Adam Wiggins avatar

Muse retrospective

Adam Wiggins // Feb 2024

Between 2019 and 2023 I worked on Muse, a canvas-based thinking tool for iPad and Mac. We raised $2M in funding, grew the team to seven, had tens of thousands of active users, and thousands of customers. Despite that, we never reached sufficient organic growth or found the right business model for financial viability. I stepped away along with majority of the team in fall of 2023, leaving the business in the hands of one of my former colleagues.

If you followed the Muse story and were curious about what happened inside, this is my perspective on that. Or if you’re building productivity software or a creative tool, maybe you can find a few useful lessons here.


Muse began life as part of a human-computer interaction lab called Ink & Switch. The lab's charter involved inventing a new computing environment that would better serve people doing important knowledge work: scientists, writers, entrepreneurs, designers, architects, and so on.

Mark McGranghan, Julia Roggatz, and I were working together in the lab and started to see a confluence of several tracks of research that we thought could become a commercial product. This included input devices and form factors (tablet + touchscreen + stylus), infinite canvas as a document type, and how digital tools can be more useful at the ideation stage of a project (which typically gets done with analog tools like whiteboards, sketchbooks, and post-its).

Julia had created a fantastic prototype that ran on iPad, and Mark had been the product visionary (and also came up with the name “Muse”). I was ready to transition from running a research lab to working on a commercial product and joined them to create a commercial spinout.

So in summer of 2019 we created Muse Software, Inc. Ink & Switch gave us the existing IP (name, product design, prototype codebase) in exchange for a stake in the new company, and we were off.

Research prototype → product

We quickly discovered that things that had tested well in the lab were simply too hard to explain to new users. For example, holding the stylus at a particular angle to activate different tools. Most of our early onboarding (done either in person or over videochat) failed in the sense that people did not go on to use the product.

Shortly after we started the company we were lucky enough to land Lennart Ziburski as a product designer. He immediately started finding ways to bring Muse closer to the status quo of what people expect from iPad apps, and before long we had our first handful of regular users.

My background is in product development (e.g. design, engineering, user research), but since we had such a strong product team already I decided to make my focus for this venture be storytelling. I challenged myself to figure out how to explain a novel product, build a brand, and generally get the word out to people we hoped to have as users and customers.

My first real try at this was starting an email newsletter. At the time this was a non-obvious idea, but the timing was perfect: independent writers, Substack, and the newsletter craze in general was just getting started. The early editions of our newsletter were a success at helping tell a larger story, test some ideas (including pricing), and generally keep folks on our waitlist engaged while we tried to get the product in shape.

Company vision

For the company and team, we sought a somewhat unusual structure that blended elements from startups, indie businesses, and professional partnerships (such as legal or accounting firms).

Our “small giants” approach optimized for mojo over growth. We wanted a small, talent-dense team with a focus on craft, autonomy, and quality of life for all team members. We banished the term “founder” in favor of “partner” and tried to be transparent with all business matters across the team.

We took some capital from investors in order to invest heavily in technology, design, and brand—just as a startup would. But we started charging a prosumer price early on, intending to avoid further funding rounds, and instead transition to being revenue-supported.

Pre-1.0 doubts

We quietly moved from TestFlight to the App Store in early 2020, and activated a subscription paywall once a user reached a data threshold. Our team was nervous about this: as craftspeople it’s easy to feel like its not ready, see all the flaws, etc. But to our delight, we pushed the paywall live and several people purchased a $100 annual subscription right away.

But by mid-2020 I was having some doubts about the viability of the company. We had a few hundred active users—some incredibly active, using the app for hours a day to do important work like evaluating investment decisions or developing a master thesis. Of those, we had a very high conversion rate, with already 50 or so customers. But we didn’t have the growth we needed, and it wasn’t clear where it would come from. I thought a lot about whether the experiment of this venture was over almost before it began.

We were also struggling to find good messaging/positioning for our homepage. New users often wanted to slot us into “note taking” or “sketchbook/whiteboarding” neither of which were a great fit for the problem we wanted to solve.

The first three iterations of

I loved “thinking canvas” but that phrase didn’t strike a chord with users. Based on early feedback, we tried a message of helping the iPad reach its potential as a creative tool.

The opportunity we were going after, one I think is still mearningful and unsolved by any product today, is to bring the earliest stages of unstructured thinking and project ideation onto the computer. Almost every other part of our information lives have long-since been digitized: accounting into spreadsheets, presentations into slide software, page layout into graphic design software, etc. But when it comes time for an individual or a team to sit down and sketch out the beginnings of a new business, a new book, a new piece of art—this almost never happens at a computer. Or if it does, it’s a cobbled-together collection of tools like Google Docs and Zoom which aren’t really made for this critical part of the creative lifecycle.

But a long-winded description of the problem doesn’t cut it for something like a marketing website, which needs to sum up the problem and the value prop in a couple sentences. I found that, for any individual person, I could sit down and ask them a few questions that would quickly lead them to understand what we were doing. Usually that conversation would go like this:

Me: What do you do when thinking about ideas for a new project?

Them: Well, I sketch some stuff in my notebook. Then I take a photo of those with my phone and upload them to Evernote. I hate Evernote, but anyway. From there, to share with my collaborator, I copy-paste some screenshots into iMessage/Slack/Google Docs, but that never works so well because … oh. All of this is what you’re trying to make better?

Me: Yep.

Them: I need that. When can I buy it?

A related problem is that we had yet to target a narrow market, like just writers, or just designers, or just architects. In the research phase we had purposely kept it generic, e.g. it could just as well be used by someone working on a book as it could be someone designing a building. This seemed to make sense because analog ideation tools are generic: there’s no whiteboard for authors that’s different from a whiteboards for architects.

But if I’m honest it’s also because I loved the diversity of our userbase. Even at this early stage, we had lawyers, doctors, fiction authors, journalists, investors, game designers, and many more using our product. I had assumed that, at one point, we’d have a segment that would emerge as the most numerous, most likely to buy, or one we thought we could best serve and then we’d find a way to focus on that. But that never happened and we never forced it, which was probably a long-term problem.

Launch rocketship

Despite our weak growth and my concerns about the messaging problems, we decided to push forward with the 1.0 launch. This would be a pure marketing launch: the product was already stable enough for real use and available for purchase in the App Store. But we needed to break out of our small circle of users who had discovered us organically.

I churned through lots of messaging ideas and was generally gnashing my teeth about this. I ultimately answered the question by coming up with a bunch of weird ideas and split-testing them on our existing website to see which converted. A clear winner emerged from among these: calling ourselves a “tool for thought."

I was surprised, since this was a niche term from the dusty archives of the 1970s and 1980s HCI research world. Little did I know that the term was about to explode in popularity, largely driven by the breakout success of Roam Research starting to build steam about this time. But I was delighted, because “thinking tool” or “thinking canvas” was how I personally always thought of Muse, I just didn’t think that would play with a wider audience.

In August of 2020 we launched on Product Hunt and Twitter. Apple featured us prominently in the App Store. To my delight and surprise, the launch was a huge success. We got mentions from high-profile tech thinkers, including Benedict Evans’ newsletter, the economics blog Marginal Revolution, and a one-word tweet from Patrick Collison.

Website traffic for the first three years of Muse.

Note that the 1.0 launch bump produced a “new normal” baseline traffic that was 4x where we were before. The 2.0 bump is only slightly bigger and didn’t adust the baseline afterward.

With buzz building nicely and growth taking off, we had inbound interest from some angel investors whom I respected. So we took another small funding round (about $600k).

Doubts set aside, we settled into a rhythm of shipping great features to our growing audience.

Distribution channels

You’ve probably seen the meme about product distribution, and I went into this venture knowing that productivity software is particularly difficult to market. Everyone needs software for email, calendar, todo, notes, etc. But your computer comes with pretty good defaults, and there's no specific market (in the Crossing the Chasm sense of people with shared identity that talk to one another) for todo software. There's no "Todo Apps Monthly" magazine or trade show. Hence these markets tend to be fragmented, with a zillion niche players, and no clear path to building a big business.

I generally like the approach outlined in Traction. Briefly: any given business will find a small number of highly-effective channels, and the rest don't matter. For Heroku, that was attending developer conferences and getting blog posts on Hacker News. For another business it might be YouTube influencer sponsorships and print ads in a niche magazine. So I set about systematically testing many channels.

Some of the things we tried that had mixed success:

The marketing/distribution angles that worked best were:

Podcast monthly listeners, estimated by unique IP downloads.

High-profile guests (some of which are labeled here) created spikes in new listeners. But after I think many stayed as regular listeners thanks to ongoing high-quality episodes.

Overall, I can’t point to any one thing that we did that worked: but rather a succession of unlocks which would work for a time, then level off or decline.

The golden age

2021 was the best year for Muse. A second engineer, Adam Wulf, had joined our team the previous year and now our team of five was cranking out product improvements at an impressive rate. The “tools for thought” trend was in full swing, and being described as note-taking or sketching app was in our review mirror. We had a good amount of runway and a small but steadily-growing revenue stream. The podcast was cranking out high-quality episodes, and we were consistently in the top 200 podcasts list in the Technology category of Apple Podcasts. Team morale was at an all-time high and I was loving every day of my work. as it looked for the 1.0, 2.0, and 3.0 releases of the product.

We launched new pricing which seemed to unlock the growth we needed to be on a track to sustainability. But we were often running up against a problem: many people followed our work via Twitter and the podcast but would say: "Love what you're doing, but I don't use an iPad."

It had always been our intention to bring Muse to other devices, but it’s clear that desktop computers is where most professionals do their work. Even if you own an iPad, if the bulk of your serious work happens elsewhere, there’s a gravitational pull away from Muse on iPad despite your best intentions.

But a Mac app alone wouldn’t be enough. We needed your Muse boards to be available on both devices. After extensive experiments with iCloud (slow, unreliable, impossible to debug) and Firebase (better but not really suitable for the large data sizes our best customers had) we decided to import another piece of bleeding-edge research technology from Ink & Switch. Namely: local-first sync with CRDTs.

Mark had previously worked on a research project which implemented this, so he ended up building a from-scratch system over the course of a year. Wulf filled in the client-side implementation in Swift and ported the app from Core Data to the new persistence layer.

After having tried to do sync with many other technologies, this bespoke system felt magical. You could draw on your iPad and see it instantly on your Mac. But you could also disable your network and continue working as long as you wanted, and reconnect to have all the changes merge back together.

The Mac app was a huge effort by Julia and Lennart. The challenge of adapting our tablet-native canvas to a keyboard/mouse/non-touchscreen was substantial but I think they did an incredible job. And Catalyst is an excellent technology for using the same codebase for both platforms.

At the end of 2021, all our metrics were up and to the right. We tweeted a teaser shot of our icon in the Mac dock which got everyone really excited. It felt like the addition of the Mac app would push us over the edge into being a serious tool for work, open us to new audiences, and juice our revenue to bring us to financial sustainability.

The crash

In the first quarter of 2022, something strange happened that I still can’t fully explain. Our active users dropped by almost half over the course of a month or two, and all our other metrics (revenue, podcast listeners, etc) seemed to level off or even decline.

Weekly active users.

To this date it remains a mystery why we had solid growth for nearly two years after the initial launch, followed by a decline only briefly interrupted by our 2.0 launch.

Our 2.0 release wasn’t out quite yet and people were getting a little annoyed/impatient, since we had implied it would be available early in the year, and in the end we didn’t launch until late May. But after speaking to some other tools developers, who were seeing similar things, I think this was more a function of the macro economy. The world was entering a period of tech market decline, rapid inflation, and high interest rates.

I’ve built businesses in time of financial recession: I started my first tech company in 2000, and Heroku was just getting started when the 2008 financial crisis occurred. But in both of those cases, I think the underlying business was more solid.

So while I could try to handwave away my failure as a CEO (“something something Ukraine war something something inflation”), I won’t do that. Instead I’ll ask the question: was Muse a zero-interest rate phenomenon?

Briefly, I think: yes. Or to put it less harshly, Muse as a brand and a product represented something aspirational. People want to be deeper thinkers, to be more strategic, and to use cool, status-quo challenging software made by small passionate teams. These kinds of aspirations are easier to indulge in times of plenty. But once you're getting laid off from your high-paying tech job, or struggling to raise your next financing round, or scrambling to protect your kids' college fund from runaway inflation and uncertain markets... I guess you don't have time to be excited about cool demos on Twitter and thoughtful podcasts on product design.

I’d speculate that another factor is the half-life of cool new productivity software. Evernote, Slack, Notion, Roam, Craft, and many others seem to get pretty far on community excitement for their first few years. After that, I think you have to be left with software that serves a deep and hard-to-replace purpose in people’s lives. Muse got there for a few thousand people, but the economics of prosumer software means that just isn’t enough. You need tens of thousands, hundreds of thousands, to make the cost of development sustainable.

2.0 launch and B2B pivot

So with this looming problem showing itself in our metrics, we pushed on through to a launch of Muse 2.0 in summer of 2022. It did pretty well, largely thanks to all the love and support our many friends showed us on Twitter, for which I’m eternally grateful.

It did make a difference for our numbers: our revenue doubled, from around 60k ARR to around 120k ARR, shortly after the launch. But the absolute numbers were just still too small. 120k ARR is a one-person business, not a team effort.

Revenue annual run rate (ARR) over time.

In the summer of 2022 I did some soul-searching, as did the rest of the team. We knew our dream of reaching sustainability with our current model wasn’t going to work. But for me, I wasn’t done. We had such incredible assets: the team, the brand, the beautifully-crafted app, the groundbreaking sync system, the podcast. Maybe we just needed a new business model.

It had always been our plan to create a multiplayer version of the app, once we had conquered single-player and multi-device. The sync engine was already built and now battle-tested on 100k+ devices, many syncing very large datasets. We already knew we wanted to move into collaboration—so why not pair it with a well-known, lucrative, venture-capable business model? Namely, B2B SaaS.

So as we convened for a team summit in August 2022, I laid everything out. Everyone agreed: we had more to do here, and keeping all the good stuff (team/product/brand/etc) but jettisoning the thing that wasn’t working (business model and financing approach) made sense.

My experience with fundraising from venture capitalists was brief, a bit painful, but also enlightening and inspiring. I had spent the last decade somewhat trying to escape the gravitational pull of Silicon Valley and venture-backed hypergrowth startups. Reconnecting to that world I discovered… a lot of really lovely people, with smart perspectives on business, founder-friendly mindsets, and general integrity and goodwill! This made me feel encouraged to plug into the startup machine a bit more than we had so far.

But not surprisingly, we struggled to raise from an institutional investor. Because our B2B product was a barely-working prototype at that point, we were effectively a pre-seed “still just an idea” type company. But we had several small rounds of financing and a product already in the market. Those things were a liability, they argued, since they weren’t directly relevant to our B2B move. Also, summer of 2022 was probably the worst time since 2008 to try to raise venture funding, so we were playing on hard mode anyway.

In the end we raised from a dozen or so goodhearted angels and scout funds, plus I put in some of my own money. It was intended to be a bridge round: enough to let us validate the market and product just a tiny bit, and then use that initial traction to raise a real round.

It didn’t work

Going into 2023, we conducted a survey to see if our existing audience contained people that would introduce us to use on their teams. The signal here looked really good!

We had hundreds of impassioned answers from people working at big and small companies talking about the burning pain point of their group ideation situation now that their team had gone permanently remote.

One team lead described to me how they would fly out new employees for orientation, put them in a hotel, etc all so that they could spend a couple of hours drawing on a whiteboard together. If they had to buy those new employees iPads and a Muse subscription they would happily do that as much cheaper and less logistically-difficult than their current solution.

I took numerous customer-discovery calls in the same vein. Over and over, team leads and CEOs described to me how they had tried everything out there (including the obvious direct competitors of Miro, FigJam, Freeform, and Whimsical) and none of them solved the problem long-term. They were acutely aware of how this was hurting their team’s ability to creatively solve problems together, but didn’t see going back to the office as a viable solution.

We envisioned Muse as the perfect combination of the freeform elements of a whiteboard, the structured text-heavy style of Notion or Google Docs, and the sense of place you get from a “virtual office” ala group chat. As a way to asynchronously trade ideas and inspiration, sketch out project ideas, and explore possibilities, the multiplayer Muse experience is, in my honest opinion, unparalleled for small creative teams working remotely. Here’s a 3-minute demo of the early beta.

A screenshot of the completed multiplayer canvas. A true thing of beauty, in my opinion.

We onboarded a couple of dozen teams, and some of them went on to integrate it into their daily work. We had a few quotes from folks describing how their work was forever changed to have this series of nested whiteboards were they could gather ideas, have a shared artifact during a meeting, easily reference older material, sketch together, etc.

But friction began almost immediately. The team lead or organizer was usually the one bringing Muse to the team, and they were already a fan of its approach. But the other team members are generally a little annoyed to have to learn any new tool, and Muse’s steeper learning curve only made that worse. Those team members would push the problem back to the team lead, treating them as customer support (rather than contacting us directly for help). The team lead often felt like too much of the burden of pushing Muse adoption was on their shoulders.

This was in addition to the obvious product gaps, like: no support for the web or Windows; minimal or no integration with other key tools like Notion and Google Docs; and no permissions or support for multiple workspaces. Had we raised $10M back during the cash party of 2020–2021, we could have hired the 15+ person team that would have been necessary to build all of that. But with only seven people (we had added two more people to the team in 2021–2022), it just wasn’t feasible.

By summer of 2023 I had started to accept that this new path, while it held a lot of promise, probably just wasn’t going to be doable by our particular team. We had a difficult all-hands where I laid out everything in our strategic landscape. We debated and brainstormed. And ultimately we decided together: we had reached a dead end in this particular idea maze.

Could we have managed to scrape together more money and keep going? Yes, I think so. But for me the question wasn’t whether we could, but whether we should. The team we had assembled was one of the best I’ve ever been a part of. Their creations inspired me every day. I had become aware of the contrast between the level of execution and the apparent market opportunity. It was my feeling that everyone here, including me, deserved to be putting their efforts toward a stronger opportunity.

A graceful end

It was important to me that we ensure continuity to our customers, and facilitate a graceful exit for our investors and team. I know of so many startups that ran themselves right up to a zero bank balance, then had to lay off the entire team without warning. And post the dreaded “Sorry, you have 30 days to export your data” to their website before shutting down their product.

I’m interested in software longevity and since we had built our app local-first, it could hypothetically run for a long time without a company to support it. We sketched out various ideas for open-sourcing the app, documenting the data format, etc. But in the end the solution came from a surprising place: one of our team members, Adam Wulf, who felt there was a lot more life left in this business if we reframed it from a team effort to a solo indie developer.

I had some conversations with potential acquirers, but I knew these were likely talent acquisitions that wouldn’t make much difference to our investors, and would basically just be getting jobs at a new company for everyone. There weren’t many companies that we would have all been excited to work at. For the ones that were, we quickly discovered that the tech shakeout going on had many teams in our exact position. The market was suddenly full of available talent, so demand was weak even for a well-respected team like ours.

So with our team and investors agreed on a plan to turn things over to Wulf, I went about the painful process of crafting a goodbye letter and subsequent podcast episode. This was difficult to do, both in striking the right tone, but also what felt like “launching” the end of some of our ambitions and my failure as a CEO.

But I’m really glad we did it this way. A different approach would have been to do some blog post that buried the bad news in the middle or tried to hide or gloss over the team's huge downsizing entirely. I hate it when companies do that, and I also wanted to be able to speak freely about my own next move.

I got so many lovely messages of support after the announcement. Folks were cautiously optimistic about Wulf’s stewardship of the business, and now (six months later) things are going well with a 3.0 release that refocused on individuals and a thriving Discord community.

And I continue to use Muse nearly every day. The ideation for this article, for example, started in Muse. I genuinely don’t know how anyone gets by without a great digital ideation tool with a freeform canvas. I’m very happy it will continue to exist and improve.

My takeaways

These are in no particular order, just some loose reflections on the whole experience.

Inventing a new document type is hard. The conventions of text editors, spreadsheets, slide deck software, photo editing software, etc provide a foundation whenever you're building a product in that space. But canvas document types are still the wild west, and if you follow tldraw's explorations you'll see that there's no shortage of tiny design details to figure out. Users can't rely on muscle memory, and we spent a lot of time iterating on foundational details like how to zoom in and out.

Helping to define an emerging category is hard. I've done this before with Heroku—there, we were weird undefinable astronauts until other products with the same model (Google App Engine, Docker, AWS Lambda) came along. Even then it was a decade later that the category of "serverless" was finally accepted. I had hoped Muse would be in a similar situation with "tools for thought" but that term came to mean Roam-inspired outliners with backlinks. In the end, working within an established category is much easier.

We were too unfocused in our target market, and ended up as an “everything app.” We neither focused on a particular vertical (academics, designers, authors...) or a narrow use case (PDF reading/annotation, collaborative whiteboarding, design sketching...). That meant we were always spread pretty thin in terms of feature development, and marketing was difficult even over and above the problem of explaining canvas software and digital thinking tools. Part of this reflected our origins: in the lab, Muse was a subset of a comprehensive OS for creative tools, so being general-purpose was in its blood from birth. Part of it was maker's hubris: don't we always dream of general-purpose tools that will be everything to everyone? And part of it was that it's truly the case that Muse excels at the ability to combine together so many different related knowledge tasks and media types into a single, minimal, powerful canvas. Not sure what I would do differently here, even with the benefit of hindsight.

Integrity and goodwill are great, but needs to be balanced with self-interest. I think Muse built a lot of its reputation on being principled, but we were maybe too cautious to do the mercenary things that help you succeed. A good example here is asking users for ratings; I felt like this was not to user benefit and distracting when the user is trying to use your app. Our App Store rating was on the low side (~3.9 stars) for most of our existence. When we finally added the standard prompt-for-rating dialog, it instantly shot up to ~4.7 stars. This was a small example of being too principled about doing good for the user, and not thinking about what would benefit our business.

I'm proud that we were so capital efficient, but I wonder about the opposite path. What would have happened if we had raised $5M – $10M back in 2020 and used that to build for more platforms, with more integrations to other products, and just a more feature-rich and useful product generally? I think that path has as many or more pitfalls as the one we took, so I don't necessarily think it would have turned out better. But I wonder about it.

Growing the team slowly was a delight. At several previous ventures, I've onboard people in the hiring-is-job-one environment of a growth startup. At Muse, we started with three founders and then hired roughly one person per year. This was absolutely fantastic for being able to really take our time to find the perfect person for the role, and then for that person to have tons of time to onboard and find their footing on the team before anyone new showed up. The resulting team was the best I've ever worked on, with minimal deadweight or emotional baggage.

Stuff that works… eventually stops working. Our newsletter was a great growth souce for a while, but then flatlined. Our 10-second demo video tweets would go viral, until they didn't anymore because of [algorithm changes|our target audience fleeing Twitter|people were bored of us|who knows].

Local-first is great for app developers. Although the digression into building our own sync system was arguably too big of an investment for a small team like ours, the result was really a pleasure to work with compared to backend-heavy / cloud systems. The only problem was that we had to build it from scratch. Happily, in the three years since we started working on it, there are a lot more off-the-shelf options.

Apple platforms are great, but you have to be on the web. In terms of development speed, quality of the resulting product, hardware integration, and a million other things: native app development wins. But ultimately your product does have to have some web presence. My biggest regret is not building a simple share-to-web function early on, which could have created some virality and a great deal of utility for users as well.

I accomplished my goal to grow as a storyteller. After decades working in product development, being on the marketing/brand/growth/storytelling side was a huge personal challenge for me. But I feel like I managed to grow into the role and find my own approach (podcasting, demo videos, etc) to create a beacon to attract potential customers to our product.

Final thoughts

That’s about it. Overall this was one of the most satisfying experiences of my professional life, even though it wasn’t a financial success. I feel grateful for all the parts of this that worked: the stellar team, the community of toolmakers we got to be a part of, the journey of becoming a podcaster. Most of all, I loved working with our wonderful users and customers, who shared our vision for a more thoughtful world.

People like to ask me “what’s next” and I’m far enough out now to be able to say I’ll be looking at new opportunities in 2024. After having been in the founder/leader role for the last eight years (four years at Ink & Switch, four years at Muse) I’m in the mood to join up with someone else’s thing. Undoubtedly I’ll still be operating in the realm of productivity software, creative tools, and products designed to help others make things. Feel free to get in touch if you have any short- or long-term opportunities that you think are relevant to my skills.