production at 72 companies: data quality 58%, legacy system integration 24%, security and privacy compliance 11%, other 7%.

For 58% of 72 companies, data quality for AI was the biggest gap to close before production. Legacy system integration came second at 24%, and security and privacy compliance third at 11%.

Study methodology: 72 executives and team leads who took AI into production answered over 21 days through MentionMatch and Featured. Each respondent named the single largest gap they had to close, so the shares add up to 100%. All figures are self-reported and were not independently audited.

What companies had to fix before AI reached production

Largest readiness gapShare of respondents
Data quality and consistency58%
Legacy system integration24%
Security and privacy compliance11%
Other, including adoption and trust7%
Total100%

Essential statistics: data quality for AI and other readiness gaps

  • 58% of 72 companies named data quality and consistency as the largest gap before AI reached production.
  • 24% named integration with legacy CRM, ERP, ticketing and in-house systems.
  • 11% named security and privacy compliance for sensitive records.
  • The remaining 7% named team adoption, trust, output consistency or context management.
  • 8Rental cut dispatch preparation from 18 minutes to 4 per booking.
  • Proximity Plumbing cut its booking error rate from 12.4% to 3.1%.
  • Sundance Networks cut legacy integration time from 6 weeks to 2 with custom hybrid APIs.

Key takeaways

  • Data quality outranked every other gap by more than 2 to 1, with 58% against 24% for legacy integration.
  • Closing the data gap meant standardizing input formats and cleaning old records. Some respondents spent weeks or months on that work.
  • Teams that standardized data before deployment reported output that held up 2 to 3 times better than teams that cleaned it later, by their own account.
  • Data quality was the most common gap at 58%, while 22% of the same sample named data readiness as the deciding factor for production.
  • Legacy integration ranked second at 24%, so roughly 1 in 4 teams spent the hardest part of readiness on connecting old systems.

Actionable insights

  • Audit and standardize source data before the pilot starts, since 58% of companies named data quality for AI as their largest gap.
  • Schedule data cleanup as its own workstream, because some respondents spent weeks or months closing the data gap.
  • Map every legacy system the AI must reach during scoping, as integration was the largest gap for 24% of companies.
  • Plan API bridges or middleware for older CRM and ERP systems, as Ketch did to connect its AI pipelines to older storage tools.
  • Log prompts and outputs and restrict access early where records are sensitive, since security and privacy compliance was the largest gap for 11%.

“Most teams already held the data their AI needed, but in formats a model could not use. For some of them, standardizing it took weeks or months before the first workflow could run.” — Kirill Funtikov, R&D Lead, SumatoSoft

SumatoSoft logo
If you have any questions, email us info@sumatosoft.com

    Please be informed that when you click the Send button Sumatosoft will process your personal data in accordance with our Privacy notice for the purpose of providing you with appropriate information.

    Vlad Fedortsov, IT Sales Expert at SumatoSoft
    Vlad Fedortsov
    Account Manager
    Book an intro call
    Thank you!
    We've received your message and will get back to you within 24 hours.
    Do you want to book a call? Book now
    SumatoSoft clients logo