Angel Investing Transparency: How We Track Our 38+ Investments
A rare look inside a 38-company angel portfolio: the tracking system, vintage analysis, wins, losses, and the decision framework that evolved over seven years.
Whether you're looking for an angel investor, a growth advisor, or just want to connect — I'm always open to great ideas.
Get in TouchAI, startups & growth insights. No spam.
TL;DR: Most angels hide their track record. I've decided to do the opposite — publishing the full structure of how I track 38+ investments across six years, including 6 wins, 5 losses, and everything in between. This post walks through the exact spreadsheet, the vintage analysis, what the wins had in common, what killed the losses, and the decision framework I use today versus what I naively used when I started in 2019.
Let me start with an uncomfortable truth: most angel investors are lying to you.
Not lying exactly — but curating. You see the wins on LinkedIn. You see the portfolio company logos on their website, carefully selected for name recognition. You hear about the $50M Series B they got into at the seed stage. What you never see is the company they quietly wrote off last quarter, the investment they made at a terrible valuation because a friend pressured them, or the sector thesis they abandoned after losing two bets in a row.
This curated self-presentation has three effects, all harmful:
First, it distorts the expectations of aspiring angels. When new investors only see the highlights reel, they underestimate the loss rate. The reality is that in a professionally managed early-stage venture portfolio, you should expect 40-60% of investments to return little to nothing. If your personal loss rate is lower than that, you either haven't been investing long enough, or you're not taking enough shots.
Second, it prevents the community from learning. Angel investing is an apprenticeship craft. The fastest way to improve is to study what actually happened — not just the successes, but the failures in granular detail. When successful angels don't publish their losses, they deprive the next generation of the most valuable information.
Third, it creates selection bias in deal flow. When you're known as someone who only publishes wins, the founders who approach you know they're getting a polished version of you. When you're known as someone who talks openly about failures, you attract founders who respect intellectual honesty — and those tend to be the founders worth backing.
I started publishing notes about my investments more openly about two years into my journey, and the quality of my deal flow changed immediately. Founders appreciated that I was willing to say "I made a mistake here" publicly. It signals a certain kind of intellectual honesty that the best founders find attractive.
Beyond individual reputation, there's a systemic problem with the opacity of angel investing as an asset class. Institutional limited partners evaluating venture funds can access audited return data. Angels operating individually have no equivalent standard. The result is that bad actors can claim excellent track records with no verification, and good actors who have genuinely strong returns get no signal advantage over those who are simply good at marketing themselves.
There is a slowly growing movement toward voluntary transparency among angels — publicly sharing portfolio structures, check sizes, return assumptions, and methodology. I want to contribute to that. This post is my attempt to be as specific and honest as I can be about what 38+ investments over six years actually look like.
I'll share the structure of every investment: stage, sector, geography, check size range, current status, and mark-to-market methodology. What I'm not sharing are company names for active portfolio companies — because that creates complications for those companies with other investors and acquirers. For companies that have exited (in either direction), I'll describe the patterns without identifying information that would be identifiable to insiders.
I made my first angel investment in early 2019. I had sold a SaaS product, had some capital, and had a founder friend raising a pre-seed round who asked if I wanted in. I had no framework. I had no diligence process. I wired $10,000 and thought I was a venture capitalist.
That company, by the way, is still alive. It's one of the ones I'm most uncertain about in terms of outcome.
2019 — Learning year (3 investments, $10K-$25K average check)
My first year was entirely opportunistic. I invested in three companies through warm introductions from founder friends. Check sizes ranged from $10,000 to $25,000. I had no sector thesis, no stage preference, no geographic view. I was just trying to figure out what I was doing.
The selection criteria in 2019 was embarrassingly simple: did I like the founder? That's it. No market sizing, no unit economics review, no reference checks. I was pattern-matching to "people I'd want to work with" and calling it a strategy.
2020 — Thesis formation (7 investments, $15K-$50K average check)
2020 changed everything. The pandemic compressed timelines. Founders were building faster, markets were shifting faster, and deal quality — ironically — improved because the tourist investors pulled back. I increased my deployment pace significantly and started developing an actual thesis: B2B SaaS tools for distributed teams, with a bias toward founder-market fit in remote-native contexts.
I made seven investments in 2020, my busiest year to that point. Check sizes grew as I got more comfortable with the asset class. I also started doing basic due diligence — reading cap tables, reviewing customer reference letters, asking for cohort retention data.
2021 — The euphoria trap (9 investments, $25K-$100K average check)
2021 was the year I overpaid. It was also the year I made some of my best investments — because valuations were high, but so was the quality of founders coming into the market. The mistake I made in 2021 was not adjusting my expected return math for the compressed entry valuations. I was accepting lower ownership percentages than I'd accepted before without adjusting my conviction threshold upward.
I made nine investments in 2021. At least two of them I would not make today, because I let the FOMO of the moment override my framework. This is a lesson I return to repeatedly: market euphoria is an especially dangerous time to invest, not because good companies aren't being built, but because the pricing environment means the same company needs to succeed more dramatically to generate the same return multiple.
2022 — The correction filter (4 investments, $25K-$75K average check)
The market correction of 2022 was clarifying. I slowed dramatically — only four investments. The companies I backed in this period were founders I had been tracking for twelve or more months, companies that had survived the market shift and were growing through it. This cohort is, so far, one of my best-performing vintage years on paper.
2023 — Rebuild and focus (6 investments, $25K-$75K average check)
In 2023, I narrowed my thesis significantly to AI-native B2B tools. I had been an early observer of the GPT-3 and GPT-4 inflection points through my work in AI product development, and I believed we were at the beginning of a productivity wave that would advantage companies building natively on large language models. Six investments in 2023, all in this space — though studying why AI startups fail shaped which teams I was willing to back.
2024–2026 — Current phase (9 investments and counting, $50K-$150K average check)
My check size has grown because my conviction threshold has grown. I'm making fewer investments than I did in my peak deployment years, but I'm writing larger checks into fewer companies. I've also started doing more follow-on investments in existing portfolio companies — something I did almost zero of in my first three years.
| Period | Average Check | Max Check | Deals/Year | Total Capital |
|---|---|---|---|---|
| 2019 | $17,500 | $25,000 | 3 | ~$52K |
| 2020 | $28,000 | $50,000 | 7 | ~$196K |
| 2021 | $48,000 | $100,000 | 9 | ~$432K |
| 2022 | $42,000 | $75,000 | 4 | ~$168K |
| 2023 | $45,000 | $75,000 | 6 | ~$270K |
| 2024–26 | $85,000 | $150,000 | 9 | ~$765K |
The growth in check size reflects three things: more available capital as earlier investments appreciated on paper, higher conviction from a more developed framework, and the practical reality that good deals at the pre-seed stage are now more competitive, requiring larger checks to get meaningful pro-rata rights.
The thing I wish someone had told me in 2019: the tracking system is not optional. It is the investment. Without it, you are not building a portfolio — you are making a series of disconnected bets that you will struggle to learn from.
My tracking system has evolved through four major iterations. The current version lives in a Notion database with a Google Sheets mirror for quantitative analysis. Here is every field we track.
| Field | Type | Notes |
|---|---|---|
| Company name | Text | Full legal name + DBA |
| Slug | Text | Short identifier for internal use |
| Investment date | Date | Date wire sent, not signed |
| Check size | Currency | Actual amount invested |
| Pro-rata rights | Boolean | Do we have the right to follow on? |
| Follow-on invested | Currency | Total follow-on capital deployed |
| Stage at investment | Select | Pre-seed / Seed / Series A |
| Instrument | Select | SAFE / Priced / Convertible Note |
| Valuation cap | Currency | For SAFEs and notes |
| Discount rate | Percentage | For convertible instruments |
| Post-money valuation | Currency | For priced rounds |
| Estimated ownership | Percentage | Modeled at conversion |
| Lead investor | Text | Who led the round |
| Round size | Currency | Total round, not just my check |
| Field | Type | Notes |
|---|---|---|
| Founder names | Text | All co-founders |
| Founder backgrounds | Text | Prior companies, roles, exits |
| Founder-market fit score | 1-10 | Scored at investment, not updated |
| Reference check completed | Boolean | Did we run references? |
| Reference check notes | Long text | Summary of calls |
| Founder communication rating | 1-5 | Updated quarterly from update quality |
| Field | Type | Notes |
|---|---|---|
| Sector | Multi-select | Primary + secondary |
| Geography | Select | HQ country |
| Total addressable market | Currency | At investment time estimate |
| Business model | Select | SaaS / Marketplace / Services / Other |
| Revenue model | Text | How they charge |
| ARR at investment | Currency | Actual ARR when we wrote the check |
| MoM growth at investment | Percentage | Trailing 3-month average |
| Unit economics at investment | Text | LTV:CAC ratio if known |
| Field | Type | Notes |
|---|---|---|
| Current status | Select | Active / Acquired / IPO / Dead / Zombie |
| Last update received | Date | Most recent investor update |
| Update quality | 1-5 | Transparency, metrics shared |
| Last known ARR | Currency | Most recent figure from updates |
| Fundraising status | Select | Closed / Raising / Not raising |
| Last round raised | Currency | Most recent capital event |
| Last round valuation | Currency | Most recent post-money |
| Mark-to-market value | Currency | See methodology section below |
| Mark methodology | Select | Last round / Write-down / Write-off / Acquired |
| Red flags | Boolean | Any active concern |
| Red flag notes | Long text | Detail on concerns |
| Field | Type | Notes |
|---|---|---|
| Total invested | Currency | Check + follow-ons |
| Current mark | Currency | Current estimated value |
| MOIC (paper) | Calculated | Current mark / total invested |
| Cash returned | Currency | For exits only |
| Exit MOIC | Calculated | For exits only |
| Exit date | Date | For exits only |
| Exit type | Select | Acquisition / IPO / Secondary / Wind-down |
This is where most angels are the least rigorous, and where I've put the most effort. My methodology:
Last round pricing: If the company has raised a priced round in the last 18 months, I mark at that round's post-money valuation times my ownership percentage. This is the standard approach.
Stale round (18-36 months ago): If the last priced round was 18-36 months ago, I apply a 25% haircut to reflect time passage, market conditions, and the real risk that the paper valuation overstates current enterprise value.
Stale round (36+ months ago): 50% haircut. Many of these are effectively zombie companies — not dead, not growing.
No revenue or revenue declining: Write-down to 25% of last round valuation regardless of timing.
Active wind-down or confirmed dead: Write to zero.
Acquisition: Mark at actual cash received or at signed deal value if pending close.
The temptation is always to mark generously. Resist it. A mark-to-market that flatters you is useless for decision-making. The only person harmed by an inaccurate mark is you, because it distorts your portfolio health view and leads to worse capital allocation decisions.
| Stage | Count | % of Portfolio | Avg Check | % of Capital |
|---|---|---|---|---|
| Pre-seed | 22 | 57.9% | $32,000 | 38.2% |
| Seed | 13 | 34.2% | $56,000 | 39.8% |
| Series A | 3 | 7.9% | $110,000 | 18.0% |
The pre-seed bias is intentional and reflects my belief that founder-market fit is easier to evaluate than traction metrics — and that the best entry points are before the market has priced in the obvious narrative. The 5 checks I run before writing a check explains exactly how I score founder-market fit at this stage.
| Sector | Count | % of Portfolio | Notes |
|---|---|---|---|
| B2B SaaS | 14 | 36.8% | Broad; includes vertical SaaS |
| AI-native tools | 9 | 23.7% | Post-2022 investments predominantly |
| Developer tools | 5 | 13.2% | Overlaps with AI tools |
| Marketplace | 4 | 10.5% | Two-sided; mix of results |
| Consumer | 3 | 7.9% | Most underperforming category |
| Fintech | 2 | 5.3% | Both active |
| Other | 1 | 2.6% | Hardware-adjacent |
The consumer allocation is my biggest regret from a sector composition standpoint. Consumer products at the pre-seed stage have fundamentally different risk profiles than B2B, and my evaluation framework was poorly suited to assess them. I backed them because the founders were exceptional — which was correct — but the structural disadvantage of consumer at early stage overwhelmed the founder quality advantage in two of three cases.
| Region | Count | % of Portfolio |
|---|---|---|
| United States | 24 | 63.2% |
| India | 7 | 18.4% |
| Europe | 5 | 13.2% |
| Southeast Asia | 2 | 5.3% |
The India allocation reflects my personal network advantage there — I understand the market, the founder archetypes, and the local competitive dynamics in a way I don't for most other geographies. I've learned to lean into edge rather than chase the geographies everyone else is focused on.
| Vintage | Investments | Write-offs | Paper MOIC | Notes |
|---|---|---|---|---|
| 2019 | 3 | 0 | 1.4x | Too early to conclude; all active |
| 2020 | 7 | 1 | 2.8x | Best vintage so far; correction filter |
| 2021 | 9 | 2 | 1.6x | Overpaid; valuation compression effect |
| 2022 | 4 | 1 | 3.2x | Correction-year discipline paid off |
| 2023 | 6 | 1 | 2.1x | AI thesis beginning to pay |
| 2024–26 | 9 | 0 | 1.3x | Too early; mostly paper marks |
The pattern is clear if uncomfortable: the vintages that were deployed into uncertain or declining markets outperform the ones deployed into enthusiastic markets. This is not a coincidence.
In 2020, the uncertainty of the early pandemic filtered out weak founders. Only founders with genuine conviction — who were building because they had to, not because it was fashionable — were raising. Valuations were reasonable because FOMO had temporarily left the room.
In 2022, the correction did the same thing with a different mechanism. Founders who continued building through the market downturn, and who were still growing despite tighter conditions, were demonstrably resilient. The price discipline I applied in 2022 was higher than in any other year.
In 2021, I paid peak prices for companies that would have been good investments at 2020 valuations. The founders were often excellent. The terms were not.
The best time to be an active angel is when everyone else is scared. This is easy to say and extremely hard to do. The social pressure to pull back when markets decline is enormous. The data from my own portfolio is the clearest possible argument against giving in to that pressure.
My 2023 investments are concentrated in AI-native tools, and they are the hardest to evaluate right now. The space is moving so fast that companies that seemed well-positioned in early 2023 are facing competitive dynamics in 2026 that didn't exist when I invested. Some of these companies have adapted. Some have not. I expect this vintage to be bimodal — a few large wins and several losses — which is consistent with the thematic concentration.
I've had six investments that I would classify as wins: four acquisitions and two companies that have raised significant follow-on capital at markups that represent clear realized or near-realized value. I'm not naming companies, but the patterns are highly consistent.
In five of six wins, the founder had either worked in the exact domain they were building in for a minimum of five years, or had previously founded a company in an adjacent space. They weren't just interested in the problem. They had lived it professionally and had developed an asymmetric information advantage about customer needs, distribution channels, or technical constraints.
The one win that didn't fit this pattern was a founder with no direct domain experience who compensated with exceptional network density in the target buyer segment. She knew every potential customer personally and could convert those relationships into early pilots. Different path, same outcome: a distribution advantage that competitors couldn't easily replicate.
Every single win in my portfolio had at least one early customer story that surprised me. Not "we got the obvious first customer." But a customer who adopted the product in a way the founder hadn't anticipated, using it for something slightly different than the designed use case, and that unexpected use case pointed toward a larger market.
This pattern — what I now call "customer surprise" — is one of my leading indicators. Founders who can articulate a surprising customer behavior they discovered in the first 90 days of selling have usually found a real pain point. Founders who describe their early customers as exactly what they expected worry me.
The first year of revenue is often misleading. Founders are exceptional at convincing their networks to become early customers. The signal comes in year two, when you've exhausted the network effect and have to sell cold. In four of six wins, the MoM revenue growth rate in the second year was within 20% of the first year. The market was pulling the product, not just the founders' personal relationships.
In three of six wins, the founder turned down a term sheet at some point — usually a Series A offer with terms they didn't like — and waited for better conditions. These founders had conviction in their trajectory and the financial discipline to not take dilutive capital just because it was available. It is the kind of founder decision-making that separates long-term builders from those chasing the fastest raise. Every one of those founders eventually raised at terms substantially better than the ones they passed on.
Every winning founder in my portfolio improved their investor update quality between year one and year three. The updates became more focused on a smaller number of metrics, more honest about what wasn't working, and more decisive about what they were going to do about it. This correlates, in my view, with founders who are genuinely learning from the business rather than managing investor perception.
In every acquisition case, the acquirer was someone the founder had a prior relationship with or a company that had been on the founder's strategic partnership list from early on. This is not coincidence. Founders who know from the beginning who might buy them build toward acqui-hire relationships. The exits didn't happen because someone discovered the company — they happened because the founder had been building the relationship for years.
Five of my 38+ investments are formally written off. Several more are what I'd call "zombies" — alive but unlikely to return capital. Here's what the losses had in common.
In three of five write-offs, the company pivoted significantly in the first 18 months — but the pivot came six months too late. The founder knew the original product wasn't working but continued investing in it out of a combination of sunk cost bias and investor pressure. By the time they pivoted, they had burned too much runway on the wrong thing and didn't have enough capital to validate the new direction.
The lesson I've taken from this is to look for evidence that a founder can kill their own ideas quickly. During diligence, I now ask: "Tell me about something you tried in this company or a previous one that you shut down, and how quickly you shut it down." The founders who can answer that question specifically and calmly are demonstrably different from the ones who struggle to name anything.
Two write-offs were in markets that were clearly going to exist but didn't exist yet at the scale required to build a venture-scale company. The founders weren't wrong about the direction of the market. They were wrong about the speed. And early-stage startups don't have the runway to wait for a market to arrive.
This is the hardest loss category for me, because these were often excellent founders with correct vision who simply needed the market to be three years further along than it was. In retrospect, I should have done more rigorous analysis of the specific triggers required for mass adoption and been more skeptical of timelines that depended on behavioral change at the consumer or enterprise level.
In two of five write-offs, the cause of death was co-founder conflict. In both cases, the signs were visible in retrospect: ambiguous equity splits that hadn't been fully negotiated, divergent views on company direction that both founders acknowledged in diligence but dismissed as "healthy tension," and different levels of commitment that became stark after the first rough quarter.
Co-founder conflict is the single most underrated startup risk. I now ask explicitly: "Have you had a serious disagreement with your co-founder in the last six months? Walk me through it." If the answer is "we rarely disagree," that's actually a yellow flag — either they're not honest with each other or one founder is clearly dominant and the other isn't fully bought in.
One write-off was a marketplace with unit economics that seemed improvable through operational efficiency but turned out to be structurally constrained by the nature of the market. No amount of execution improvement was going to get the LTV:CAC ratio to a sustainable number because the supply side of the market had too much power and too little lock-in.
The lesson: distinguish between unit economics problems that are early-stage immaturity (real, can be fixed) versus structural market problems (will not get better). The way to do this is to model the unit economics at the end state — not assuming growth will fix it, but assuming the competitive dynamics of a mature version of this market, with incumbent suppliers, commoditized demand, and normalizing margins.
Counter-intuitively, one of my cleanest losses came from a company that raised an unusually large seed round in 2021. The excess capital allowed them to build before they had learned. They hired aggressively, built features their customers hadn't asked for, and came to Series A with a product that was technically sophisticated but commercially confused. They ran out of options before they could recalibrate.
There is a real phenomenon in startup investing where too much early capital is as dangerous as too little. The best companies I've seen are ones that had to be extremely capital-efficient in the early months, because scarcity forced clarity about what mattered — a dynamic explored in depth in profitable growth without burning cash.
| Status | Count | % of Portfolio | Notes |
|---|---|---|---|
| Active / Growing | 19 | 50.0% | Regular updates, positive trajectory |
| Active / Uncertain | 9 | 23.7% | Update quality declining or metrics mixed |
| Zombie | 5 | 13.2% | Alive but no meaningful growth |
| Acquired (Win) | 4 | 10.5% | Successful exits |
| Wind-down (Loss) | 5 | 13.2% | Written off |
Note: percentages exceed 100% because some companies moved between categories during the year and are counted in their current status.
I don't publish specific absolute numbers because they change with every mark. What I can share is the MOIC distribution:
The classic power law distribution is visible. The 4 companies with 4x+ marks account for a disproportionate share of the portfolio's total paper value. The diversification strategy post explains what this power law distribution means for how many companies you need to back.
One of the best leading indicators of portfolio company health is the investor update. I track this religiously:
The correlation between update frequency and company performance is not perfectly causal, but it's strong enough to be actionable. Founders who are growing have good news to share and share it. Founders who are struggling avoid the communication. When a company goes silent, I treat it as a yellow flag and reach out proactively.
If all three were yes, I wired money.
The current framework is organized around five dimensions, each scored 1-10:
1. Founder-market fit (weight: 35%) Does this specific person have an asymmetric advantage in building this specific company? Evidence can include domain expertise, distribution relationships, prior failed attempts in the space (seriously — prior failure in a domain often produces the best founders), or technical capabilities that competitors can't easily match.
2. Market timing (weight: 20%) Is this the right moment? I look for three signals: a technology enabler that has recently crossed a viability threshold, a behavioral shift that has occurred in the last 2-3 years (not a shift that will happen), and a regulatory or structural change that has opened a new space.
3. Business model integrity (weight: 20%) Does the unit economics math work at scale? I build a simple model: at 1,000 customers, what are the margins, the payback period, the capital required to grow, and the competitive moat? If the math doesn't work at 1,000 customers, it probably doesn't work at 10,000.
4. Reference quality (weight: 15%) What do former colleagues, customers, and (if available) prior investors say? I run a minimum of three reference calls for any check above $50K.
5. Momentum signal (weight: 10%) What has happened in the last 90 days? Are there specific milestones that indicate the business is moving? A company that just signed its first three paying customers is in a different position than one that signed its first customer six months ago and has signed no others.
| Total Score | Action |
|---|---|
| 8.5 - 10.0 | Invest at full check size; request pro-rata rights |
| 7.0 - 8.5 | Invest at reduced check size (50-75% of standard) |
| 5.5 - 7.0 | Pass, but stay in relationship; revisit at next raise |
| Below 5.5 | Pass with no follow-up |
The most important change from my 2019 approach to my 2026 approach is not the framework itself — it's the discipline to apply it even when social pressure pushes toward yes. The investments I most regret are ones where a friend pressured me to invest, the opportunity looked exciting, and I skipped one or more elements of the framework to move quickly. Every time I've done that, I've regretted it.
If you're an angel investor with fewer than 10 investments, start simple. You don't need the full schema I described above. You need enough structure to answer three questions at any moment:
Start with a simple spreadsheet with these columns:
| Field | Why You Need It |
|---|---|
| Company name | Obvious |
| Investment date | For calculating hold time |
| Amount invested | For capital accounting |
| Stage | For return expectation calibration |
| Instrument + cap | For modeling ownership |
| Current status | Active / Acquired / Dead |
| Last update date | Forces you to notice silence |
| Current mark | For portfolio valuation |
| Notes | For anything that doesn't fit |
Once your portfolio reaches 10+ companies, add:
Every quarter, I do a two-hour portfolio review. The agenda:
The last item is the most important and the most frequently skipped. The learning only compounds if you articulate it.
The tool matters far less than the discipline. A perfect system you use inconsistently is worth less than a simple spreadsheet you update religiously every quarter.
I define a win as any investment that has returned or is highly likely to return more than 3x my invested capital in cash (not paper). This threshold is higher than break-even but lower than the "venture return" standard that institutional funds target. For an individual angel writing checks from personal capital, a 3x cash return on a single investment is a meaningful outcome, and I think it's intellectually dishonest to only count 10x+ outcomes as wins.
By this definition, I have 4 confirmed wins (acquired companies with cash exits) and 2 near-wins (companies with priced rounds that put my paper mark above 3x and are actively on a path to liquidity).
For the four acquired companies in my portfolio, the average time from my initial investment to acquisition close was 4.3 years. This is consistent with the broader venture data that suggests the typical early-stage exit timeline is 5-7 years, though acquisitions at the seed stage tend to happen faster than IPO paths. I expect some of my 2020-2022 investments to reach liquidity events in the 2026-2028 window.
I have strong pro-rata rights in about 40% of my portfolio companies. I exercise those rights selectively — only in companies where my conviction has increased since the initial investment, not simply maintained. The fact that I have a right to invest more doesn't mean I should. Follow-on capital is a fresh decision evaluated with the same rigor as an initial investment.
In practice, I've exercised pro-rata rights in about half the cases where I've had them. The companies where I chose not to follow on have performed worse on average than the ones where I did, which suggests my conviction signal was calibrated reasonably well.
I care about valuation — but I care about it differently at different stages. At pre-seed, where there's often no revenue and minimal traction, the valuation is primarily a signal about the founder's self-assessment and negotiating posture. I'm wary of founders who push very hard on valuation at pre-seed without commensurate traction, because it suggests either overconfidence or a focus on optics over fundamentals.
That said, I've made investments at pre-seed valuations ranging from $3M to $12M cap, and the outcomes have been more correlated with founder quality than with entry valuation at that range.
This is the hardest practical challenge in angel investing. When you invest in a friend's company and it's struggling, the personal relationship creates pressure to be supportive rather than honest. I've learned to separate my role as a friend from my role as an investor very explicitly. In my investor role, I owe the founder honest assessment. In my personal role, I owe them emotional support. Conflating the two helps neither.
I've also had to recuse myself from providing advice in a few situations where I had invested in a direct competitor. This is uncommon but requires immediate transparency when it arises.
All investments are made through a single-member LLC for tax efficiency and liability management. This is not advice — consult a tax attorney — but I'd strongly recommend any angel doing more than 2-3 investments per year evaluate the entity structure question early rather than retrofitting it later.
Investing in the first ten opportunities they see without establishing a thesis. The first few deals you see will seem exceptional because you have no comparison set. Building a comparison set — watching 50 or more opportunities before investing, or investing small amounts in the first year explicitly as a learning exercise — dramatically improves your judgment. The investors I've seen fail fastest are those who came in with significant capital and deployed it all in the first year.
Badly, at first. Better now.
The honest answer is that a write-off hurts, especially when you liked the founder and wanted the company to succeed. The way I've gotten better at managing it is to separate outcome quality from decision quality. A bad outcome doesn't necessarily mean a bad decision — early-stage investing is fundamentally probabilistic. If I made a good decision with the information I had at the time, and the company still failed, that's the nature of the asset class.
What I don't accept is a bad outcome from a bad decision. When I look back at my losses and see that I skipped steps in my framework or ignored red flags because of social pressure, that stings more than the financial loss — because it was avoidable.
There's no universally correct answer, but my working recommendation is: watch 30-50 opportunities before writing the first check, then write the first 3-5 checks at the smallest size you can negotiate (often $5K-$25K at pre-seed), explicitly framing them as tuition. The goal of the first year is not to generate returns. The goal is to develop a framework, build your deal flow network, and make the expensive mistakes with small amounts.
Two things: the update quality score and the founder communication rating.
Most angels track whether a company is alive and what its latest valuation is. Very few track the quality of how the founder communicates with investors over time. In my experience, this is one of the best leading indicators of company health and founder development. A founder who starts writing terrible quarterly updates and transitions to excellent ones over two years is a founder who is genuinely learning. A founder who was great at investor communication in year one and has gone increasingly dark by year three is usually navigating something difficult they don't want to surface.
The second thing I track that most don't: my original conviction score versus the current trajectory. Comparing what I thought about a company at investment to what the evidence shows now is the most direct way to improve my judgment over time.
Transparency in angel investing isn't just a virtue signal. It's a practice that makes you a better investor. When you commit to sharing your track record — the real one, not the curated one — you create accountability to yourself that forces honesty in your marking, your learning, and your decision-making.
Thirty-eight investments in. A lot more to learn. But the system is working, and I'm confident enough in it now to share the whole thing.
The exact pre-investment checklist I run before every angel check: founder-market fit, timing, unit economics, references, and momentum — with scoring tables and decision matrix.
A post-mortem across 38+ angel investments: the 3 patterns behind every win, 3 patterns behind every loss, and the risk framework that changed everything.
Data-driven framework for angel portfolio construction — optimal portfolio size, stage mix, sector focus, check sizing, and follow-on allocation by power law.