Michael Mignano

The through-line

Everything Mignano has built or backed comes out of one conviction: that creative work which used to require expensive tools, formal training and gatekeepers keeps collapsing into something anyone can do with a tap, and that the interesting companies sit exactly at the moment of collapse. He absorbed the framing at Aviary, whose mission was to democratize creativity through photo tools, watched Instagram prove it, then asked the same question of audio and got Anchor. "This was happening for videos photos text everything was being democratized except audio" [1]. He now applies the same lens to AI, and to his own new company, OBO: the industry spent billions making machines smart, and "we believe with OBO that it's time for the machines to make the humans smarter" [2].

The second half of the through-line is that the tool is only half the story. What made the App Store a revelation for him was not creation but distribution, and the same instinct runs through his read on podcasting today, where he is unsentimental about the open format he helped scale. He also keeps returning to the fact that none of this was planned: the career, the company, the product that worked were all found by iteration, which colours the advice he gives founders now.

On democratizing creativity

The lens came from Aviary, where he joined around 2012 during the company's shift from desktop to mobile, and where founder Avi Muchnick's product filter was that things should be simple enough for "a kid picking up a phone or an iPad for the first time or maybe your grandparent" to take, edit and share beautiful photos [3]. Mignano's mother was a professional photographer, so he had watched the old version of the work up close, and Instagram's arrival was the light-bulb: a photo that would have taken two hours in Photoshop now looks beautiful with one tap and reaches anyone in the world with another [3]. He describes that as the moment he understood he was living in a "creative revolution" where devices and software had reached the point of dramatically simplifying what used to be very hard [3].

Once you have the lens you cannot switch it off. At Adobe he and his co-founder got deep into podcasts and immediately asked "why is this so hard why does it take you know really complicated software and really expensive hardware" for someone just to speak and be heard [3]. He points to the same pattern elsewhere: website building is still mostly a desktop activity, which is why he angel invested in Universe, which puts it on the phone, and why webflow makes something just as powerful as hand-written code far simpler [3]. And he notes the loop closing on itself: at Aviary they used to talk about a product so smart you tap a button and it just knows what you want, and now you type a sentence into a prompt with products like stability and it appears [3].

On how Anchor actually found its market

Anchor was not born as a podcast tool. "The original vision for Anor actually was not to be a podcast platform it was it was very very idealistic we wanted to be a social media platform like audio social media" [2], essentially Clubhouse a couple of years early, because in 2016 "everything was a social network back then like Snap was loing and we're like oh of course this should be a social network right we should have a feed we should have a button right at the bottom of the app in the middle" [2]. The first build was a feed of two-minute audio "waves" you could reply to with your voice, made nights and weekends while they were still at Adobe [1]. His verdict on the whole category is blunt: "social audio is really really freaking hard" [2]. They gave one format a year, pivoted to a second iteration of social audio, gave that a year, and only then listened to what users had been telling them, which was that they wanted to publish podcasts to Apple and Spotify. "When we did that it was just like instant pmf from that moment" [2].

Two decisions inside that story matter to him. First, launching before it felt ready: the beta community was small and "kind of floundering to be honest," but after more than a year of building they concluded "we have to just get this thing out there we don't we don't know what we don't know," and the launch blew up [1][3]. Second, the willingness to keep going through non-linear years at all. He left Adobe in 2015 on the reasoning that "when else are we gonna be able to do this in our lives," raised a small amount from beta works, then a proper seed, and launched in 2016 [1][2]. In total Anchor raised just under fifteen million dollars and was about twenty-five people when Spotify acquired it in 2019 [2].

On platform moments

He is precise about when technology stops being interesting and starts being generative. Studying C, C++ and lisp in college stimulated him intellectually but never connected to what anyone could build or distribute at scale, because software still shipped in boxes from CompUSA and Best Buy [3]. The iPhone and, more importantly, the App Store changed that: software "went from being this thing that was very rigid this sort of boxed software distribution model to being kind of like a form of media," where a small team presses a button and reaches millions overnight, "not too dissimilar from music or film or TV shows" [1]. That reframing is what pulled him out of a record label job and into startups, alongside a cultural moment he thinks is underrated, The Social Network, which his whole department at Atlantic Records went to see together [1].

His current view is that we are overdue. "We haven't really had one since then we haven't had like a new platform since then that totally resets um or extends people's imaginations to new heights" [3]. People have nominated web3, the metaverse, VR and AR, and now AI, but he holds that it is impossible to know in advance which one lands, and that when it does it will inspire a new generation to chase their interests [3]. On AI education specifically he thinks the field is at "the App Store moment where there's a bunch of we need to build products around this stuff we have the new technology but you know what is the dominant product Factor product form factor for each of these uh use cases" [2].

On RSS, Spotify and the video pivot

Having built the infrastructure much of podcasting runs on, he is willing to say the open format may not survive. He acknowledges the constituency: "you have the podcast purists that will live and die by the RSS feed I mean they will fight you to the death" [2]. But RSS is not a two-way channel and is severely limited in the functionality it can support, and it is "probably no coincidence" that the two biggest podcast platforms, Spotify and YouTube, are no longer really RSS based and layer capabilities on top that have nothing to do with the format [2]. His conclusion is that it is the centralization versus decentralization argument again, "and in most cases I know people don't want to hear this but centralization often wins" [2].

Spotify's advantage, in his reading, was bundling rather than podcasting per se. Apple ran music and podcasts as two separate apps, which "was a total gift to Spotify," because Spotify could place podcasts one tab away from a built-in audience of hundreds of millions and never make you leave the experience [2]. He expects video to be the next thing bundled in the same way, starting with podcasts because Spotify already has every show in the world, then extending to other video content, where taking even a few market share points off YouTube's catalogue would be significant [2]. He speaks about this explicitly as an outsider who has not been at the company in three years [2]. For creators, he sees no path back to audio-only: covid pushed everyone onto Zoom and Riverside, which meant "all these podcasters got video for free," they uploaded it to YouTube, discovered the distribution, "and now you can't go back the the algo craves the video" [2].

On AI that makes humans smarter

OBO, co-founded with his former Anchor co-founder, who runs it as CEO while Mignano stays a full-time investor, starts from the premise that the personalized tutor use case for chat gbt is hyped precisely because chat gbt is not built for it [2]. His objection is about context and continuity: "it doesn't know everything you've ever learned and it doesn't know what your interests are and it doesn't know what your goals are," whether you are learning a language or some theory of finance [2]. He frames the wave of competing AI learning products as a knockout drag out fight that is a good sign, because fast iteration means consumers get tested, improved products and eventually a real business emerges [2].

On high calorie and low calorie ARR

His main warning about the current AI cycle is that revenue charts have stopped being evidence. He distinguishes "high calorie uh ARR and low calorie AR," arguing there is a lot of cheap ARR growth around, driven by the fact that AI products feel magical, so users hand over a credit card on the strength of the demo [2]. That says nothing about stickiness. Many of the explosive-growth companies of the past couple of years "are very very churny," full of trialists chasing one magic moment and never returning [2]. His advice to founders facing pressure to post ridiculous numbers is to hold onto the old metrics: "consistent user growth and retention and making sure like the retention corts flatten out they don't just like Drop to zero," because a zero-to-a-hundred-million business is phenomenal but "you can't ignore the traditional metrics that tell you whether or not a product is good" [2].

On where value accrues in AI, and how venture should adjust

His broader investing thesis is that the moment now favours the application layer, and that the foundation model labs, OpenAI and Anthropic among them, will not be the winners there [4]. He argues startups should be TokenMaxxing, and that venture investors should put less weight on price and ownership than they traditionally have in an age of AI [4].

On careers, and why his own path is not a template

He is openly uncomfortable with career advice because his own route was unplanned. He took computer science on the recommendation of roughly two people, including his sister's boyfriend, without really knowing what the degree was, because music seemed risky and CS seemed responsible [1][3]. He went straight into consulting at Accenture and "immediately was like oh my goodness I've made a horrible mistake," programming in a proprietary SAP language on a project whose purpose he still does not know [1]. The correction was to go work where his passion was, building artist websites at Atlantic Records, which let him use the degree somewhere he felt at home [1]. He notes that graduating seniors then faced no question about what they wanted from their life, only whether they had applied for fifty jobs, and that it took years of asking "what the hell am I doing with my life" to break the pattern [1].

He finds today's students unrecognizable by comparison, meeting nineteen-year-olds who already know they want to do early stage SAS investing between pre-seed and seed, when he did not know what venture capital was at graduation [1]. So when people ask how he got here, his honest answer is that the story is "not particularly relevant to your experience at all like what I did isn't going to work today at all" [1]. What he does endorse is the startup ethos he found at Aviary, an office with no amenities in a bad part of Manhattan where they were robbed three times, which he loved because "it felt like being in a band" [1][3].

Takeaways

• Social audio is extremely hard: Anchor spent two full years across two social audio iterations before user demand for a simple way to publish to Apple and Spotify produced instant product-market fit [2].
• Launch before you feel ready. Anchor shipped out of a floundering beta on the logic that "we don't know what we don't know," and it took off [1][3].
• Treat AI revenue skeptically: separate "high calorie" from "low calorie" ARR and judge companies on consistent user growth and retention cohorts that flatten rather than drop to zero [2].
• RSS is limited and one-way, the two dominant podcast platforms have already moved past it, and in these fights "centralization often wins" [2].
• Audio-only is a dead end for creators because covid gave podcasters free video, YouTube rewarded it, and the algorithm now requires it [2].
• Spotify's podcast success came from bundling podcasts one tab away from music while Apple split them across two apps; video is the next bundle [2].
• The App Store, not the internet itself, was the platform shift that mattered, because it turned software into media distributed with one button, and no comparable reset has happened since [1][3].
• Chat gbt is not built to be a tutor because it lacks memory of what you have learned, your interests and your goals, which is the gap OBO is built for [2].
• He argues the application layer, not OpenAI or Anthropic, is where AI value accrues, that startups should be TokenMaxxing, and that VCs should de-emphasize price and ownership [4].