AMA: Lovable B2B PMM Lead, Talia Moyal on How to build an AI-first PMM function and team

AI-native teams will design workflows / org structure around the new economics AI creates. Drafting, summarizing, automating, etc. are now cheap. They'll also think about what humans can uniquely contribute to the equation which is now centered around decision making / judgement and taste. AI-augmented means you plug AI into existing workflows. Some examples across workflows / assets: Competitive intelligence - create an app with all things competitive an app with key competitors tracking all ne ...Read More

It's helpful to not treat AI transformation as a side project (otherwise it just feels looming). I try to just make changes to processes as we go that way we are sacrificing minimal speed upfront in favor of being able to move quickly later. Some questions I ask myself to figure out where to start in a process (including examples of low hanging fruit): 1. Is this work being reused elsewhere? If the output needs to show up in multiple places, we redesign it as a structured source of truth: Messag ...Read More

People with taste!! People who can look at an output and know if it's hit a quality bar. What separates good from great from amazing from inspiring is someones ability to raise the bar on quality and figure out how to create an emotional connection with an audience. That is not something AI can do. Practically, what does this look like in interviews? Ask people to show you the work they are most proud of Ask people who show you where they think brands they follow have AI slop Ask them what contr ...Read More

Anything that includes generating drafts, automating workflows, or repurposing materials, definitely uses AI. Anything that gets to an editorial stage, requires nuance, or is in an execution phase, needs a human-in-the-loop (i.e. coordinating across different vendors, reviewing writing drafts, reviewing creative). Humans are here to provide judgement / taste controls. Human-in-the-loop is really the key for me here. You should treat AI as your thought partner. It's not there to do all the work f ...Read More

There is really no navigating around these. When it comes to governance, it's important your product meets baseline level of requirements. This may include things like SSO, SCIM, audit logs, granular permissionings, deployment model considerations (depending on the space you're in). As a PMM, you work in here by showing your company what the baseline entry requirements are to sell your product, then what the scale requirements are to sell your product. Both are important if you want the company ...Read More

Have dropped this answer in a few places now, but these are my top three use cases: comp intelligence, enablement, product launches. I love these because the apps you create can be repurposed to support other workflows!Competitive intelligence - create an app with all things competitive an app with key competitors tracking all news mentions how their websites change email comms they're sending mentions of them in sales calls (automatically scrape all transcripts) win/loss analysis objection hand ...Read More

I personally haven’t found a ton of benefit in organized courses. I’ve found it far more beneficial to encourage people to try problem-solving with AI directly. For example, pick something that feels broken and try fixing it with AI, or ask AI how to approach solving the problem. I’d also suggest not being too restrictive about which tools it recommends. When it comes to resistance, there are always people who think, “that will be really difficult.” The best way to get them unstuck from that fee ...Read More

It's important to separate this into two questions: 1. concerns about AI-generated content and 2. concerns about data privacy. Two is easier to address as your company / product either does these things or it doesn't. You should clearly explain: What data is accessed Where it is processed Whether it is stored Who can see it Whether it trains models Just make sure to define data boundaries, explain access controls and traceability, and take any data-related incidents seriously. For question one, ...Read More