AMA: Google Product Marketing Lead, Vishal Naik on Product Launches

I've 100% found the failed launches in my career to be more impactful towards my learning and development as any successful launch. Two jump out at me as ones that didnt go according to plan. The first, more recently, was that I launched image generation in Gemini. We had some really strong usage, but we also ran into a scenario where our product wasn't delivering on the product principles we have as a responsible AI company. So we paused the usage of specific components of the product. What I learned here is that when you're building a consumer product, you need to proactively seek out the opinions of users who may not be your target or your hero user. Because in a consumer product, every voice is important. The other, which is a foundational case study in how i think about product marketing to this day, was the HP TouchPad launch. I worked at a reseller, and we were very strong at selling iPads, we knew what customers wanted in tablets, and we knew the market needed a second player. (mind you, this was many years ago and Android tablets hadnt taken off yet). So we were very bullish in our projections with HP, and they had a strategy that also wanted to ensure the market didnt see the TouchPad as inferior to the iPad, so it was priced similar to the iPad with similar features. However, what we all overlooked was that the iPad was already out as a market incumbent. The TouchPad was slated to come out when the iPad 2 was coming out, so the comparison wasnt to the iPad 1 at its launch (where TouchPad had favorable comps), it was compared to the iPad 2 at that time (where TouchPad had negative comps). And so it was a really really strong lesson in not drinking the kool-aid. You need to approach launches with a healthy dose of optimism for the launch potential, but also a realistic view of how the end-user will react to your positioning/pricing/product strategy based on present market dynamics. I had to read this book in college called Dissent, Injustice and the Meanings of America and one of the main takeaways I had from that book is that dissent is a good thing. Not blind disagreement, but healthy debate about the topic at hand and bringing a "devils advocate" point of view that forces you to see past the hype engine and look at true product market fit. The TouchPad experience proved that lesson and I still think about the learnings from that launch, probably monthly at this point, and that original launch was in 2011.

No launch is the same. A Bill of Materials checklist is nice, but no launch should ever map 1:1 with your checklist. Use it as a starting point, but flex items on or off the checklist based on what your users need out of that feature.

in my opinion, each launch has its own nuance. Not every product goes to the same audience (there are likely subsets of your user base that are more or less excited by some launches than others), and so when youre looking for how to deliver value to your new launch, there is likely a part of the launch plan that you turn on for one launch that doesnt play to another. We have a framework for launches where the list of tasks a PMM can do gets broken down into four components: user insights/opportunity sizing -> positioning and val prop creation -> GTM -> growth. and in each of those four compnents theres a list of tasks from reasearch to messaging frameworks to naming strategies to blogs to social to analyst engagement and press briefings and so on. So I'd suggest you think less about a routine bill of materials and think more about a line card of marketing deliverables across the different stages of a launch, and rather than your PMM checklist being a set ~10 activities, come up with your list of 20-30 options and then pick which 10 you need to deploy for each subsequent launch.

I'm honestly not super bullish on trying to find the new flashy marketing vehicle. I interviewed at a company once who stated they were looking for some out of the box, early salesforce-esque guerrilla marketing tactics to drive their launches; but at the same time admitted they also struggled at following up with leads that they generated at events. You'll probably get the value you're seeking out of your marketing work by focusing on the tried and true promotional vehicles, but doing them really well. For example, people may hate on emails these days (its spam, emails get auto filtered, etc.), but in one of our recent launches for Gemini, there was a direct correlation with the time in which our email was sent and a spike in usage so high that it caused us to need to reallocate compute capacity so that the product didnt crash.

My opinion is that if its in the UI, it needs to be very intuitive. So if you cant drive usage/awareness with a single in-product notification, then the UI needs some refinement to make it clearer to users. Of course, you can still educate users in other areas, blogs, help center articles, etc. but I would consider the main driver being an in-product notification.

In my opinion, you need clearer goals of who is responsible for what, before you go into a launch. For example, our team specifically focuses launches on marketing driven trials of a product. But product retention is on Product to build a product that is good enough to keep users. So similarly, you'll want to look at what your business goal is (assuming its a revenue number) and then map across the journey to get to that number what teams come into play to hit that goal. Example: Marketing is accountable for a leads quota or a specific conversion rate from first meeting to second meeting, etc. Other departments also need to own their own metrics (product retention, sales conversion, forecasting, etc.). In terms of validating forecasts, you need to understand where the numbers are coming from. For that, I'd suggest you get into the weeds on other sales opportunities. What numbers attributed to the forecast itself?Where did customers drop in the journey? etc. My guess is you'll probably find a range of reasons like where pricing was off, where one sales rep didnt deliver the pitch as well as another one does, where marketing overpromised but the product couldnt deliver, where the product just wasnt good enough to keep users, and so on. As you explore that, youll start to get a sense of what's reliable vs what isnt.

cross promoting the playbook for work my team ran on this at a previous company! https://sharebird.com/h/product-marketing/playbooks/how-docusign-increased-their-developer-marketing-conversion-rates-by-50

I worked at a startup who took pride in the fact that we launched new features with every sprint. It sounded like innovation to talk about how we released X new features that year, but the reality was that customers didnt know what was a thing they needed to care about vs what was just noise. Because there's no customer that NEEDs every feature that's launched. I also worked at a large org that had three set launches a year. It felt like the magic of a specific product could easily get lost because it was abstracted into a large bundle. So there's no set approach that will work for every company. I'd suggest you think about what the best way to resonate with your user base is and then flex your timing. There's nothing wrong with a feature that goes out when ready, but isnt marketed heavily until the bundle. There's also nothing wrong with launching something in a bundle but leveraging the bundle as the launch vehicle where you also follow up post launch with an independent set of deliverables. I'd recommend you consider a tiering approach where you have something (roughly) like a tier 4 launch is silent, and it just launches when ready and theres no comms around it. A tier 3 launch rolls out when ready, and maybe you do an in-product notification around it at launch and then its included in a blog or bundled webinar at launch. A tier 2 launch GA's at the time of the bundle and most of the bill of materials falls into the bundle activities. A tier 1 launch is announced (or GAs) at the time of the bundle, but has a continued body of work to drive awareness and interest in the feature outside of the launch bundle. This way you keep the value of the bundle, limit noise to your users but also dont water down hero features.

By the numbers. Understand your audience for each channel and where your target demo for that product is going to engage, and then show up in those areas. For example, if youre a B2B product and you're targeting end-users within your customer base, a sales pitch isnt going to get the message across--you'll also need targeted ads or emails. If youre a B2C product and youre targeting Gen Z, you might lean away from facebook in favor of tiktok (or vice versa if youre targeting Milennials) and so on. To simplify: 1) understand your product's target demo 2) understand what channel mix you have available 3) measure performance of historical usage of those channels, specifically by your target demo 4) choose your channel mix based on a balance of impact of the channel and time you have to create that content 5) optimize messaging per channel and run your programs

Both are viable goals, just depends on what's better for your business. focusing on current customer adoption can make you a stickier product within your install base and thus harder to replace. New customer adoption can give you a deeper pipeline for who to follow up with to cross sell your other products.

They both require a clear user story. For a product you've acquired, they probably have existing messaging that you then need to nuance into a cohesive customer story. For a product you've built, you probably are messaging it in line with how youve always marketed and what customer expect to hear from you. But beyond that, customers wont care if you bought or built a product, just that it works.

It's harder to measure usage if the product is launched before its available. So if product readiness doesnt map to launch dates, you'll need to consider follow up touchpoints for when the product is actually available. Otherwise you risk the launch being ignored.

Different teams might own different channels. I tend to think that I know enough that I could probably write a good email or create a good social strategy for a launch, but im not an expert in those areas--and we have people in parallel marketing teams who are experts there. So for my launches, we have weekly XFN GTM syncs where those stakeholders are included and we walk through where we are for launch and look for status updates on key deliverables like emails, data sheets, help center articles, etc. My goal out of these meetings is to ensure that my marketing stakeholders have enough to do their job well, so as as the expert of a channel, they can deliver on their goal of driving a great asset and I can deliver on my goal of ensuring the launch goes as planned.

I work at a company where there's a specific challenge where there's an entire press ecosystem around scooping new news. Thus we tend to make announcements and run launches when a staged rollout starts. The opposite side of this equation is you dont want to announce something and then have customers asking where the product is. Thus I tend to look at it as mitigating risk. Imagine you log into Amazon and there is a new UI on the checkout page and you dont see any news around it--you might abandon your cart thinking its a security breach / phishing attempt / etc. Here, not announcing at the start of the rollout might degrade brand trust. But if you're not at the scale of an Amazon or Instagram, it's probably more optimal to make your announcements when all users can act upon that announcement and try the product (at 100% rollout), because you're minimizing customer service tickets or customer complaints about why they don't have access to the new feature you just announced.

You have to articulate the cost. And specifically the cost to the business, not just the cost to your department. For the most part, I cant say I've worked for many leaders that would value the human element (wasted time on your part) over the business value to a last minute shift to a launch (potentially because of alignment to another launch/bundle or product readiness, etc.). While there is an opportunity cost in terms of what you cant get to because youre still focused on the launch that could have been out the door, realistically speaking, you might be met with a response of .. "well its short term, you'll just have to do extra for this window of time". So focus on the opportunity cost to that it creates to the product -- we cant develop the next feature because we're still working on this one, our time to revenue generation is longer because we dont have this feature to help close deals, etc. We had a launch recently where the launch date moved 6 times within 2 months. The cost in this case was that because of other features rolling out in parallel, it was a lot of wasted eng effort to keep the product ready. What i mean by that is that our process at that time was that if we had two teams launching new features at the same time, each launch team was responsible to not break features already live in production, but features that were also in the queue to be launched were the responsibility of that individual launch team. So because our launch date shifted, we couldnt ship the fast follow features we had in the queue because of the tech debt created by not letting the team launch the feature at the original date, and as such shifting the responsibility for product maintenance to the other teams launching other new features.

One tactic that I like to employ when I dont have buy-in yet from XFN partners is to pilot something with no cost to the other team. My guess is that your stakeholders dont yet grok the value youre proposing. So in this instance, dont force your stakeholders to adhere to your launch plan. But create one anyway and measure its effectiveness. So let your team ship items when ready, but in parallel create a marketing moment for after launch. Then measure key metrics -- leads generated, product trials, customer counts, active usage, etc. and if you can showcase that when you create a marketing moment you can drive more product value to the organization, then you can start leading the conversation around the business impact created/lost by not committing to dates and becoming a more mature product development function.