Jellyfish vs PMG

Brandtech's Rebecca Sykes on AI Implementation Challenges

In a podcast interview, Rebecca Sykes, partner at The Brandtech Group and head of its emerging technology practice, argues that generative AI can work with messy, siloed data, contrary to conventional wisdom. She emphasizes that brand-specific data, even if disorganized, prevents AI outputs from becoming generic. Sykes highlights that many AI pilots fail at scale because companies only automate isolated steps without re-engineering the entire workflow, using the 'frog in a sock' analogy. She advocates for building new operating models around AI capabilities, focusing on people, processes, and new roles like AI governance. The episode also discusses Brandtech's new global data and AI practice, Jellyfish's 'share of model' tool, and the importance of video in AI brand recommendations.

• Rebecca Sykes is a partner at The Brandtech Group and leads its emerging technology practice.
• Sykes argues that generative AI can work with messy, siloed data, not just pristine data.
• She uses the 'frog in a sock' analogy to illustrate how existing operating models hinder AI transformation.

Black History Month: Culture Chats, with Global Chief People Officer, Annabelle Walters

Annabelle Walters, Global Chief People Officer at Jellyfish, joined The Digital Voice™ Black History Month Culture Chats for a candid discussion, unpacking what it truly means for black women to own their space in executive leadership, whilst navigating the nuanced reality of vulnerability at the top.

Harvard Study: AI Does Not Truly Boost Developer Productivity

A study by Harvard researchers Fiona Chen and James Stratton, based on data from the engineering analytics platform Jellyfish covering 300 million work outputs from over 700,000 employees at 718 companies (January 2021 to March 2026), finds that while AI coding tools increase code output, these gains do not translate into actual productivity improvements. Reactive AI assistants led to a 12% increase in lines of code and a 9% increase in commits, but only the commit effect was statistically significant. Semi-autonomous AI agents showed larger increases (30% more code lines, 20% more commits, 23% more pull requests), but the review bottleneck negated these gains. The average time to review a pull request rose by 49%, the share of pull requests requiring changes nearly doubled, and comments per pull request increased by 35%. The proportion of engineers doing code reviews rose by 14%. Qualitative interviews with 21 professionals confirmed that AI speeds up routine tasks but requires more manual checking for complex tasks.

• Harvard researchers analyzed 300 million work outputs from Jellyfish, covering 718 companies and over 700,000 employees from January 2021 to March 2026.
• Reactive AI assistants increased code lines by 12%, commits by 9%, and pull requests by 5%, but only the commit effect was statistically significant.
• Semi-autonomous AI agents increased code lines by 30%, commits by 20%, and pull requests by 23%.

AI Disrupts Job Applications, Skills-Based Hiring Rises

The traditional resume is becoming obsolete as AI allows candidates to generate endless applications, forcing companies to rethink hiring. A 2025 study shows UK graduate employers received a record 140 applications per vacancy. Experts like Barb Hyman (Sapia.ai) advocate for skills-based hiring, which assesses candidates on demonstrated abilities rather than employment history. Around two-thirds of UK organizations have adopted this approach, per Michael Page research. Companies like Taktile and PMG now assess candidates through practical exercises and interpersonal skills, moving away from paper-based evaluation. AI is also seen as reducing human bias in screening, though final decisions remain with human teams. The shift emphasizes continuous learning and adaptability over static credentials.

• AI enables candidates to generate unlimited resumes, pressuring employers to adopt new hiring methods.
• UK graduate employers received an average of 140 applications per vacancy in 2025, a 30-year high.
• Two-thirds of UK organizations have adopted skills-based hiring, per Michael Page research.

Creator Marketing Matures as PMG Acquires Digital Voices

In this AdExchanger Talks podcast episode, Jennifer Quigley-Jones, VP of Strategy & Partnerships at PMG, discusses the evolution of creator marketing. She highlights that a decade ago, influencer marketing agencies primarily focused on Instagram, missing opportunities on YouTube. This gap led her to leave YouTube in 2017 and found Digital Voices, an agency that was acquired by PMG earlier in 2026. Quigley-Jones emphasizes that creator marketing is now a strategic imperative, with brands investing across multiple platforms. She discusses how technology, including AI models and performance data analysis tools, can improve creator-brand relationships and scale campaigns from 25 to 1000 creators, democratizing media spend. The episode also covers why influencer marketing should be viewed as a strategy rather than a tactic, and the importance of aligning goals between advertisers and creators.

• Digital Voices was acquired by PMG earlier in 2026.
• Jennifer Quigley-Jones founded Digital Voices in 2017 after leaving YouTube.
• Quigley-Jones is VP of Strategy & Partnerships at PMG.

Agencies Audit AI Agent Costs; Micro1 Outbids Google for Spirit Data

This AdExchanger daily news roundup covers several stories. A Gartner survey of 1,300 marketers finds 60% of AI users incur unexpected costs due to lack of oversight, prompting agencies like PMG and Rise to build auditing tools to control AI token spending. AI startup micro1 has bid $12.5 million for Spirit Airlines' data, outbidding Google's $10 million, though Google's deal is already approved. Separately, the FBI is investigating a dark web service selling 153 million driver's license scans from IDScan.net. The roundup also includes opinion pieces on the Google ad tech remedies, The Trade Desk layoffs, PepsiCo moving its media account to Publicis, and Instagram's AI detection issues.

• Gartner survey: 60% of companies using AI incur unexpected charges or exceed budgets.
• PMG built a tool to limit AI token usage and provide recommendations.
• Rise, a Quad agency, uses PubMatic's audit log feature to track AI agent behavior.