TechBreakdowns 8 Rules For Finding Quality Tech Stocks

Over the years I have toyed around with various rules to find the best stocks out there. You can quantitatively screen all day long, but it’s the qualitative stuff that gets hard to grep. Well, thanks to AI, you can now run your qualitative rules against hundreds of companies for a few dollars!

So, I did exactly that.

I’ve finalized 8 core rules to surface potential winners. Let’s break them down using a duel of SaaS titans: Snowflake and Datadog.

If you subscribe to the newsletter, there’s a download at the bottom of this post with more than 500 names as of December 30, 2025. Premium subscribers, you can find frequent updates on the Premium page.

The Rules

Rule 1. Founder Led

This is one of my favorite rules. While it feels anecdotal, data from firms like Bain & Co. consistently shows that founder-led companies significantly outperform the broader market index.

Founders take risks and look for long term outcomes, while hired CEOs tend to lean a little more towards the near term and quarterly earnings optimizations.

Snowflake: Fail. Frank Slootman has left, and the new CEO is also a non-founder.

Datadog: Pass. Olivier Pomel is still at the helm.

Rule 2. Is the company #1/#2 in its primary market?

We want dominant companies. So, we look at their primary market and check the data to see if they’re ranked among the top companies. If #1 or #2 it’s an obvious pass.

Snowflake: Fail. While massive, they technically fail the "Monopoly" test. They sit in 3rd place (~17% share) behind the hyperscalers themselves—AWS Redshift (27%) and Google BigQuery (19%). They are fighting a multi-front war.

Datadog: Pass. They are the undisputed #1 in cloud observability. With $2.6B+ in revenue, they dwarf competitors like Dynatrace and have effectively become the default "central nervous system" for modern cloud stacks.

Rule 3. SBC discipline

As an investor, I like when my money grows. It’s harder to grow if the company is consistently diluting your stake.

For this rule we’re checking if stock based compensation (SBC) is less than 20% of revenue.

Snowflake: Fail. This is the ugliest part of the report. Snowflake spent a staggering ~37% of its trailing revenue on stock-based comp. For every dollar they bring in, nearly 40 cents goes out the door in equity. It’s an employee-enrichment machine first, and a shareholder vehicle second.

Datadog: Caution. Datadog isn't perfect here either, clocking in at ~23% of revenue. However, it is significantly more disciplined than Snowflake.

Rule 4. Data Gravity

Data is key in the economy of the future, so we’re checking for it here. This is a 0-10 score that we give out that asks “does the product get better/stickier the more data users put in?”

SCORING:

• 9-10: Strong data moat explicitly mentioned, clear network effects
• 7-8: Some data stickiness, mentioned in filings
• 5-6: Moderate - some accumulated value but not core
• 3-4: Weak - mostly a tool, data is portable
• 1-2: No data moat - easily replaceable

Snowflake: Score: 9/10. This is their strongest moat. With customers storing 10+ petabytes of data, the "gravity" is immense. The Snowflake Marketplace (3,500+ listings) creates a network effect where the data becomes more valuable because it can be shared instantly with other Snowflake customers.

Datadog: Score: 8/10. While they don't store "records" like a database, they store "context." With 750+ integrations, Datadog holds the history of every crash, error, and performance metric. Removing it means losing the institutional memory of how your infrastructure actually works.

Rule 5. Switching Costs

This is a “high” or “low” rule. We want to know if it is difficult for a user to just up and leave. It’s easy for a consumer to switch from Pepsi to Coke, but it’s not easy for an enterprise to rip up and reestablish well built systems.

HIGH if:

• Product is "mission critical" or "embedded in workflows"
• Requires system integration (not just a login)
• Switching would be a 6+ month migration project
• Data/history loss would be painful

LOW if:

• Simple web app with easy export
• Could switch to competitor in an afternoon
• No deep integration required
• Commodity product

Snowflake: High. The friction here is financial and technical. Egress fees (paying to move data out) are punitive, and proprietary features like "Snowpark" mean you can't just lift-and-shift your code to another provider.

Datadog: High. The report describes a Datadog outage as a "Code Red" event. If a company turned it off, they would be flying blind. Replicating the custom dashboards and alerts for a large enterprise is a multi-quarter nightmare.

Rule 6. Pricing Power

We’re looking for the ability to raise prices while maintaining stable customer retention. This usually manifests in earnings calls, or sometimes in 10-K / 10-Q filings.

Snowflake: Pass. Despite the competition, they boast a 125%+ Net Revenue Retention rate. Customers aren't just staying; they are spending significantly more every year, proving Snowflake can command a premium.

Datadog: Pass. They maintain a 115% retention rate and have successfully pushed up-market. The "$100k+ ARR" customer cohort is their fastest growing segment, proving large enterprises are willing to pay six-figures for the platform.

Rule 7. Rule of 40

The rule of 40 takes revenue growth and FCF margin. It sums them both up and, if the number is over 40, that company gets a pass.

Snowflake: Pass (Score: ~47). They are growing at 29% with a ~18% FCF margin. Respectable.

Datadog: Strong Pass (Score: ~55). This is where the quality difference shows. Datadog is growing almost as fast (26%) but doing it with significantly higher profitability (29% FCF margin). They are simply the more efficient machine.

Rule 8. R&D Efficiency

Is R&D spending translating to revenue growth?

There are a couple of ways to go about this, but I tend to follow the following rules when making a call:

PASS if:

• Revenue growth > R&D % of revenue (efficient)
• Or high growth with reasonable R&D spend
• Clear product innovation pipeline

FAIL if:

• High R&D spend (>20%) with low growth (<10%)
• R&D appears to be maintenance, not innovation
• No new products despite spending

Snowflake: Fail. They are spending nearly 40% of their revenue on R&D, yet revenue growth has decelerated to 28%. They are running harder just to stay in place against the hyperscalers.

Datadog: Pass. They spend roughly 29% of revenue on R&D, but that spend is birthing entirely new profitable product lines like Cloud Security and LLM Observability. Their innovation engine is actually expanding their TAM, not just defending it.

The Data

Below is a download of the December 2025 run of this score across all of technology. Subscribers to the newsletter (free) can download it.

A warning though! The rules were run via AI. AI can and will make mistakes. So, if you see something you like, you should double check those rules.

We run all 500 stocks relatively cheaply to surface ideas and quickly check assumptions. The best use of this data is from a high level, find your company of interest, then go deeper.

That’s how we do things here at TechBreakdowns. We run the data to find interesting edge cases, then dive in to find the actual details, bull cases, bear cases, and everything else in between.

If you’re interested in going deeper on the companies this surfaces, consider signing up for our Investor+ membership. With that you’ll get frequent updates to the scores, more in-depth AI reporting, the TechBreakdowns portfolio, and more as time goes on including a complete AI reporting app which should go live in January or February.