AI Transformation
Lattice CEO: Build for the Success of People First
Companies have spent the last few years measuring the value of AI by the headcount it lets them cut. Sarah Franklin thinks that’s the strategic error of the era, and that the market is already correcting it.
In this episode, Sarah Franklin, CEO of Lattice, joins host Joelle Emerson to talk about why AI transformation is really a people transformation and what it looks like to build for the success of people first. She also gets into what evidence does to the performance review, why even OpenAI buys people software rather than building it, the AI coach that sits in on Lattice’s one-on-ones, and why she thinks companies should retire the sabbatical.
Key Takeaways
Sarah Franklin is CEO of Lattice, the people platform used by companies from OpenAI to Duolingo, and the first non-founder to lead the company. She arrived after 15 years at Salesforce, where she built Trailhead into a learning movement of millions and rose to president and chief marketing officer.
AI transformation is a people transformation. Sarah says she didn’t expect the people function to be thrust into the epicenter of relevance this fast. Lattice’s own chief people officer was just promoted to lead the company’s AI transformation — a trend she now sees everywhere, because aligning people, technology, and process is the transformation.
The system of people is harder than the technical system. Technical systems are complicated; people are nuanced, emotional, and unpredictable. Sarah notes that roughly 70% of a company’s combined technology and people investment is anchored in its human resource — so tuning how those people work isn’t an HR side quest, it’s the biggest lever on the business.
Evidence beats memory in performance reviews. Lattice’s evidence-based reviews pull the work someone actually did — from the CRM, the code repository, the care system — and surface it inside the review. It’s hard to debate evidence, Sarah says, and it fills the gaps that recency and primacy bias otherwise fill. It also gives calibration meetings, which take management teams offline for weeks each year, a shared factual baseline instead of a debate about what “exceeds expectations” means.
Building it yourself is a rerun of the cloud era. AWS made it possible for any company to build its own ERP or CRM. Nobody did, Sarah points out, because a build is only the down payment — the maintenance is the mortgage. Her proof that the same logic holds for AI: OpenAI itself is a Lattice customer rather than hacking together its own version.
Cutting people is a temporary strategic error. Pricing AI software by the human capital expense it can replace is, in Sarah’s view, the mistake of the moment. Tokens aren’t cheap, prices change on a dime, and AI doesn’t replace knowledge or judgment. The durable math is unlocking more from the people you already pay for — when the top line grows, the bottom line isn’t something you need to cut.
Managers are the most important, least-trained layer. Companies invest in executive coaching and onboarding, then pluck their best ICs into management and tell them to figure it out. Lattice’s AI coach joins one-on-ones and privately flags what a human coach would — you talked over this person, you gave feedback without data, you’ve drifted from the goals. Sarah’s observation: critical feedback is easier to hear from an AI, because nobody feels judged by it.
Every company that kills the performance review brings it back. Joelle has watched three or four companies retire reviews or scores — and every one reinstated them, because by year-end people were asking “but how am I doing?” Sarah’s frame: continuous feedback is an and, not an or. The review is a human ritual people trust — at least once a year, a real conversation about performance, compensation, and growth.
Retire the sabbatical. Sarah’s unpopular opinion: people don’t burn out in five-year increments. A formalized clock on rest reminds people to leave; wellness principles should be reality-based, not time-based — especially in tech, where she sees women burning out and leaving.
Chapters
[00:00] Introduction
[01:23] Meet Sarah Franklin
[02:17] Why HR, and why now
[03:45] AI transformation is a people transformation
[05:01] How the people team’s job is changing
[07:46] People systems are harder than technical ones
[09:57] Removing bias from performance reviews
[15:40] Should companies build their own HR tech
[22:20] Cutting people is a strategic error
[27:04] Customers putting people first
[30:55] Prepping a board deck with AI
[33:41] Generalists or specialists on people teams
[36:36] When leadership doesn’t value HR
[40:18] The performance review comes back
[45:46] The AI coach in your one-on-ones
[50:45] Lightning round
[52:46] Closing thoughts
Show Links
Watch on Youtube
Listen on Spotify
Listen on Apple Podcast
Learn more about Lattice
Follow Sarah Franklin on LinkedIn
Connect with Joelle Emerson on LinkedIn
Follow Paradigm on LinkedIn
