There’s this moment in every engineering leader’s career when some exec drops “We need to improve our EBITDA” in a meeting, and you’re sitting there wondering if that’s a new frontend framework or database technology.
Look, the hard truth is that technical excellence alone won’t get you a seat at the table where important decisions happen. You can architect beautiful systems all day long, but if you can’t translate that work into business impact using language leadership understands, you’re screwed.
I spent years nodding along in meetings while secretly Googling acronyms under the table. What a waste of time. The reality is that this stuff isn’t actually complicated – it’s just deliberately obscure to maintain the MBA industrial complex.
So here’s my gift to you: 20 business terms explained in plain language that will make you dangerous in budget meetings. No MBA required.
Revenue Metrics
Let’s start with how companies track their income:
1. ARR (Annual Recurring Revenue)
This is the holy grail metric that VCs obsess over. Take the money customers pay you every month, multiply by 12, and voilà – ARR. $10k MRR = $120k ARR. It’s that simple. If your customers are paying yearly, it’s even simpler. Unlike that distributed consensus algorithm you’re debugging, this math actually IS straightforward. ARR is one of the predictors of “will we all have jobs next year?” so pay attention.
2. MoM (Month-over-Month Growth)
The percentage your metrics went up (hopefully not down) since last month. 10% MoM growth sounds cute until you realize that compounds to ~3x yearly. That’s the difference between “we’re raising another round” and “we’re updating our LinkedIn profiles.” YoY (Year-over-Year) is the same but for years, for when your growth is too depressing to look at monthly.
3. NRR (Net Revenue Retention)
Measures revenue growth from existing customers, including upsells and churn. Over 100% means you’re growing without adding a single new customer – basically printing money. Below 100% means you’re in a leaky bucket, and no amount of new customers will save you. Investors will pay multiples for 120%+ NRR because it means your product isn’t garbage.
4. GRR (Gross Revenue Retention)
Unlike its sexier cousin NRR, GRR ignores upsells and only measures if customers are sticking around. If 9 out of 10 customers bail, but the last one doubles their spend, your NRR looks great (congrats on growing!), but your GRR is a horrifying 10% (everyone hates you). Good GRR means your product is sticky; good NRR means you’re also extracting more value. You want both.
Cost Metrics
Now let’s look at how costs are categorized:
5. COGS (Cost of Goods Sold)
This is what it ACTUALLY costs you to deliver your service – the “if we add one more customer, these costs go up” stuff. For software companies, this includes cloud infra directly tied to usage, support people who handle tickets, and those third-party APIs you’re praying don’t jack up their prices.
Here’s the part most engineers miss: Not ALL your AWS bill is COGS. That k8s cluster you’ve overprovisioned that could handle 10x your current load? The base cost isn’t COGS because it doesn’t scale with each new user. But the S3 storage that grows with user data? Pure COGS. This distinction matters because gross margins depend on it, and VCs obsess over gross margins.
6. OpEx (Operating Expenses)
Everything else you’re burning money on: your salary, that fancy office nobody wants to commute to, the marketing team’s budget for branded socks. Basically, stuff you could theoretically cut without immediately turning off the service. I say “theoretically” because we’ve all seen what happens when companies take this too literally – just look at the “Twitter experiment” where they fired 80% of staff and somehow the site still (mostly) runs — for now…
7. CapEx (Capital Expenditures)
CapEx is what you spend on stuff you could theoretically sell later – physical assets like servers, buildings, or that ping pong table nobody uses anymore. Unlike the money you burn on SaaS subscriptions or salaries, CapEx gets “depreciated” over time, meaning the accountants pretend it loses value gradually rather than being an immediate expense. Yes, this matters for taxes and balance sheets. No, you probably don’t need to care unless you’re running the company.
When you spend money on COGS (Cost of Goods Sold) and OpEx (Operating Expenses), that money is gone – it’s a pure expense that you can’t recover.
But purchasing something like a car works differently. The car is an asset you own. If you decide to sell it after 2 years, you’ll get some money back. However, the car’s value typically decreases over time. That decrease in value is called depreciation, and it’s calculated as part of the OpEx for each year you own the asset.
8. Burn Rate
The speed at which your company is setting money on fire. If you’re spending $250K monthly but only bringing in $50K, you’re burning $200K per month. With $1M in the bank, your runway is 5 months. After that, it’s game over – no more payroll, no more free snacks, no more jobs. This is why founders look increasingly haunted as the runway shortens. When an exec starts talking about “focusing on efficiency,” they’re telling you the burn rate is terrifying them.
Profitability Metrics
How do we measure if a business is actually making money?
9. Gross Margin
After you deliver your service, what percentage of revenue do you actually keep? That’s gross margin. Take your revenue, subtract COGS, divide by revenue, and you get your gross margin percentage. If you make $1M and spend $500K on direct costs, you’ve got a 50% gross margin.
This number varies WILDLY by industry. Hardware companies like Tesla (18%) and driver-dependent services like Uber (33%) have brutally low margins because physics and humans are expensive. Meanwhile, SaaS companies like Figma are printing money at 90% margins because bits are basically free to copy. This is why VCs throw money at software and not hardware – software margins are pornographic by comparison.
OpenAI is a fascinating case with negative gross margins – they’re literally paying for you to use their service. Not a strategy I’d recommend unless you have Sam Altman’s Rolodex of investors.
10. Net Margin
Gross margin is a lie – it pretends your engineers and office space are free. Net margin tells the real truth by accounting for ALL your expenses. Figma might have 90% gross margins, but if they’re blowing billions on marketing, their net margin could still be negative. This is the “are we actually making money or just shuffling VC cash around?” metric.
11. EBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization)
The most obnoxious acronym in finance, and finance loves acronyms. EBITDA is basically “how profitable would we be if we ignored all this annoying accounting stuff?” It strips out interest payments, taxes, and how quickly your assets lose value.
Why this nonsense metric? Two reasons:
- It lets you compare companies that have different tax situations or debt levels
- It’s a way for unprofitable companies to pretend they’re almost profitable.
When a company trumpets they are “EBITDA positive,” translate that to “we’re not actually profitable, but we’re getting closer, please don’t sell our stock.”
12. Default Alive/Default Dead
Paul Graham’s brutal binary: based on current growth and burn rate, will your company reach profitability before running out of money? Default alive startups can breathe – they’ll survive even if the VC money printer stops. Default dead startups are in a race against the clock, desperately trying to grow faster or cut costs before the cash runs out.
This is why the “move fast and break things” era is ending – companies with insane burn rates, but mediocre growth are realizing they’re default dead. No amount of ping pong and cold brew will save them.
On the other hand, a startup might have a high burn rate, but if it has healthy growth and good margins, it might reach a break-even point before it runs out of cash.
Customer Metrics
How do we track customer behavior and satisfaction?
13. CAC (Customer Acquisition Cost)
How much cold, hard cash does it take to acquire one paying customer? If you’re blowing $50K on Google ads and landing 500 customers, that’s $100 CAC. This is the metric that exposes the “growth at all costs” lie – any idiot can buy growth by setting money on fire in the Google and Facebook ad engines. Companies with long-running astronomical CAC are just doing customer acquisition wrong, full stop.
14. LTV (Lifetime Value)
How much revenue will a customer generate before they inevitably leave you? If customers pay $100/month and stick around for 14 months on average, that’s $1,400 LTV. This is pure financial modeling, not reality, but investors will judge you on it anyway.
The golden ratio is LTV:CAC, and you want at least 3:1. Spending $1,300 to acquire a customer worth $1,400 means you’re basically breaking even after COGS and OpEx – congrats on building a sophisticated money-laundering operation! At 4:1 or higher, you’ve got a real business. But if it’s too high (like 10:1), you’re probably not aggressive enough on growth and leaving market share on the table.
15. Retention
Are your customers sticking around or running for the exits? Measured as the percentage that stay with you month-over-month or year-over-year. What “good” looks like varies wildly – B2B SaaS might expect 98%+ monthly retention, while consumer apps would kill for 40%.
Know your industry benchmarks, or you’ll either panic needlessly or miss catastrophic problems.
16. Churn
Retention’s evil twin. If 5% of users bail each month, that’s 5% churn. Sounds small until you do the math and realize you’re losing 46% of your customers yearly. Compounding works against you here – a seemingly modest monthly churn can mean you’re replacing your entire customer base annually. Not great for that LTV formula!
17. NPS (Net Promoter Score)
The “how much do users love us” vanity metric based on a single question: “On a scale of 0-10, how likely are you to recommend us?” Customers scoring 9-10 are “Promoters” (they love you), 7-8 are “Passives” (meh), and 0-6 are “Detractors” (they actively hate you). NPS = % Promoters – % Detractors.
If 40% love you, 30% are neutral, and 30% are writing angry tweets about you, your NPS is 10%. Anything above 0 means more people love than hate you – congrats on clearing that low bar! Above 30% is genuinely good. But remember: customers lie on surveys and happy customers don’t respond, so take NPS with a massive grain of salt.
Real behavior > survey responses.
Strategic Frameworks
Finally, let’s look at how businesses think about their strategic position:
18. AARRR Framework (Pirate Metrics)
The “pirate metrics” framework (because it sounds like “arrr!”) breaks down what actually matters in your product lifecycle:
- Acquisition: How do users find you?
- Activation: Do they have that “holy shit, this is awesome” moment?
- Retention: Do they stick around or ghost you?
- Revenue: Do they pay you actual money?
- Referral: Do they tell friends or keep you as their dirty little secret?
The brilliance is in knowing where your biggest problem lies. Building cool new features for a product nobody discovers is pointless. Adding monetization before users see value is predatory. As an engineering manager, if your team is building “engagement features” when your activation funnel is broken, you’re rearranging deck chairs on the Titanic.
Fix the biggest leak first.
19. TAM (Total Addressable Market)
The theoretical ceiling of how big your business could get if you captured literally every potential customer. It’s the “we’re only capturing 1% of a $10B market!” slide in every pitch deck ever.
VCs obsess over TAM because they need massive returns, not cute lifestyle businesses. TAM is mostly bullshit math – take a huge market, claim some percentage of it could use your product, and voilà.
But if your TAM is genuinely tiny, no amount of execution can create a unicorn. You can’t be the world’s best buggy whip manufacturer when cars exist.
20. Moat
What stops a competitor from copying everything you do and eating your lunch? That’s your moat. Network effects, proprietary data, switching costs, brand loyalty – these are modern moats. In the AI era, data moats are becoming even more critical – the company with the most user data to train on wins.
If your company doesn’t have a clear moat, you’re in trouble. Engineers often think technical sophistication is a moat, but unless you’re doing quantum computing, it probably isn’t – someone will reverse-engineer your approach. Great UX isn’t a moat either unless it’s backed by uniquely deep user understanding.
If you can’t articulate why customers would stay with you even if a cheaper alternative appeared, you don’t have a moat.
Will This Make You Better at Your Job? Absolutely.
Holy hell, we’ve only scratched the surface here. There’s a whole universe of funnel metrics, cohort analysis, cap tables, and other corporate hieroglyphics that business folks throw around. But this covers the crucial 20 that have repeatedly saved my ass when advocating for my team.
Look, I’ve watched brilliant technical leaders get steamrolled in planning meetings because they couldn’t articulate how their work impacted the metrics the company actually cares about. Don’t be that person.
Next time you hear a term you don’t understand, write it down, look it up later, and add it to your arsenal. Do yourself a favor though – don’t pretend to know something when you don’t. Just ask. “I’m not familiar with that term – could you explain it?” works wonders and actually builds respect. Most people are making half this stuff up anyway.
The moment you start connecting your team’s work to metrics like gross margin or COGS reduction, you’ll watch executive eyes light up. Your boring refactoring project suddenly becomes a strategic initiative that “addresses technical debt to maintain our gross margin as we scale.” Same work, completely different reaction.