AMA: Asana Head of AI & Platform, Product Marketing, Neel Patel on AI and Product Marketing

This is something I've been grappling with for some time too, but have learned a few best practices through trial and error (and seeing what resonates with customers): 1. Speak to how AI supercharges your overall narrative/platform value: often times, we get caught up in launching new products and thinking about what makes them unique/different but instead, I'd focus on how does it amplify your existing product value. AI is a catalyst in many respects, so it should further demonstrate your established product's offering vs. pull away from it. This can be done by doubling down on pain points, value props, and what you already offer today. 2. Drill into key use cases or common jobs to be done: for your established product, you've most likely already defined your "product market fit" and what customer use cases you uniquely support. To make it more relatable when you launch new products, you want to make it about the customer. Not you. This means speaking to their core use cases and how your AI solution further supports them. 3. Lean into "how-to" forums/content: I've personally been overwhelmed by all the possibilities of AI and when you stare at the blank screen with the cursor flashing from the chat bot, I don't always know where to start. Customers are the same way. Educating them via simple use cases and "how-to" content is integral to ensuring product adoption. It can be a simple video series, customer edu campaign, or something else—as long as it solves for the range of learning curves we're seeing & experiencing with AI.

There's a bunch out there but I've found the following most helpful: 1. Sharebird community, which is full of great insights & learings :) 2. Coursera AI for marketing 3. Udemy AI in marketing 4. Prompt engineering: there's a lot of them out there with different price points; I'd recommend starting with some of the free online tutorials before investing in a deeper dive curriculum (you'll be surprised how much you can do with just a few best practices)

Like a good loaf of sourdough bread, it only takes a few ingredients. But the ratios are key to getting it right. The same can be said about training AI bots for any sort of role or task you're giving it. It all starts with a clear, actionable prompt. I've learned through AI experts at Asana about a 4-sentence prompting framework that's really helped me see great results from building PMM-specific bots: 1. Persona: “You are an expert at _____” 2. Goal: “Your goal is to _____” 3. Task: “Your task is to _____” 4. Context: “Here is all the context you need” (think formatting expectations, things to avoid, things to highlight, etc.) Highly recommend trying this out, and you can learn more here.

There's no doubt that AI will transform our careers and starting now is important. I've found myself diving deep into a variety of trainings, boot camps, certifications, and research articles to help with my own learning curve. What it's shown me is this isn't a "one and done" motion but rather an on-going skill investment as the technology continues to evolve. Essential skills I'd recommend include: 1. Prompt engineering 2. AI 101 & 201 baseline certifications 3. Automation education (e.g. creating workflows with AI) 4. AI storytelling (e.g. messaging & positioning) 5. AI ethics & principles 6. AI for marketing applications I think of my AI learning cycle like going to the gym. It takes time, discipline, and constant trial & error to get things right. And before you know it, you might be looking like The Rock (wishful thinking for me, realistic thinking for you)!

There are so many AI writing bots out there, but I primarily use Poe and Claude with my own prompts built in. I've started experimenting with Jasper and Grammarly to just see what they can offer, but haven't tested them deeply just yet. What I've found is most helpful is having a really solid prompt. If you just put in a request without a lot of guidance, AI will spit out something that might not resonate. But if you give it more prompting or guidance, you'll see a night & day difference in the output. My recommendation would be focusing more on the prompt and you'll see great results from a variety of solutions.

I've recently been reading more in Gartner’s Hype Cycle for Emerging Technologies report that last year was all about the “Peak of Inflated Expectations” with the rise of LLMs and promise of Co-pilots—or as I like to call it, "the year of AI billboards on every city corner!" And now, we're in this "Trough of Disillusionment" where we're trying to find 1-2 clear applications of AI within organizations. But if we fast forward past those stages, you'll see we soon turn to a "Slope of Enlightenment" and ultimately, the "Plateau of Productivity" where core use cases & workflows are supercharged by AI. It's embedded in all the critical operations of a company. So why do I bring this up, besides to reference some cool names? Well, for many organizations we're still in the peaks & troughs. Still trying to identify a clear use case for AI applications within our companies. So I encourage you to think about your core Product Marketing workstreams, responsibilities, etc. and try testing a few applications of AI to see what sticks. You'll quickly find things that don't work but also a few powerful concepts that do. And from there, it can spark ideas across the company. While your company might not offer an AI solution for your end consumers, it doesn't mean you can't think about how to transform the business with the help of AI to better serve those same customers.

When it comes to measuring AI's impact for product marketing, I've found it really helps in several areas: 1. Message crafting & testing 2. Competitive & industry analysis 3. Predictive analytics on consumer trends/behaviors 4. Content writing support (e.g. press releases, keynotes, etc.) 5. Actioning various stages of a product launch workflow I'm also discovering more and more applications—whether that's helping synthesize key trends from customer calls or helping inform what we should think about launching next. What I'll do is block off an hour each week to think about the core things I'm working on and see where/how I can try injecting AI into those workstreams. It's more time upfront, but I've found it leads to the best results long-term.

I'm using AI to help with automating a lot of the rote, manual work that often bogs us down (e.g. writing status updates for XFN teams, assigning out work & next steps, taking a first pass at generating content, etc.). It's been a powerful addition to our product launch process where we have created workflows with AI embedded in various stages. For example when new launch requests come in, we have AI triage the request, validate it, assign it to the right team/individual, and provide a first pass at copy recommendations to give us a starting point. We've seen it supercharge our speed and quality of bringing new products to market. We also have an AI Council where we're experimenting with a variety of new solutions—almost like our think tank for early testing before we broadly make it available and enable our teams.

It's important to be transparent and consistent in addressing security & privacy concerns of enterprise customers. I've found that ~70% of their questions are the same from customer to customer but then the nuances of regional teams, specific data sharing protocols, etc. make up the remaining 30% of differences. In the past few months, we've found engaging with AI Councils within these enterprise organizations leads to more in-depth conversations around business use cases, ROI, AI application, etc. which lead to better results. It's less about filling out a questionnaire and more about showing up as an invested partner in your customer's AI journey. This can mean dedicated conversations with AI Councils, bringing your security/privacy/legal teams into strategic meetings, and using internal customer champions to advocate for the impact your solution can have on their organization.

Balancing the good and bad of AI is incredibly important, and something I think about regularly for what we build and launch at Asana. How that plays into our marketing plans really comes down to a few key themes: 1. Ethics for a "human-centered" approach: think about the ethics or principles for AI that you want your customers to know as they evaluate your AI solutions. Ex: Before we even built anything at Asana, we established a set of "human-centered" principles that grounds everything we do—from ensuring customers are always in the driver seat to the standards we hold our partners to, and much more. 2. Think global, act local: with any fast-paced innovations, it's easy to get caught up in a "one-size-fits-all" approach but 1) that's not scalable and 2) it will not help you achieve the goals you care about. After learning this the harder way early on in my career, I now have regular touchpoints with teams across the globe to understand what is & isn't resonating in different geos/markets. We're then able to apply an AI maturity model assessment to inform our campaigns & messaging so we're treating each region in a more targeted way. 3. Be transparent & upfront: I've found that the more we can share with customers upfront, the more their minds are at ease with what AI can do. Don't make them go hunting for the common information they're looking for. Being transparent & upfront will lead to more productive conversations, especially when it's coupled with a "tiger team" of experts who can help address data/privacy concerns. 4. Listen to your customers: I'm constantly talking to our customers to understand where they are in the "AI Hype Cycle" and how they're thinking about AI transformation for their business. As a member of our own AI Council, it's equally important to hear from employees on what are their fears, concerns, or learning curves—and all versions of that are okay! It helps us stay close to reality of adoption timelines and think about the solutions we're bringing to market for our customers. I'm also seeing from industry research that over 30% of knowledge workers are worried that AI will replace or displace their jobs. And while the concern is valid, the real opportunity with AI is that it can help amplify our human skills and give us more time & space to focus on the strategic work we were hired to do. That's at the core of our messaging: seeing AI as more of your teammate that can help you be more creative, strategic, and more intentional with your time. Transformation doesn't happen overnight but if we embrace it and use it to uplevel our skills, we'll see how it can create new opportunities for us all.

We've seen our technical teams really lean in to help with customer-facing documentation, fact sheets, resource guides, etc. Essentially, an enablement partner to PMM across a few areas: 1. Validating documentation: for many of our customers, understanding what data AI has access to, how it interacts with that data, and when/where is it shared with LLM partners is what we're often fielding. Our technical documentation team has helped us create very clear diagrams, wireframes, and fact sheets that break all of this down for a customer to easily self-serve/understand. 2. Supporting customer inquiries: our technical documentation team is comprised of product, security, privacy, and legal teams who we've built into a "tiger team" for strategic customer calls/inquiries where their expertise will help with the evaluation process. 3. How-to videos/public forums: this team also supports some of our campaigns where they're able to directly join a Q&A expert forum with customers to answer their questions, break things down, and just reassure audiences as they evaluate new technologies. This group is incredibly important to an organization, and often on the frontlines with PMM enabling our global sellers/field teams. An integral partnership for bringing AI products to market.

I really like this question, and worry there's not enough characters to do it justice here! But let me give it a shot by sharing my high-level thoughts based on the past 2 years of bringing new AI solutions to market: For me, this comes down to what's at the core of Product Marketing and what really drew me to the craft: storytelling. I love telling stories. And with the help of AI, I've seen how it can make Product Marketers even better storytellers (or in some cases, even worse if we're lazy about it). AI can help take a kernel of an idea, a good prompt with some thought-starters, or historical performance data and turn it into something relatable, targeted, and highly engaging. I recently read an article in a surf magazine (I promise not to make this about my personal hobbies) and felt so moved, so seen, so inspired by how it was written to only find out moments later...that it was written by AI! So, to answer your question about "how will AI shape product marketing?"...it really comes down to a few areas where I've already seen the most impact from my own experience: 1. Messaging: turning key points or themes and fleshing out into clear messaging & positioning 2. Content writing: demos, webinars, scripts, press releases, etc. 3. User research/testing: customer validation and sentiment analysis 4. Competitive analysis: cross-referencing messages against competitors 5. Strategic planning: asking AI for recommendations based on historical performance The use case where I've found AI most helpful in sharpening my PMM skills, is product launches. From helping build out a plan and executing various stages of a workflow to crafting ideas and testing new messages to support the launch. And I'm always looking to be inspired by other applications of AI for PMM! Whether that's from a community forum, like Sharebird, or surf magazine. :)