The People Who Did This Before
AI-written, not Charlie-written.
This research was compiled by Charlie’s AI while he considered a less glamorous kind of software business: one built around a problem already visible inside his Dad’s practice, with early interest from other accountants.
The familiar software-founder story begins with a vast market, a venture round and a promise to change the world. These ten begin with printable bingo cards, store locators, background jobs, student-feedback forms, proposals and accountancy admin. Their founders started close to a real problem, chose a narrow group of customers and built with little or no conventional venture funding.
Some sold. Some kept going. Several became better known for explaining what they had learnt than for the supposedly boring products that taught them.
That is why they matter to Charlie.
In the emails leading here, he described fifteen years of building businesses: online marketing, retreats, the Deep Work design studio, freelancing, a therapy app and Design Canvas. The businesses differed, but the lesson kept repeating. They worked when distribution worked. When the channel disappeared—or never existed—the quality of the thing itself could not rescue them.
His current bet looks different. For months, he has been building accountancy software inside his Dad’s practice. The team uses it, and it has already created measurable value. A live presentation to sixty accountants produced unsolicited interest. It may be less fashionable than another design tool, but it begins with a real problem and one early signal of a possible route to more users.
Researching people who had made comparable choices produced ten stories with oddly familiar contours.
They are not templates, and they do not prove that Charlie’s business will succeed. A set of successful companies is not a base rate. But they do establish that narrow, unglamorous, independently built software can lead to a valuable company—and, in some cases, a good life.
A note on the numbers
Most operating figures below were published by the founders or their companies rather than independently audited. Acquisition and incorporation dates come from public records; unusually uncertain claims are labelled in place. Undisclosed prices remain undisclosed, approximate figures remain approximate, and popular versions of several stories are corrected where the available record does not support them.
That matters. The useful comparison is with what actually happened, not with a polished founder myth.
Part I: Built inside the profession
The closest parallels come from UK accountancy software itself. Both found their decisive form inside a real practice. Both turned administrative friction into a product. Both were eventually bought by larger software groups.
1. GoProposal — James Ashford and Paul Barnes
The sale lost to paperwork
James Ashford’s route into accountancy software began with close-up magic.
Performing at tables of sceptical strangers taught him to earn attention and establish belief quickly. He later ran a marketing agency in Doncaster during the recession, after a long run of ventures that had not worked. The accounting-software success came much later than the confidence associated with him now.
The decisive moment was not a technical breakthrough. It was a failed sale.
Ashford approached Paul Barnes, whose Manchester accountancy practice specialised in digital and creative agencies. Barnes had a sophisticated pricing method, and Ashford was ready to become a client. But the proposal and engagement paperwork took too long to arrive. By the time it did, the buyer’s enthusiasm had gone.
The pricing was good. The administrative gap killed the deal.
Ashford had already experimented with pricing software for other industries. Barnes’s practice gave it the right home. The practice became the working laboratory; Barnes’s pricing method became the substance of the product. Other accountants heard about the internal tool, and GoProposal launched commercially in February 2017.
Its first version cost £4,000 and ran on WordPress Multisite. Ashford chose an existing engine that already handled users, databases and separate sites rather than building every layer of a software platform from scratch. That modest implementation went on to support more than 1,100 customers without a wholesale rewrite.
The company did not raise venture capital. Ashford owned 90 per cent and Barnes 10 per cent: one founder brought product and marketing, the other lived inside the profession and supplied the pricing expertise that made the tool credible.
Distribution through teaching and proximity
GoProposal’s main distribution channel was Ashford himself.
He wrote Selling to Serve, a sales book for accountants, in two weeks. It reached number one in its Amazon category and created a waiting list before the company could onboard everyone. Rather than spend heavily on repeated conference sponsorship, he hired a full-time videographer and produced material that continued working after an event ended. That decision became particularly valuable in 2020, when in-person industry events stopped.
The content was not a substitute for contact with customers. It came from it. Ashford has said that he spoke with an accountant every day for five years, learning what they struggled with and feeding those conversations back into both the product and the teaching.
On 5 October 2021, Sage announced its acquisition of GoProposal. At the time, the company had £1.5 million in annual recurring revenue, roughly 1,100 customers, a Net Promoter Score of 78 and 12 employees. The team joined Sage’s accountants group and remained in Manchester.
The price was not disclosed.
The often-repeated description of an eight-figure exit comes from Ashford’s own later accounts, not from Sage or a filed deal value. It may be plausible, but it is still self-reported. Ashford’s detailed account describes a seven-month sale process involving three interested buyers, six management presentations, two final offers and extensive diligence.
Ashford has also described the emotional drop after the acquisition. The financial result arrived; the expected sense of completion did not. It is an important correction to the idea that an exit is automatically a satisfying ending.
Where this rhymes with Charlie’s story
The strongest parallel is not the acquisition. It is proximity to actual work.
GoProposal brings together two things Charlie’s emails have shown separately: software shaped by daily use inside an accountancy practice, and a proposal tool built during a client sprint.
That does not establish that Charlie’s accountancy product follows GoProposal’s exact wedge. It does show why the gap around pricing, proposals and paperwork can support a serious company.
The other lesson is distribution. Ashford’s teaching worked because it was grounded in years of direct contact with accountants. The audience and product were not separate projects. Each conversation improved both.
2. AccountancyManager → Bright
The administrator inside his father’s practice
James Byrne joined his father’s accountancy practice in 2010, working on marketing and client onboarding. The practice reportedly quadrupled over the next four years. During that period, Byrne could not ignore the administrative burden.
Staff spent large parts of their days chasing information, reminding clients about payments and gathering the documents needed to bring a new client into the firm. Byrne was not a software developer, so he recruited developer Alex Hawke. The pair had already tried a lead-generation product that failed. This time, they built in response to work happening around them in a live practice.
AccountancyManager covered onboarding, document chasing, customer records, deadline-based workflows, identity and compliance checks, a branded portal, electronic signatures, time tracking and integrations with the tools accountants already used. The features did not come from an abstract market map. They came from the queue of problems inside the office.
The company was incorporated on 8 March 2017. Its early years were unromantic: Byrne answered support messages late at night while Hawke shipped features. The team grew from two people in 2017 to 12 in 2018 and roughly 36 by 2022.
By the time it was sold, AccountancyManager served more than 2,000 accountancy firms. It had won repeated industry awards and appeared in the Deloitte UK Technology Fast 50 with reported growth of 1,281 per cent. No institutional funding round is documented, so the cautious description is a company that appears to have grown without conventional venture backing.
The outcome—and what can actually be claimed
Bright acquired AccountancyManager in March 2022. The terms were not disclosed.
Bright had been formed from the combination of payroll and accounting-software companies and went on to assemble more products built by people with direct experience of accountancy. In 2023, AccountancyManager was renamed BrightManager.
That history gives this story more weight than a generic tale of “industry disruption”. A practice insider and a technologist built around real administrative work, reached thousands of small firms and became valuable to a larger software group. The mundane origin was not something the buyer had to overlook. It was part of what made the product useful.
Where this rhymes with Charlie’s story
The resemblance begins with a son initiating a software product inside his father’s accountancy practice, with a real team using it every day.
Charlie arrived with public experience in design and business-building, and AI coding tools now allow him to implement ideas himself. The practice gives him something equally important: daily use by real people, rather than guesses about an imagined customer.
AccountancyManager does not prove that every insider-built tool becomes a large company. It shows that “software made inside an accountancy practice” is not an eccentric category. It is a documented precedent for a route from local irritation to widely adopted product.
Part II: The boring product and the story around it
The next three founders show a second pattern. The product did not need to be culturally exciting for the work around it to earn attention. Clear teaching, honest numbers and a recognisable point of view made small businesses legible far beyond their original niches.
3. Patrick McKenzie — patio11
Bingo cards after work
For six years after graduating, Patrick McKenzie was a salaryman in Japan. He built software during evenings and weekends around punishing working hours.
His side project was Bingo Card Creator, a program that generated printable bingo cards for primary-school teachers. It is a useful antidote to the belief that a software business needs an impressive subject.
McKenzie published unusually precise accounts of how it performed. By August 2009, lifetime sales had passed $50,000, with roughly $27,000 in profit. In its best year, 2012, it produced $64,791.81 in net sales and $38,598 in profit, despite receiving little attention from him. He also wrote about the unappealing parts: low conversion, repetitive support and the exhaustion of answering the same questions for years.
His next product, Appointment Reminder, supplied a sharper lesson in customer pain. McKenzie initially assumed hair salons would pay to reduce missed appointments. In his published account, a walk-in could often fill the chair, while a missed appointment mattered far more to a trades business that might lose $500 to $3,000 of work.
He reported testing the idea by visiting fifteen Chicago businesses with an iPad prototype built for less than $400, and later signing a $75,000 enterprise contract while still running the product alone from Tokyo. It also remained, in his description, a slow and emotionally unrewarding SaaS climb. He quietly sold both products through a broker around 2015–16.
The writing became the compounding asset
The products made money. The writing made McKenzie visible.
On his Kalzumeus blog he published revenue, conversion data, pricing decisions, mistakes and the economics of small software businesses. Essays on careers, names, negotiation and money travelled far beyond the users of his products. A tiny software company became a laboratory from which he could explain much larger systems.
His phrase for the effect was that writing online turns an illegible person into a legible one.
Clients had read McKenzie’s work, and that legibility helped raise his consulting rates and bring opportunities that the bingo-card market never could. In 2016, Stripe recruited him to work on Atlas. He stayed for six years and later continued writing and broadcasting about financial infrastructure. His unaudited estimate that one salary-negotiation essay helped readers gain about $9 million in aggregate pay rises is best treated exactly as that: his own estimate, not an independently verified total.
Where this rhymes with Charlie’s story
Design Canvas was technically ambitious, but readers saw it hit the distribution problem. The accountancy software appears less glamorous, yet it already has users, evidence and a story anchored in real work.
McKenzie’s example explains why that need not be a trade-off. Writing can make an obscure business visible without pretending that attention is the product’s only value. The software creates the evidence; the evidence makes the writing worth reading; the writing makes the builder legible to people outside the immediate customer group.
Charlie does not need to make accountancy sound fashionable. He needs to describe the genuine surprises, mistakes and decisions clearly enough that the lessons travel.
4. Arvid Kahl — FeedbackPanda
A weekend tool for unpaid admin
Arvid Kahl’s partner taught English online to children in China. After the paid teaching ended, she spent roughly two unpaid hours each evening producing repetitive student-feedback reports for the platform.
Kahl built a weekend prototype from reusable snippets because he wanted their evenings back. That small household irritation became FeedbackPanda.
The product found an unusually concentrated market: online teachers doing the same compulsory administrative task, often discussing their work in trusted Facebook groups. Kahl’s partner was not marketing to those groups from outside; she was a member of the profession who understood the task and spoke its language.
Kahl has reported that FeedbackPanda reached roughly $20,000 in monthly recurring revenue within nine months and about $55,000 per month around two years in, with more than 5,000 customers. He and his partner remained the entire team. More than half of customers, he has said, never contacted support.
They sold the company to SureSwift Capital in June 2019. The price remains undisclosed. Kahl has called it life-changing money, but that is a description, not a disclosed figure.
The sale also complicates the idea that a profitable two-person company must feel calm from the inside. Kahl has described reaching the edge of burnout and feeling anxious about having nearly all of the couple’s wealth and income tied to one platform-dependent product.
The second life of a small company
Kahl had documented the business while building it. After the sale, that record became the foundation for books, a newsletter and a podcast about calm, sustainable companies. By Kahl’s account, his audience grew fastest after FeedbackPanda itself had been sold.
His most useful distinction is between building an audience first and building with and for an audience from the start. FeedbackPanda did the latter. The teachers were not a generic following assembled in advance; they were the people whose problem justified the product.
Where this rhymes with Charlie’s story
The parallel is proximity.
FeedbackPanda began beside one person performing a painful task and spread through a high-trust professional community. Charlie’s software began beside a team doing real accountancy work and received early interest after being shown to a room of accountants.
Neither story says, “collect attention, then decide what to sell”. They say, “work with a specific group, solve something that matters to them, and let useful evidence move through the relationships that already connect them”.
There is a warning here too: high revenue and a two-person team do not eliminate platform or concentration risk.
5. Nathan Barry — ConvertKit, now Kit
A designer who narrowed the market
Nathan Barry began as a designer, author and teacher. He has reported that his first book launch made roughly $12,500 in 24 hours from an email list of about 800 people, built in two months. At one point, books and courses produced around $250,000 a year, giving him the means to fund a software company.
He launched ConvertKit in 2013 into a market dominated by Mailchimp. For eighteen months, it struggled. By October 2014, monthly recurring revenue had fallen to $1,207, and an adviser recommended shutting it down.
Barry instead made the market smaller.
“Email marketing for anyone” became email marketing for professional bloggers. He stopped the profitable course business, committed around $50,000 of savings and focused on a customer he understood.
The turnaround depended on work that looked inefficient in isolation. Barry contacted creators one by one, demonstrated the product personally and offered to migrate their lists from existing tools himself. Moving forms, subscribers and automated sequences was low-status manual work, but it removed a major practical cost of switching.
One migration involved a creator’s list of roughly 135,000 subscribers. The subsequent recommendation reportedly generated 48 per cent of the next month’s revenue. Barry said concierge migrations cut churn to around 1.5 per cent, compared with roughly 5.5 per cent he considered normal at the time.
Company-reported figures moved from $1,207 in monthly recurring revenue in October 2014 to $5,020 in March 2015, then to $1 million a month in March 2018, with about 18,900 customers and 34 staff. The company reported $15 million in annual recurring revenue in 2019 and roughly $44 million by 2024, the year ConvertKit became Kit.
It did so without traditional venture capital. Investors had rejected the niche as too small, and a bank had refused a loan. Barry kept control and shared profits with the team.
Correcting the “150 demos” story
A popular version says Barry personally completed 150 sales demos. The underlying record does not support that exact claim.
The documented figure is more than 150 webinars run during 2016, operationally led by a growth employee. Barry’s one-to-one outreach and personal migrations are well documented; attaching a neat total of 150 personal demos turns a real lesson into a cleaner myth.
The real lesson does not need embellishment. Unscalable founder effort enabled a scalable referral and webinar channel.
Where this rhymes with Charlie’s story
Barry’s turning point combined two choices that recur in Charlie’s account: define the buyer narrowly, then get close enough to remove the practical obstacles between interest and use.
Charlie has already demonstrated the accountancy software live rather than waiting for a polished mass-market funnel. The early signal came from a professional audience that could recognise the problem immediately.
The scale will not necessarily resemble Kit, and it does not need to. What transfers is the sequence: specificity before reach, founder-led contact before a mass-market funnel, and earned proof before broad claims.
Part III: The calm-company path
“Bootstrapped” describes financing. It does not, by itself, describe the life a company creates. Rob Walling and Tyler Tringas gave more precise names to the gradual, optional and deliberately limited version of the path.
6. Rob Walling — the stair-step
Learning on small rungs
Rob Walling did not jump directly from consulting into his largest software company.
Walling has described an early portfolio that included DotNetInvoice, a downloadable invoicing script making a few thousand dollars a month, along with a job board for electrical linemen, personalised beach towels, a wedding-site builder and ebooks. Each was small. Together, they replaced consulting income and taught him search marketing, copy, pricing and support without making one irreversible bet.
In 2011, Walling bought HitTail, an existing keyword tool, improved it and sold it in late 2015. He later wrote that trying to launch his next company without the skills, confidence and money gained from HitTail would have ended badly.
That sequence became his stair-step method:
- Start with one simple product that can use an existing marketplace or channel.
- Stack small products until they buy back the founder’s time and reduce risk.
- Only then attempt a standalone recurring-revenue company.
His third step was Drip, an email-automation product co-founded with Derrick Reimer. Walling says that before building it, he emailed 17 founder friends and received 11 commitments at $50 a month. Drip launched in November 2013, reached roughly $2 million in annual recurring revenue with about 1,500 customers and nine employees, and was acquired by Leadpages in 2016.
The sale price was not disclosed. In Walling’s account, two conditions reveal more than a speculative number would: he tied his earn-out to features he knew the team could ship rather than revenue targets outside his control, and he protected the team and customers from a deal-dependent round of layoffs or a destructive product change.
The asset around the assets
Walling had started the Startups For the Rest of Us podcast in 2010 and co-founded MicroConf in 2011, before Drip existed. The audience and community persisted across his products and outlasted his ownership of individual companies. In 2018 he started TinySeed, a funding model aimed at software companies that did not fit the conventional venture pattern.
His work made an important distinction visible: a niche can be small enough to deter a large competitor while still being large enough to support a valuable business.
Where this rhymes with Charlie’s story
Charlie’s earlier businesses were not wasted detours. Online marketing taught selling. Retreats taught operations and channel economics. Deep Work taught how to build a specialist team and serve demanding technology companies. Freelancing kept the craft and customer contact alive.
The public story of the accountancy product begins after those rungs, not before them. He arrives with design, sales, operations and implementation experience, plus a real practice in which to test the work.
Walling’s framework does not say every rung must become a permanent company. It says small, revenue-producing work can fund the skill and freedom required for a more durable bet. Seen that way, the apparently wandering route is cumulative.
7. Tyler Tringas — Storemapper
A small product that stayed small on purpose
Tyler Tringas worked in renewable-energy finance before co-founding a solar business, failing to find a technical partner and teaching himself to code. After the solar company failed, he freelanced as a Shopify developer to get paid while learning.
A client needed a store-locator map. Tringas says he built the first Storemapper version on a flight from San Francisco to Buenos Aires, launched it at $5 a month and found three to five paying customers within 24 hours.
Popular retellings often stop at Storemapper’s early $50,000-a-year milestone from August 2014. The company went on to reach roughly $100,000 in annual recurring revenue in early 2015, pass $200,000 a year in 2016, and enter an approximate $250,000–$600,000 annual recurring revenue range by the time it was sold.
In his account, Tringas ran it while travelling, first alongside heavy freelance weeks and later in roughly ten hours a week with a small remote team. He was explicit that he did not want to become the king of store locators. The objective was a healthy internet business, not dominance of a category.
He sold Storemapper to SureSwift Capital around August 2017, after a five-month process and several offers. The exact price was not disclosed. He called it “level-up money”, and said diversification was one of the best reasons to sell. In his telling, this was a sale from a stable position, not a public claim of an enormous exit.
Micro-SaaS and the right to choose
Tringas popularised “micro-SaaS” for a narrowly focused subscription business run by one person or a small team, with low costs, a dedicated customer group and no outside funding. “Micro” described the scope of the problem, not a mandatory ceiling on quality or revenue.
He later founded Earnest Capital, renamed Calm Company Fund, to back businesses that did not risk survival for growth. By the end of 2024, the fund had stopped making new investments while continuing to support its existing portfolio of roughly 77 companies. It is another example of deciding to stop rather than treating continuation as mandatory.
Where this rhymes with Charlie’s story
The relevant idea is not the digital-nomad setting or the store-locator market. It is that the founder can choose the amount and kind of growth a company is built to absorb.
Charlie has described wanting accountancy software to create the best life for Coco, not merely the most impressive company. Storemapper shows that a narrow product can be valuable without being pushed towards maximum headcount or maximum market share.
Part IV: One wedge, two possible endings
The final three cases focus on the shape of the product and the definition of success: a proposal tool expanding into signatures, a one-person infrastructure company engineered around constraints, and a designer who declined a sale to keep building.
8. Ruben Gamez — Bidsketch and SignWell
Proposals first, signatures next
Ruben Gamez spent nine years as a developer and technology manager at a payroll company, where he had helped build a proposal system used in seven- and eight-figure deals. When he later helped a freelance friend prepare a proposal and found only outdated tools, he already understood that the document between a conversation and a signed project could have disproportionate value.
Before building, Gamez checked search demand. More people searched for proposal templates than proposal software, suggesting that the need existed even when customers did not use the product category’s language.
He launched Bidsketch in late 2009 for designers, freelancers and agencies. The first version cost roughly $4,000–$5,000 in outsourced development before he ran out of money and completed it himself.
The private launch produced 21 paid accounts; the public launch was slower. One of Gamez’s most repeated experiments was removing the free plan and requiring a card. Paid sign-ups reportedly increased by roughly ten times. Gamez reports that about eighteen months after launch, revenue reached $6,000–$7,000 a month, enough to replace his salary. It later crossed a series of plateaux to roughly $30,000 a month and around 1,500 paying customers, with a small remote team.
Customers then exposed an adjacent problem. After creating a proposal, they still needed expensive or cumbersome software to collect a signature. Gamez built Docsketch in 2019 and renamed it SignWell in 2021. By 2024, SignWell reported roughly $5 million in annual recurring revenue and more than 60,000 businesses served.
The sequence was proposals, then signatures: the second product emerged from friction experienced by users of the first.
Two corrections worth keeping
Gamez was a guest on the Bootstrapped Web podcast, not its co-host.
And “no venture capital ever” is too broad. Bidsketch was bootstrapped, but SignWell joined TinySeed’s 2020 programme. TinySeed was designed for independent software companies, but it was still outside capital. Describing it accurately does not weaken the story; it clarifies what kind of company each product was.
Where this rhymes with Charlie’s story
The proposal is already part of Charlie’s public build log. Gamez shows that proposal software can be commercially important rather than trivial: it sits at the point where an interested buyer becomes signed work.
There is also a lesson in endurance. Bidsketch grew through long plateaux, not one launch spike. Gamez’s observation was that many competitors removed themselves from the market; his companies survived long enough to compound.
The comparison has limits. Gamez served a broad market of freelancers and small businesses; Charlie is building for accountants. What transfers is the commercial importance of the proposal itself, not a predetermined sequence of adjacent features.
9. Mike Perham — Sidekiq
One owner, boring infrastructure
Mike Perham repeatedly encountered inadequate background-job systems while working as a Ruby developer. In January 2012, he released Sidekiq, a more memory-efficient way for Ruby applications to process work in the background.
Its subject is nearly invisible by design. Background infrastructure is noticed mainly when it fails.
The business model made that invisible tool unusually powerful. Sidekiq has a free open-source core and paid Pro and Enterprise editions sold as annual licences. Customers install and run the software in their own infrastructure. Perham sells maintained software rather than hosting every customer’s production workload himself.
His own published figures show the progression:
- About $100,000 a year in 2013.
- Roughly $80,000 a month in 2017, from 726 customers.
- 1,850 customers and $13.5 million in cumulative gross sales by January 2022.
- In April 2023, annual revenue closer to $10 million than $1 million, in Perham’s own description, from closer to 2,000 customers.
That final comparison is self-reported rather than a filed revenue figure. So is the striking claim that his licensing infrastructure cost $18 a month. The important point is not to convert either into false precision; it is that a narrow developer tool supported a large business run by its sole owner without employees, according to Perham.
The founder did not hand it off
Another popular retelling says Perham stepped away from Sidekiq. More recent activity shows otherwise.
Perham released Sidekiq 8.0 in March 2025 and continued releasing versions and building a new administration tool through 2026. He remained the sole owner-maintainer rather than handing the company to an operating team; in April 2023, he reported having no employees.
The low headcount was engineered through product and policy choices. Perham set a limited support entitlement, offered no private support for free users, used public issues and documentation so answers accumulated, and priced inconvenient procurement accordingly. He treated support demand as something the business model could shape.
Where this rhymes with Charlie’s story
Sidekiq is not a direct blueprint for software used by accountants. Developers can install packages, read technical documentation and diagnose problems in a way that most professional-service users should never be expected to.
The transferable principle is narrower: product and policy choices shape the maintenance load a small software company must absorb. Charlie’s public story shows how AI has increased one person’s building capacity; Perham’s adds the longer-term question of whether that software can remain dependable and maintainable.
10. Peldi Guilizzoni — Balsamiq
A product builder who wanted the whole job
Giacomo “Peldi” Guilizzoni spent about six and a half years at Macromedia and Adobe, working on web-conferencing software. He was not escaping a disastrous career. He wanted to learn the parts of a business that his role did not include: marketing, sales, support and pricing.
He built a wireframing tool in the evenings, saved roughly a year of runway, left Adobe on 15 June 2008 and launched Balsamiq Mockups four days later. The first customer found the live site before the formal announcement.
Balsamiq’s deliberately sketchy appearance was both a design decision and a distribution mechanism. A polished mock-up invites arguments about surface detail. A hand-drawn one keeps attention on structure and flow. The visual style made the product more useful and recognisable at the same time.
Guilizzoni published the company’s performance early:
- More than $10,000 in the first six weeks, later updated to $11,488.
- More than $100,000 and 800 customers in under five months.
- $2 million in cumulative sales within 18 months.
- $6.99 million in revenue in 2023 and $6.58 million in 2024, while remaining profitable.
Transparency was part of the company’s way of earning trust. Balsamiq later applied the same instinct internally through open salaries and profit-sharing rather than employee equity—a choice Guilizzoni said helped avoid cap-table pressure towards a sale.
Saying no to the exit
Around 2015, Guilizzoni received a serious acquisition offer and considered it. He ultimately returned to the reason he had started: to build the online equivalent of an excellent, durable, family-run restaurant, not a company whose success depended on being sold.
He declined.
The latest financial record, from 2024, showed Balsamiq profitable; the research compiled in 2026 found no sale or outside investment. The company now lists an interim chief executive, suggesting that Guilizzoni stepped back from some day-to-day operations after 2022. Keeping a company forever does not require the founder to keep the same operating job forever.
Where this rhymes with Charlie’s story
This is the closest craft parallel. A product-minded builder makes an opinionated design decision, learns the whole commercial job and refuses to remain inside one speciality.
Charlie’s emails have documented a similar expansion: from product design to code, AI systems, sales and the operations of a real business. They have also made family part of the definition of good work, rather than a reward postponed until after it.
Balsamiq shows that independence can be a durable outcome. A profitable company can mature and rely less on its founder without first being sold.
What the ten stories add up to
Put side by side, the founders differ too much to support one neat formula. Their markets ranged from teachers to developers; their companies ranged from one person to dozens; their outcomes ranged from quiet brokerage sales to strategic acquisitions and long-term independence.
But six patterns recur.
1. They began close to an observable problem
GoProposal began with a sale lost to delayed paperwork. AccountancyManager was written in response to tasks piling up inside a practice. FeedbackPanda began with two unpaid hours at the end of a teacher’s day. Storemapper began with a client request. Bidsketch began with an awkward proposal process.
The idea did not have to be visionary. The evidence had to be near.
2. Narrowness was an advantage
Accountants, online teachers, professional bloggers, Ruby developers: each group was specific enough for the founder to understand its language and reach its members. The niche reduced abstraction. It made the first customer, the useful feature and the credible message easier to identify.
“Boring” often meant “too particular for outsiders to notice”. That could be a form of protection.
3. Early distribution was borrowed or interpersonal
The first channel was commonly a professional community, an existing marketplace, direct demonstrations, one influential referral or the relationships around a working practice.
None of this was passive. It often required the founder to teach, migrate data, answer questions or talk with customers one at a time. But it started where attention already existed instead of asking the founder to manufacture a large generic audience before solving a problem.
4. Public explanation and community made the work legible
McKenzie published precise economics. Kahl documented the company he later sold. Barry exposed revenue and churn. Walling established a podcast and co-founded a conference. Tringas wrote the language of micro-SaaS. Guilizzoni made openness part of the company’s character.
Their writing did not compensate for useless products. It made useful, easily overlooked work understandable. In several cases, the reputation created by explaining the work lasted longer than the original software.
5. Small did not mean effortless
FeedbackPanda’s two-person success brought burnout and concentration risk. Sidekiq’s one-person scale depended on unusually deliberate product and support boundaries. ConvertKit’s turnaround depended on manual work that would look irrational in a mature company. AccountancyManager began with late-night support and grew into a team of roughly 36.
The calm-company path is real, but “calm” describes a designed outcome, not the absence of hard periods.
6. Both endings exist
Sage bought GoProposal. Bright bought AccountancyManager. Leadpages bought Drip. SureSwift bought FeedbackPanda and Storemapper. Balsamiq stayed independent. Sidekiq remained with its sole owner.
The evidence does not demand one preferred ending. It supports optionality. A founder can sell a useful company to a larger operator, or keep a profitable company and gradually redesign their role inside it.
So, is Charlie’s path really well trodden?
Not in the sense that someone else has lived his exact life.
No founder here combines the same history: sailing-beach summers, online marketing, retreats, a 30-plus-person design studio, years of consulting, products that lacked distribution, AI-assisted building, a daughter and a place inside his Dad’s accountancy practice.
But the commercial shape is familiar.
A builder has accumulated skills through several imperfect businesses. He has learnt that a product without a route to customers is still a product without a business. He is now building inside his Dad’s practice, with its team using the software each day. One presentation to sixty accountants prompted unsolicited interest. And the story around the work is beginning while the outcome is still uncertain.
None of this guarantees success. It does describe a credible starting position.
The deepest parallel across these ten stories is not “boring software gets acquired”. It is simpler:
Useful work begins near real pain. Distribution begins with a real route to reachable people. A good story helps the evidence travel.
Charlie spent years learning what happens when a channel disappears—or when a product is built without one.
This time, the problem was already in the room, and the presentation offered an early signal of a possible route to other users.
Companion long read to email six. Previous emails: ellington.design/emails.
Continue the Journey
Personal updates on what I'm building and where I am.