“My Best Investment”: Small Business Leaders Share Their Top Picks

I’m Runbo Li, Co-founder & CEO at Magic Hour.

The most impactful investment I’ve ever made cost zero dollars. It was time. Specifically, the six months I spent posting AI-generated videos every single day on social media before Magic Hour was even a company.

Most people think of “investment” as writing a check or buying equity. But the highest-ROI investment available to anyone right now is investing time into learning AI tools before the market prices that skill in. I did this in early 2023, when Stable Diffusion was still rough around the edges and most people dismissed AI video as a toy. I was spending nights and weekends after my day job at Meta, figuring out how to chain together models, tweak outputs, and produce content that actually stopped people mid-scroll.

One NBA edit I made went viral. Mark Cuban saw it, followed me, became a paying customer. The Dallas Mavericks reached out organically. That single piece of content, born from months of daily reps, became the proof of concept that gave us conviction to leave our jobs and build Magic Hour full-time.

The reason this stands out is the asymmetry. I didn’t risk capital. I risked comfort. I traded Netflix hours for render times. And the return wasn’t 10x, it was infinite, because the denominator was essentially zero dollars. It gave us our first customers, our narrative for Y Combinator, and the distribution insight that shaped our entire product philosophy.

People overthink investments. They wait for the perfect deal, the perfect valuation, the perfect moment. The best investment I’ve made was just showing up every day to a new medium before it was obvious, and letting compounding do its thing. Consistency in an emerging space is the closest thing to a cheat code that exists.

One investment I’m genuinely proud of at Scale By SEO happened about two years ago when I decided to invest heavily in building an in-house content team instead of continuing to outsource to freelancers. We were spending around $8,000 monthly on freelance writers, and while the content was decent, it lacked the consistency and deep understanding of our clients’ industries that really moves the needle in SEO.

The switch wasn’t cheap. Hiring two full-time writers, investing in training, and building content workflows cost us roughly $30,000 upfront. That’s a significant chunk for a growing agency. But within six months, our client retention rate jumped from 72% to 89%. Our content wasn’t just technically good anymore. It was genuinely helpful, industry-specific, and written by people who understood our clients’ businesses inside and out.

What makes this investment particularly meaningful is how it changed our entire approach to SEO. We stopped chasing algorithm updates and started focusing on what actually matters: creating content that real people find valuable. Our writers attend client onboarding calls, they research competitors extensively, and they’ve developed expertise in niches like healthcare, legal, and home services that you simply can’t get from generalist freelancers.

The ROI has been incredible. We’ve been able to increase our retainers because clients see tangible results faster. Our case studies are stronger because the content actually ranks and converts. We’ve even had clients specifically mention our content quality as a reason they referred us to others.

Looking back, I think what makes this investment special is that it forced me to bet on people rather than just processes. In digital marketing, there’s always a new tool or tactic promising shortcuts. But building a team of skilled writers who care about our clients’ success? That’s created sustainable growth that no algorithm change can take away from us.

The most impactful investment I’ve made as a founder was committing six figures of personal capital to building our internal evaluation harness before we shipped any customer-facing AI feature. Most early-stage AI companies treat evals as a phase-two problem; we treated them as core infrastructure. Every prompt, model, and agent gets a regression test on a versioned dataset before it touches production. The reason this stands out is that it compounds: every customer escalation becomes a permanent test case, model swaps take hours instead of weeks, and we can confidently swap GPT for Claude or a fine-tuned open model when economics shift. In a market where model providers change pricing monthly, that optionality has been worth more than any feature we shipped.

The investment I’m most proud of was hiring a dedicated quality control manager in our India supply chain — not a flashy tech tool or market expansion, but a single hire. Before that investment, we were losing about 3-4% of every order to rework or replacement due to quality issues, which was quietly eating our margin. After bringing that person on full-time, defect rates dropped to under 0.8% within six months, and client retention improved because repeat issues stopped. That one hire paid for itself in eight months and unlocked higher-margin contracts with clients who needed consistent quality. The best investments often aren’t the most exciting ones on paper.

Of the many impactful investments we’ve made, building our own low-latency trading platform from the ground up has to be one of the best at return on investment and also changing how we can service our institutional clients. I put $200K into custom server architecture at the end of 2019 for a custom trading platform to power our low-latency trading and we hit a return of over 400% within 3 years off of that base. The core of the TradingFXVPS network is a collection of colocated servers running in key financial centers around the world (London, New York, Chicago, Hong Kong, Singapore, Tokyo, Amsterdam and Frankfurt). We engineered a sub-millisecond interconnect between servers so that the network functions as a highly efficient single computing resource. Building out our own servers and interconnects as opposed to leasing a typical white-label hosting environment for high volume trading was a high-risk move at the time. But as with most things in life, there is a big difference between doing something the easy way and doing something right. And in this case, working with prop trading shops that are running tens to hundreds of millions of dollars in assets to push the boundaries of trading execution speed was the right thing to do. Execution speed is not just a matter of latency, but also consistent, real-world performance under extreme stress. That is what we built. The proudest moment I can think of regarding this issue was during the March 2020 crash, when the trading volumes surged 800% overnight (for some assets even 1000%). I was on the phone with one of our largest clients (a prop trading firm of around $20 million) who was extremely panicked, due to potential losses of hundreds of millions caused by trading at even 50ms higher latency than usual (his servers were not able to handle the high traffic). I managed to reassure him that our servers are running as usual (99.999% uptime) and that he wouldn’t face any losses thanks to our low latency trading environment. What a great lesson it has been to build a fintech business that creates value in the market by projecting the market’s technical requirements for the next 5 years and building that for today’s customers. Our current trading platform now services over 10 million trades a month and forms the basis of all the services that we provide at TradingFXVPS.

The most impactful investment I have made is one most financial advisors would not put at the top of a list — I bought a small piece of land last year.

It was not a large plot. It was not in a premium location. But the reasoning behind it came from a place that no spreadsheet captures cleanly.

Our family lost nearly everything in 2011 when our publication business collapsed. The bank began proceedings to auction our house. We spent a decade recovering. In 2021 we finally settled the debt and kept the home. That experience changed how I think about money permanently — not in the abstract way that finance books describe, but in the visceral way that comes from nearly losing the roof over your family’s head.

When I bought that land last year, the calculation was simple. If my job disappears tomorrow, if something unexpected happens again, my family will not be homeless. We can build something modest on that land. We have somewhere to go. That certainty is worth more to me than the return figure on any investment statement.

The financial case also holds up. Land prices in India have been appreciating at rates that consistently outpace the interest on a fixed deposit. A bank FD gives you 6 to 7 percent annually and taxes eat into that further. Land in developing areas has been returning significantly more over five to ten year horizons, with zero management overhead and no counterparty risk beyond the title itself.

But honestly, the financial case was secondary. What makes this investment stand out is that it solved a problem no bank account can solve — the fear of having no fallback. After 2011, I know what it feels like to have nowhere to go. That land means my family will always have somewhere.

One investment I’m particularly proud of is my early participation in the fintech education space, specifically platforms building financial literacy infrastructure for underserved populations.

What stood out wasn’t the valuation or exit potential. It was recognising a structural gap that traditional finance kept ignoring: retail participation was exploding faster than financial understanding. Regulators saw the risk. Institutions saw a compliance problem. I saw an opportunity.

The investment made sense because it addresses a real market failure. Over 15 crore mutual fund folios in India, yet investor outcomes remain fragile. Women account for 30% of individual mutual fund AUM, but vulnerability to mis-selling remains disproportionately high. The gap between access and informed participation keeps widening.

What makes it stand out: unlike venture-backed fintech chasing disruption narratives, this investment plays the longer game. Governance compounds. Trust compounds. You don’t see returns in a Series A. You see them when women stay invested through downturns, when first-time investors make rational decisions instead of emotional ones, and when families build multi-generational wealth conversations.

The fintech industry obsesses over speed and scale. But the most impactful investments are often the unglamorous ones—infrastructure that strengthens decision-making, transparency systems that reduce behavioural risk, and education platforms that close the knowledge gap.

That’s the thesis I’ve backed. Not because it’s trendy. Because it solves the actual problem.

The best investment I have ever made was in a small, rundown campground in a small rural town. I purchased the property before anyone else showed any interest in it. The land was a mess, and the water lines were broken. Nobody thought it was worth saving. What makes this investment so special is the changes we made. We put in the water lines. Cleaned up the site. We put in electric hookups for RVs. It was not long, and the whole campground was filled with enthusiastic families.

The business is now bringing substantial cash flow to my company. Though the greatest part has nothing to do with cash flow. We provided comfortable accommodation for the visitors. We took a neglected tract of land and made it useful. That is what Investorade is all about. Buy as is. Fix what matters. Treat people right.

The investment we are proudest of was not a campaign or a hire. It was giving the team enough AI seats and credits to build a culture of constant prototyping. Many teams underfund exploration and then wonder why innovation feels slow. We wanted people to test ideas early while they were still rough.

The change created a strong second effect across the team over time. Once the barrier to testing dropped people stopped waiting for perfect briefs and asked better questions. We became less performative and more driven by evidence in our work. From our work in SEO and growth we learned the best investment makes learning cheaper and faster.

I put $847,000 into building a 140,000 square foot fulfillment center from scratch when I was running my 3PL company. Not buying an existing warehouse. Building one. Most people thought I was insane because I was in my twenties and could have just leased space like everyone else.

That investment stands out because it forced me to think ten years ahead instead of quarter to quarter. When you’re signing construction contracts and designing dock configurations, you can’t half-a** your vision. I had to commit to what I believed the future of fulfillment would look like. We designed for automation before most 3PLs even had warehouse management systems. We built in redundancy for power and internet that seemed excessive at the time but saved our clients during outages that crippled competitors.

The real payoff wasn’t just the facility itself. It was what building something permanent did to my team and client relationships. When you own the building, clients know you’re not going anywhere. We landed three major accounts in the first six months specifically because they’d been burned by 3PLs who abandoned ship when their leases ended. One of those clients is still with the company today, years after I sold it.

Here’s what made it truly impactful though. That experience taught me the difference between investing in assets and investing in capabilities. The building was an asset. The systems we built, the team we hired, the processes we refined – those were capabilities that transferred when I exited. The buyer paid a premium specifically because we’d created something that couldn’t be easily replicated.

Now with Fulfill.com, I invest differently. I’m putting capital into technology that helps brands avoid the mistakes I made when I was on the other side searching for fulfillment partners. But the lesson is the same. Real investments force you to commit to a vision and build something that outlasts your direct involvement. That’s when you know it matters.

The investment I’m most proud of was hiring our first full-time developer in 2008 when Thrive was barely breaking even. Revenue couldn’t justify the salary, but I recognized that outsourcing development created bottlenecks limiting our ability to serve clients quickly.

That hire cost $65,000 annually when our profit was maybe $40,000. It felt terrifying, forcing me to operate leaner everywhere else. But having in-house development meant we could implement client requests within days instead of waiting weeks for external contractors to fit us into their schedules.

Within eight months, that speed advantage helped us close three major clients specifically because we could promise fast implementation. Those clients generated enough revenue to justify the developer salary twice over. More importantly, the competitive advantage of speed compounded as we grew.

Fifteen years later, our development team is eight people, and that initial scary hire proved that investing ahead of obvious affordability creates capability competitors cannot easily replicate. Most agency founders wait until they can comfortably afford new hires. We grew faster by investing before comfort arrived.

The lesson: invest in capability that removes bottlenecks before profitability makes it obviously affordable. Waiting for safety guarantees slow growth competitors eventually outpace.

The investment I’m proudest of isn’t in a company. It’s the two years we spent building our own investor intelligence layer at spectup. We were a small team. We had paying clients. Every data feed we wanted was four figures annually, and we wired together 80+ of them ourselves over time, mostly evenings and weekends, no external capital.

It stands out because it changed what we could promise founders. Before, we said we had a curated investor network. After, we could tell a client exactly which 12 funds had closed in the last 90 days, which partners had shifted thesis, and which firms were quietly opening Series B mandates. That moved us from boutique advisor to neo-investment bank in how clients positioned us. The capital it took was small. The compound return on every client engagement since has been the highest-ROI bet we’ve made.

The most impactful investment I have made was the decision to never take outside capital for Paperless Pipeline. I started the company in 2009 by cold-pitching real estate brokers during the housing downturn. The first dollars in were customer dollars. They are still the only dollars in. 16 years later we process around 6% of every home sale in the United States, with over 1,700 brokerages, 90,000+ users, and 4.6M+ transactions through the platform. None of that exists if I had taken a Series A in 2011.

What makes this stand out to me is the compounding effect on customer behaviour, not on my equity. When you are funded by customers you spend your time on customers. When you are funded by VCs you spend your time on the next round. The two roads look similar in year one. They go to very different places by year five.

A concrete example. Tony Garrant at Abundant Realty replaced a $35,000-per-year office manager with Paperless Pipeline at $125 per month. Over 14 years that brokerage has saved around $470,000 in operating cost. That number only exists because we never had a board telling us to push pricing every quarter. We charge per transaction. No annual contracts. No per-seat ramp. The customer wins on every cycle. Customers like Tony stay for over a decade.

The same pattern repeats in our 6-week release cadence. Every six weeks, every year, for 16+ years, we ship product upgrades. The cadence is possible because we are profitable from operations, not from a runway. The pressure is to make the product better for the same brokerage that paid us last month, not to chase a feature that demos well at a board meeting.

What makes the bootstrap investment hard is the years one through three. You make less money. You hire slower. You miss conferences. You watch competitors raise. The temptation is real. The reason to keep going is that the company you build during those years is the company you actually want to run a decade later. By year ten, the founders who took capital are usually out. The founders who bootstrapped are still on the screen-share with a customer in Rochester or British Columbia.

I am not anti-venture. Some businesses cannot exist without it. But for vertical SaaS serving small businesses, the customer-funded path is usually the right one. The investment in saying no is the biggest investment I ever made.

We are most proud of investing in operational clarity before pursuing scale as a priority decision. In fleet operations, growth without visibility usually multiplies mistakes over time. We focused on building systems that show problems early instead of reacting after costs appear in operations every day. This created a cleaner decision environment for managers and leaders across teams in practice.

It improved judgment across the business beyond any single metric we use. Safety conversations became more objective and easier to align over time. Driver coaching became less personal and more constructive through regular reviews. Dispatch and leadership now work from the same facts and reduce waste in planning every day.

The best investment I made was killing my own assumption that more ad spend was the answer, and putting money into speed-to-lead instead. I learned this the expensive way from call logs. Across HVAC clients, 41% of inbound leads from paid ads were not answered within 60 seconds, and 22% went to voicemail and were never returned. That means businesses were paying for demand they literally did not pick up.

So instead of pushing harder on media buying, I invested in building AI voice coverage that answered immediately after hours and during overflow. One example stands out. An HVAC client had been leaning on Google Performance Max, but too many paid calls were slipping through when the office was busy or closed. We capped the ad budget, added AI answering to catch and qualify every inbound call, and booked the same pipeline without spending more on ads. Cost per booked job dropped 38% with no extra ad spend. That was the moment it clicked for me. The highest-leverage investment is often not getting more leads; it is rescuing the ones you already paid for.

What makes it stand out is that it changed how I think about growth. The flashy move is buying more traffic. The durable move is fixing the leak at the point of contact. My rule now is simple: do not buy another lead until you can answer the current one in under 60 seconds.

The investment I am most proud of is one that most founders would have considered wasteful at the time: building our own GPU health monitoring system from scratch instead of relying on cloud provider dashboards.

When I started GpuPerHour, we were a marketplace connecting ML teams with GPU compute. The obvious move was to lean entirely on the monitoring tools that came bundled with each cloud provider. They were free, they were adequate, and building our own system would cost us roughly three months of engineering time we did not have.

But the provider dashboards only showed us what the provider wanted us to see. They could not tell us about thermal throttling patterns across different GPU models, or how memory utilization correlated with job failure rates on specific hardware generations. We were flying blind on the metrics that actually determined whether our customers had a good experience.

So we invested those three months. We built a monitoring layer that tracks 14 performance metrics per GPU session at 10-second intervals. The system flags hardware that is degrading before it fails, routes workloads away from underperforming nodes, and gives customers transparency into exactly what their rented GPU is doing.

What makes this investment stand out is the compound return. In the first year, it reduced customer-reported failures by 62 percent. That drove our NPS from 34 to 71. The higher NPS translated into word-of-mouth referrals that now account for 40 percent of new customer acquisition, which means lower marketing spend per customer. A single infrastructure investment cascaded into retention, reputation, and acquisition gains that are still compounding two years later.

The lesson: the best investments often look like overhead until they start paying dividends across multiple parts of the business simultaneously.

The most impactful investment was committing early to efficiency-focused product curation, not volume. Plenty of sellers expanded endlessly, yet smarter assortments better served changing homeowner priorities. That required researching performance standards, rebate trends, and long-term operating cost advantages. I preferred helping customers buy better rather than simply buy more.

That focus created stronger trust because savings continued long after installation finished. Efficient systems lowered utility bills, supported sustainability goals, and reduced ownership regret. The strategy also aligned demand with products offering clearer value during uncertainty. It stands out because thoughtful curation improved outcomes for customers and the business.

Building Interseller was one, but the investment I keep coming back to is the three months we spent before building anything at Pin just talking to recruiters. Not demos, not surveys. Actual conversations where we watched them work and asked dumb questions about their day. That time felt like a cost at the time, the team was pushing to start building, and there was real pressure to have something to show investors.

What it bought us was a product that recruiters didn’t feel the need to fight. That sounds like a low bar, but most recruiting software has a 40% adoption problem because it was designed for the buyer, not the user. We didn’t have to retrofit that lesson later, which I think saved us probably six months of rework.

The most impactful investment we’ve made at ChainClarity is one that doesn’t show up on any balance sheet: investing deeply in whitepaper analysis infrastructure before it was fashionable to care about fundamentals.

In 2021-2022, the prevailing wisdom was that retail crypto investors didn’t want rigor — they wanted price predictions and moon calls. We bet the opposite. We built systematic processes to parse and translate the technical documents that actually define what a crypto project does: its tokenomics, consensus mechanism, governance model, and real-world utility.

That bet has stood out for one reason: it compounds. Every whitepaper we’ve analyzed — now covering 500+ projects — adds to a body of knowledge that makes the next analysis sharper. When Ethereum transitioned to proof-of-stake, we already had deep context on every competing L1’s staking model. When DeFi protocols started failing in 2022, we could explain exactly why — because we’d read the mechanisms that broke.

What makes it stand out isn’t just the research output. It’s that investing in genuine understanding created trust that marketing spend never could. Our readers are developers building on-chain, fund analysts doing due diligence, and first-time investors trying not to get burned. They return because we don’t have an opinion on price — only on what a project actually does.

The lesson: in a market saturated with noise, depth is the scarce resource. Investing in infrastructure that produces real understanding — not content that mimics it — is the highest-ROI move we’ve made.

Best investment of the year wasn’t money. It was a decision to stop spending it on ads.

Late last year I pulled most of our paid acquisition spend at Streamrise — we work with Twitch and Kick streamers on audience growth — out of Google and Reddit promoted posts, and put the team’s hours into writing instead. Three things specifically. A Substack on streaming mechanics, one long post a week. Quora answers on creator monetization, two or three a day max, because past that you start sounding like a content farm. And bylined answers through Featured.com and Connectively when a query actually fits something I’d thought hard about.

I won’t pretend I had a clean spreadsheet justifying the swap. Our paid CPMs had been drifting up for two years and our organic mentions had been drifting up at the same time, and at some point the ratio stopped making sense.

The part I keep coming back to happened in March. I asked Perplexity a boring benchmark question I’d written a Substack post about, something about concurrent-viewer-to-follower ratios on Twitch, and a paragraph I’d written came back as the cited source. Tried it on ChatGPT. Same thing. I hadn’t optimized for that. I’d just written the piece because someone on our team kept asking the question.

A paid click stops the day the card declines. A post that an AI model has decided is a reasonable source keeps answering the question for you, at 3am, in a country you don’t sell into yet, for free. That asymmetry is where I’m putting the hours now.

Investing in internal documentation before it felt necessary. Early in product development, it seems like overhead — everyone knows how things work. But knowledge that lives in people’s heads doesn’t scale. When team composition changes or projects hand off, undocumented systems become expensive to recover. The return isn’t visible immediately, which is why most teams skip it. That’s exactly what makes it worth doing early.

My most important investment was my own entrepreneurial development.

When I left employment and launched my construction company, I was not simply investing money. I was investing in my ability to make decisions, take responsibility, and build something independently.

That investment led to a successful construction business that thrived for seven years until the global financial crisis brought the market to a standstill. Later, the same mindset helped me build and rebuild an educational institution multiple times, one that continues to operate successfully today.

That experience led me to an important realization: I no longer believe that the investment itself is the most important factor. What matters far more is the person behind it.

A good idea alone is rarely enough. But when the right person stands behind it, even a simple idea can become a successful business.

There is room for debate about how much of that comes from natural ability and how much can be learned. What I do know is that mindset, responsibility, and perseverance can all be developed.

I also know that the success of my ventures was never the result of a single investment. It came from the ability to recognize opportunities in new situations, make decisions, and follow through on them consistently.

Today, I am building a new system. Nilvara, along with the LifeCode and LifeCommand books, represents the culmination of decades of entrepreneurial experience, decisions, successes, failures, and lessons learned. They are not simply individual projects. They are proof that accumulated knowledge can continue to create value in new forms.

That is why I am not most proud of any single company or investment. I am proud that across different industries, under very different circumstances, I have repeatedly been able to create value.

To me, that is the greatest return on the investment I have made in myself throughout my life.

I have been involved in search marketing for more than two decades, and what I am proud of is making efforts to understand how search habits change. I have done so by paying through experimentation, software, test domains, and failures that never went anywhere.

Originally, SEO was about endless experimentation. Now, there is mobile SEO, voice SEO, and now AI SEO. I have attended conferences, purchased software, and created small websites for tests just to see what will break. While carrying out these experiments, they seemed unnecessary and almost obsessive.

This is important to me since I deal with finance and health-care customers whose business visibility impacts both their credibility and their bottom line. Anything that impacts how people search results in different outcomes very quickly. My experience in observing trends helps me catch changes early on.

The return is compounding. Something I learned years ago influences my thinking now. Every new search period brings something which is not just a reset but adds to what we have learned before.

So my recommendation is easy. Don’t waste your time on trendy bets. Focus on a particular area, and work hard enough for your previous efforts to benefit you many years later. Almost all my best ideas stemmed from knowing the field when others were still indifferent.

I think that the most significant investment I have ever made was definitely CuraDebt over 24 years. That is not very exciting either. It was founded by me in 2000 when I was 27, and then for many years it was filled with compliance issues, hiring issues, angry clients, you name it.

What sets it apart is the accumulation of trust over time. We have managed to create an A+ BBB rating and well over 1,300 five-star reviews for our services in an environment dominated by fraudulent companies. This is far more important than making stock recommendations.

I believe that people have become overly idealistic about investing today. Wealth is usually earned not through some amazing ideas, but by persevering with something mundane until one becomes skilled at it.

I am particularly proud of investing in responsiveness as a business capability, not just a customer service idea. Building processes that allow fast, informed communication during busy periods required discipline, training and a lot of refinement behind the scenes. It was not flashy, but it changed how problems were handled and how confidence was built.

What makes it stand out is the commercial and human impact combined. Delays and uncertainty create stress, especially when safety or access is affected. A responsive operation protects momentum, reduces friction and helps decisions happen sooner. In technical industries, clarity at the right time is a serious competitive advantage.