What Makes a Consumer AI Product Stick? | Josh Elman

1. Key Themes

The Adoption Funnel: Intrigue, Adoption, Substitution, Spread, Retention

Josh Elman lays out an explicit five-part framework for evaluating consumer products, which doubles as his personal investment checklist. It moves beyond simple "growth" metrics to focus on genuine behavioral replacement. "Things that I look for in a product are where it has a really clear value that you can describe to somebody... The second is, when that person does go and try it, do they have a good experience?... I always like to ask people, what did you stop doing when you started using this new thing?" [00:00:22] He stresses that final-stage retention is the hardest and most valuable part: "It's never been easier to get consumers' attention... What's harder, I think, is to your point, getting me not just to try something, but to actually stick." [00:00:00]

Trust as the Next Moat in Consumer AI

As AI moves from utility tools to agents with access to money, calendars, and personal data, Elman argues trust — not just capability — becomes the deciding factor for mainstream adoption. "In these early days of adoption, utility often trumps trust... But if you truly want to break into the mainstream and really get to the hundreds of millions or billions, trust ends up becoming paramount... I think the things that will break into the mainstream will have a fundamental tenet of trust." [00:10:19]

"Earn the Right" — Narrow Wedge Before Horizontal Expansion

Elman's core company-building thesis is that products must dominate one narrow use case before expanding, using Discord and Siri as contrasting case studies. "If you want to become a very big company, you can't start by trying to do everything. But you have to start by doing a few things really, really well and earn the right to do more... Discord... earned the right to be the place that we played games and that we had a live chat during games. And as it earned that right so well, it became a place we chatted more." [00:00:22] By contrast, "Siri was actually amazing to earn the right to do a few things for you really well... But when you try to do more things with it, it often couldn't." [00:36:32]

The Rebalancing Post-COVID: From Screen Time Back to Real Life

Elman frames a multi-year arc: COVID pushed digital engagement to its ceiling, and the current wave of consumer tech (AI, live shopping, IRL social apps) is about pulling people back into the physical world using digital tools as connective tissue. "We saw so many companies, but I almost say they kind of hit a cap and we realized that we still need time for ourselves in the world... we've kind of seen this rebalance where a lot of things that happened, we're still spending time on. But now we want to be more in the world." [00:04:19]

AI as Social Glue, Not Social Replacement

Rather than AI replacing human connection, Elman envisions agents coordinating between people's individual AIs to reduce the social vulnerability of reaching out, and ultimately re-engineer how people spend free time together. "What if I could go to my agent and I'm like, can you just root around and see if anybody's up for something this weekend?... I think in some ways, it's actually AI being the glue that brings us together." [00:31:40] He extends this to a "crazy big" vision: "You just wake up every Saturday morning... 'I've got a rich day for you.'... your agent all of a sudden is just becoming your social calendar." [00:33:30]

Consumer Willingness-to-Pay Has Structurally Changed

Unlike the earlier internet era, Elman argues consumers are now primed to pay for digital value directly, which changes the monetization playbook for AI-native products even amid high inference costs. "With the rise of mobile gaming, the rise of subscriptions, consumers are much more used to paying for digital things than they were 10 or 15 years ago... When people hit a certain level of engagement with a product, asking them to pay isn't the same... hand to the face that it used to be. It's now very much an invitation." [00:43:54]

Clonability Is Not the Threat People Think It Is

In a market where products can be copied in days, Elman argues durable differentiation comes from vision, trajectory, and relationship/personality — not feature sets. "I can only clone what I see. I can't clone where it's going and I can't clone the next thing... So much of building an AI right now isn't just a product that does things, but that has personality, that has relationship, that has an identity that you actually stick to." [00:50:42]

The Coming Wave of Vertical, Agent-Powered Services

Rather than one super-agent, Elman expects a proliferation of specialized AI-native services across shopping, travel, health, finance, and home — each replicating the "trusted concierge" experience at scale. "The best experience is when you can have a concierge experience at a store... even better is when you know that their only incentive is your success... I'm really excited to find the companies that can sort of bring us that experience." [00:12:50]

Microdramas and AI-Native Entertainment as a New Creator Economy Layer

Elman sees short-form AI-assisted storytelling (building on the Musical.ly/TikTok lineage) as opening entertainment creation to people without studios, potentially spawning entirely new marketplaces for narrative content. "With AI, we can open up this creativity to so many more people, people who might not have a studio or a cast or sets to go film something... I think there might be incredible new marketplaces where people come together to see the creations from others." [00:21:29]

2. Contrarian Perspectives

No New Major Social Network Is Actually a Sign of Health, Not Failure

Against the common narrative that the lack of new breakout social networks in years signals stagnation, Elman argues incumbents already satisfy the core need, and most "new" attempts have simply been derivative. "We haven't needed to. Like, the social networks we have are at billions and billions of users... I also say, you know what? We have great ones. And maybe we're pretty satisfied." [00:26:04] He adds that most attempts failed because they weren't differentiated: "A lot of the attempts at social media between then and now were things that felt derivative... it's hard for those to catch and truly become mainstream." [00:27:02]

Attention Is Now Cheap; Retention Is the Only Scarce Resource

Contrary to the assumption that customer acquisition is the hard part of consumer tech, Elman flips it: distribution has essentially been solved by viral platforms, and the real bottleneck is now behavioral substitution. "I actually think it's never been easier to get consumers' attention... I've seen people go really, really fast just by getting in front of people. What's harder, I think, is... getting me not just to try something, but to actually stick." [00:48:27]

SF Is Not the Best Place to Start a Consumer Company — Only to Scale One

Despite being an a16z partner in the Bay Area, Elman pushes back on Silicon-Valley-centrism, arguing consumer culture is inherently local and SF's culture doesn't universally transfer. "I don't think it's the best place in the world to start a company, but I think it's a very important place in the world to scale a company... I also think if you are building for consumers, a lot of what you're also building for is culture. And I'm not sure that the SF tech culture is 100% transferable to everywhere else." [00:53:36] He also directly rebuts the idea that the best consumer apps are American-built: "I've got to think more if that's true. I mean, a company like Lovable came out of Sweden... Musically, for example, started actually as a team in China." [00:51:48]

Voice-Only Agents Are a Dead End

Despite being deeply associated with Siri and voice AI, Elman rejects the industry momentum toward voice-first/voice-only interfaces as the future paradigm. "I'm not yet a believer though that we're all going to move to voice for everything... I think the idea of a voice only agent doesn't really make sense." [00:38:37]

Group Chats May Already Be "The New Social Network" We're Looking For

Rather than a splashy new app, Elman suggests the successor to traditional social networks has already quietly emerged in a form nobody credits as such. "Group chats have become such a popular form of social media. Maybe that is in some ways the new network that we aren't thinking about that we really have developed over the past few years. I have way more local group chats than I probably had 10 years ago." [00:30:25]

3. Companies Identified

Whatnot — Live shopping marketplace. Cited as proof that live commerce, long assumed to be a China-only phenomenon, can work in the US/Europe through participation and storytelling rather than static transactions. "I've just been, I've been blown away by their execution over the number of years... it makes the shopping, when you make your purchase, you feel much more both confident, engaged, and have a connection to what you're buying." [00:19:47]

Discord — Gaming-first chat/voice platform, on whose board Elman sat. Used as the canonical example of "earning the right" to expand from a narrow wedge (game chat) into a broader social habit. "Discord was great for people who were trying to come together to play games and engage... it earned the right to be something that you relied on for more." [00:26:33]

Musical.ly (→ TikTok) — Short-form video app Elman backed at Greylock in 2015, later sold to ByteDance. Framed as the originator of the short-form-video-with-emotional-connection format that shaped modern entertainment and microdrama trends. "They had an idea that short form video... can create a whole different experience that you watch and you have a meaningful connection before you go on to the next one." [00:20:34]

Partiful — a16z portfolio event-planning app. Cited as an example of a network purpose-built to move people from online coordination back into real-world gatherings. "Partiful actually is the places you really want to be and it becomes a great way to go out and explore the world much more with the community that you have. In real life, though, it's super hard." [00:28:11]

Substack — a16z portfolio publishing platform, mentioned as part of the shifting landscape of where consumer attention now lives relative to legacy social feeds. [00:29:10]

Lovable — Swedish AI company cited as counter-evidence that great consumer/AI products don't have to originate in the US. "A company like Lovable came out of Sweden, for example, that's just an incredible company." [00:51:48]

Eleven Labs (ElevenLabs) — Voice AI company referenced as another non-US-originated standout. "All the things Eleven Lab does... didn't start here in San Francisco." [00:51:48]

Plot Party — Early-stage AI microdrama/storytelling startup that Ollie Forsyth's team collaborated with to produce a Nike-origin-story microdrama, illustrating the emerging creator tooling layer for AI video. "We collaborated with this company called Plot Party, which is like an early stage startup." [00:22:53]

Roblox — Referenced repeatedly as a durable phenomenon of shared, generative social gaming that grew through and beyond COVID, and as a template for the next wave of AI-generated shared worlds. "Roblox has been an incredible phenomenon that's continued growing... I think that opportunity is even bigger for so many more types of generated experiences." [00:04:19]

OpenClaw — Referenced as a landmark agentic coding/AI tool built essentially solo, credited with proving what long-running, memory-persistent agent loops could do and catalyzing the current agentic wave. "OpenClaw... that one person wrote by himself almost in his apartment and came up with a whole new way of interacting with models." [00:08:34]

Reid AI — Reid Hoffman's personal AI "double" project, cited as a model for how public figures/thought leaders can create interactive proxies of themselves. "One of the fun things I've seen him doing is building Reid AI... building something that has ingested and learned about him and creating sort of a double for him." [00:40:48]

LinkedIn, Facebook, Twitter/X, Robinhood, Apple — All named as companies where Elman worked in operating roles across his career, used to substantiate his operator credibility on scaling consumer products (Twitter user growth from ~10M to 100M+ actives during his tenure).

4. People Identified

Josh Elman — Guest; now a partner at a16z, previously an operator at LinkedIn, Facebook, Twitter (2009–2010 era), Robinhood, and most recently ~6.5 years at Apple working on Siri/Siri AI; earlier a VC at Greylock where he backed Musical.ly and Discord in 2015. Described as an "early joiner" rather than founder, specializing in scaling nascent products into durable platforms. "I was never a founder myself... but I was always what I call an early joiner, where I'd go find these things and say, we're going to go make something great." [00:55:21]

Evan Williams — Former Twitter CEO during Elman's tenure (founder of Blogger), credited with instilling a listen-but-hold-a-point-of-view leadership philosophy, exemplified by the retweet design decision. "The single best lesson that I remember from Evan was, hey, we've got a lot of smart people in the building... Let's listen to them. Let's learn. And then when we make decisions, let's make sure we engage with them on our thinking as well." [00:56:50]

Reid Hoffman — Referenced as a former close colleague of Elman's at Greylock and as an exemplary founder/thought leader, notable for building "Reid AI" as an interactive extension of himself. "Reid's, you know, one of the most incredible founders and thought leaders of our time." [00:40:48]

Michael Birch — Founder of Bebo, referenced by Ollie Forsyth as a recent podcast guest who was asked the same question about why no new social networks have emerged, positioning him as a veteran voice on social network cycles. [00:29:10]

Jack Dorsey — Referenced as a well-known figure from Twitter's history, though Elman clarifies he didn't work directly under Dorsey during his tenure (it was the Evan Williams era). [00:56:20]

5. Operating Insights

Diagnostic Question: "What Did You Stop Doing?"

Elman offers a specific, reusable diagnostic for evaluating whether a product is achieving true behavioral substitution versus superficial engagement — a sharper filter than typical engagement metrics. "I always like to ask people, what did you stop doing when you started using this new thing?... If you can get a good answer to that question, now you know it can intrigue people." [00:46:27]

Sequence Product Expansion Like Discord, Not Like a Feature Roadmap

The tactical implication of "earning the right" is a specific sequencing discipline: nail one behavior completely, let usage habits form organically, and only then layer in adjacent use cases — rather than shipping broad but shallow functionality upfront to appear more valuable. "Discord... slowly earned the right to be something that you relied on for more... even people who didn't play games came over to use the service to chat with friends who might also be using it to play games." [00:36:03]

Manage the "Confidence-to-Capability Gap" Explicitly

Drawing from Siri's failure mode, Elman highlights an underrated tactical lever: products should be judged not just on what they can do, but on how gracefully they communicate what they can't do yet — this materially affects whether users grant the product "one more chance." "It can explain to you that it can't do it and really understand that as opposed to just like put up a hand and say, sorry, I can't help you." [00:37:01]

Price to the Inference-Cost Curve, Not to Historical Consumer Norms

Elman implies a concrete operating framework for AI-native founders: model your subsidy of inference costs against the trajectory of falling model costs, rather than pricing based on old SaaS/freemium playbooks, since the "crossover point" between cost and willingness-to-pay is time-bound and shifting fast. "I think over the next couple of years, we're going to intersect with how much cheaper these can get to build and serve, with how much customers are willing to pay for things that are of value." [00:44:53]

6. Overlooked Insights

The Retweet Design Decision Encodes a Durable Product Philosophy Most Miss

The story of Twitter's retweet is told almost as a nostalgic aside, but it actually contains a generalizable and non-obvious product principle rarely articulated this clearly: when building virality/sharing mechanics, preserving the original creator's authenticity over the sharer's social capital is what makes amplification mechanics durable rather than exploitative. This is directly relevant to today's AI content and microdrama platforms, where attribution design choices will determine creator trust. "Everyone wants to retweet and share voices from others, but decided Twitter that the authenticity of the original poster was more important than switching the words to look like they came from your name." [00:57:18] This principle — protect the originator, not the amplifier — is a quietly load-bearing idea for anyone building creator marketplaces in the AI content wave Elman spends most of the episode discussing.

Group Chats as an Unrecognized, Already-Massive "Social Network"

Buried in a broader riff on AI and social products, Elman casually suggests that the industry has been searching for "the next social network" while missing that it already exists in an unbundled, unmonetized, and largely un-instrumented form: private group chats. This is a significant, underexplored investment thesis — nobody has built infrastructure/monetization/AI-augmentation specifically for the group chat layer despite its evident scale in people's lives, and Elman flags AI's entry into group chats ("they're talking to each other") as a potential unlock nobody is yet building for directly. "Maybe that is in some ways the new network that we aren't thinking about that we really have developed over the past few years." [00:30:25]