Before you write your first line of code, answer these four questions.
Not for investors. For yourself.

Here’s what I see every week: founders on meeting 5 with a “potential customer,” or running an unpaid pilot, or doing a “design partnership” where they’re building custom features for free.

They think things are going great.

Meanwhile, the "customer" has no budget. Or won't pay. Or is never going to be a customer at all.

Three months later, the founder realizes they built the wrong thing for the wrong person. Again.

I've watched this happen hundreds of times working with founders. These four questions would have caught it in Week 1.

Most founders aren't stupid. You're not chasing people who don't know they have a problem.

But you ARE wasting time on the wrong customer group.

Here's how customers break down:

They don’t know they have a problem yet.
They’re just “getting by” with duct-taped processes and haven’t noticed how painful the problem really is.

Example: A founder manually updating spreadsheets every week thinking “this is fine,” while losing 10+ hours/week.
(You skip these.)

They know the problem exists — but they’re not trying to solve it.
They complain about the pain but haven’t looked for tools, alternatives, or workarounds.

Example: “Our lead routing is such a mess,” but they’re not Googling solutions or exploring options.
(You skip these too.)

They’re actively trying to solve the problem.
They’re researching categories, comparing approaches, and evaluating different ways of fixing it.

Example: “We need to fix onboarding — we’re comparing tools vs building something custom.”
(This is where many founders get stuck.)

They know the solution landscape (not necessarily your product yet), understand the options, and are narrowing down vendors.
They’re deciding between specific tools, workflows, or partners.

Example: “We’ve looked at three tools, but we’re not sure which direction to go — build vs buy vs partner.”
(This is where most founders really get stuck.)

They have the problem, have budget, have urgency, and are actively moving toward purchase.
They have a deadline, a KPI tied to solving the problem, and money allocated to fix it.

Example: “We’re losing $120K/year on this and will choose a solution this quarter. If you can solve it, we’ll pay.”
(These are the ONLY customers you should be talking to.)

Only Group 5 buys. Everyone else burns your time.

Here’s a founder conversation I had last week:

Founder: “We’re building for the pet care industry.”
Me: “Cool — who specifically pays you?”
Founder: “Well, pet owners want this, and—”
Me: “Wait. I’m confused. If pet care businesses are paying you, why are you talking about pet owners?”
Founder: [pause] “Oh. You’re right. Pet care centers are the ones paying. They’re losing $120K a year on reception admin work.”
Me: “Okay, but WHO at the pet care center controls that budget? The receptionist who has the problem, or the owner who sees the $120K line item?”
Founder: “…The owner.”
Me: “There it is. THAT’S your customer.”

See what happened?
The founder finally identified a Group 5 customer (pet care centers losing $120K)…
but was still talking to the wrong person inside the company.

The person with the problem (e.g. pet care receptionist who is overworked) is often not the person with the budget (e.g. pet care center owner).

You need to target the budget decision maker — the person who can actually approve the spend.

Ask yourself:

If you're saving their employee 20 hours a week but it doesn’t map to a KPI they care about, they won’t buy — even if the end user loves you.

Pricing isn’t guesswork — it’s math.

Start by looking at what they're already paying for their broken solution:

Cost of duct-taped tools:
Add up all the software they’re using to solve this badly.

Cost of human hours:
Hours per week × hourly rate × 52 weeks.

Your price should be less than that total — ideally meaningfully less — so the ROI is instant and obvious.

Example:
If a company is spending $30K/year on tools + $80K/year in human time = $110K total, you can charge $50K/year and it’s still a yes.

Even if it means a 90% discount because you don’t have all the features yet.

Or get them to sign a quantified LOI:
“If you build X, Y, Z features to this level, we will pay $[amount].”

I don’t care about waitlists.
I don’t care about “interested” users.

I care about people who will put money down, even a symbolic amount.
Because that’s the only signal that actually matters.

This is harder — you can’t get quantified LOIs from individual consumers.

But you can test if they’ll pay.

Instead of a free waitlist:
Charge $5–$10 to jump the line
or
Offer a “lifetime deal” pre-order at 50% off.

Tesla did this: $1K deposits for the Cybertruck before it existed.
Thousands paid.

If people won’t pay $5 for early access, they’re almost definitely not paying $10/month later.

A founder proudly tells me they have 1,000 signups.

Cool.
But how often do they use it?

“Well, they signed up…”

I’d rather see 20 beta users who spend 2 hours on your product daily.
That’s someone who will pay you one day.

A million users who log in once a month for 2 minutes?
That’s not a business.

Every successful startup has a "why now" moment. Something shifted that makes this problem urgent today when it wasn't urgent two years ago.

Good answers are TANGIBLE changes from the last 2 years:

✅ Technology breakthrough unlocked economic viability: "OpenAI's API dropped costs 10x in 2024, making real-time voice agents economically viable for the first time. What cost $50K/month in 2023 now costs $5K."

✅ Regulation created forced spending: "EU AI Act passed in 2024 requires companies to audit AI decision-making. Enterprises now have compliance budgets for AI tooling that didn't exist 18 months ago."

✅ Platform/ecosystem shift broke existing solutions: "Twitter/X increased API pricing from free to $42K/month in 2023, instantly breaking every social listening and customer service tool. B2B companies now need alternatives to monitor brand mentions."

✅ Market consolidation opened opportunity: "The top 3 players in [industry] all got acquired by PE firms in 2023-24, and they immediately raised prices 40%. Mid-market customers are now desperate for alternatives."

✅ Critical mass unlocked new capabilities: "Stripe reached $1T in payment volume, making fraud patterns finally detectable at scale. Before this threshold, the data didn't exist to build effective fraud prevention."

❌ Bad answers:

The test: If nothing tangibly changed in the last 2 years, why would someone switch to your solution today instead of waiting another year?

If you want to raise VC money one day, your market needs to be at least $10 billion in total addressable market (TAM).

Here’s why: VCs need one company in their portfolio to return the entire fund. If you raise venture capital, they’re betting you can become a unicorn ($1B+ valuation). And you can’t build a billion-dollar company in a $500M market.

So before you spend months building, make sure the market is actually big enough.

It’s simple. You already figured out your ACV (annual contract value) in Question 1 when you calculated pricing.

TAM = ACV × Total Addressable Customers

From Question 1, you calculated what you'd charge based on:

Let’s say you figured out you can charge $50K/year.

This is NOT “every company in America.” Be realistic.

Ask yourself:

Don’t say “all restaurants.” Say mid-market restaurant chains with 10–50 locations and centralized operations.

Let’s say there are 200,000 mid-market restaurant chains in the US.

$50K × 200,000 companies = $10B TAM

Done.

A founder says: There are 30 million small businesses in the US, so our TAM is…

No. Most small businesses can’t afford you, don’t have this problem, or would never adopt your type of solution.
Be honest about who your real addressable customer is.

You calculated $50K/year as your future ACV when all features are built, but then use $5K/year in your TAM calculation because that’s what you charge now for the MVP.

Don’t do this. Use your target ACV, not your discounted MVP price.

A founder says: We’re starting with Series B SaaS companies, but eventually we’ll sell to all B2B companies, so our TAM is…

Stop. You’re not going to sell to “all B2B companies.” Even Salesforce couldn't capture everyone - they dominated enterprise but lost mid-market to HubSpot and SMB to simpler tools.

Your TAM should include realistic adjacent expansion, not fantasy expansion.

Real example: If you're building CRM for real estate agencies, your realistic TAM includes different sizes of real estate companies (boutique → regional → national brokerages).

Your unrealistic TAM includes "all sales teams" just because they also need CRM. Insurance sales teams, SaaS sales teams, and car dealership sales teams have completely different workflows.

Rule of thumb:
Stop expanding your TAM once you reach a customer segment that already has 5+ recently VC-backed early-stage competitors or 10+ mature players serving it well.
At that point, you’re not going to take meaningful share — don’t include that segment just to make your TAM bigger.

A $500M TAM can still be a great business.
You can bootstrap, raise from angels, or use alternative funding.
But don’t raise VC if the economics don’t support unicorn-scale.

Now that you know your market is big enough, let's validate that it's real.

Here's how I think about competition:

0 competitors at seed/Series A stage:

Either you're a genius who found a massive white space, or the market is too small and that's why no one VC-backed is building this.

I'm usually betting on "market too small."

Think about it: If this is such a huge problem in a $10B+ market, why hasn't anyone else tried to solve it? Maybe the unit economics don't work. Maybe customers won't actually pay. Maybe it's harder than you think.

Zero competition can be a red flag.

2-3 competitors at seed/Series A stage:

Perfect. This tells me there's a real market, VCs are interested in the space, but it's not oversaturated yet.

You can still win with a better approach, different wedge, or superior execution.

This is the sweet spot.

10+ competitors doing exactly what you're doing:

Too crowded. You're going to spend all your time and money fighting for scraps. The differentiation better be massive - not just "we're 10% faster" or "our UI is better."

Unless you have an unfair advantage (proprietary data, exclusive partnerships, technical breakthrough), you're going to struggle.

Step 1: Search for seed/Series A companies

Go to Crunchbase, AngelList, or just Google:

Step 2: Look for companies that raised in the last 3 years

Older than that and they might be dead or pivoted. You want to see recent validation that VCs think this market is real.

Step 3: Understand their approach

Don't panic if competitors exist. Ask:

If they're doing something different enough, you might both be right.

Then you need to answer honestly:

Talk to 10 potential customers and ask questions like:

These questions force people to quantify pain and urgency.
Their answers will tell you whether you're onto something — or chasing a low-priority problem.

If you can't answer these four questions clearly and specifically right now, don't write code yet.

I know that's painful. You're an engineer. You want to build.

But here's what I see in my inbox every day: founders who spent 6 months building the wrong thing because they didn't answer these questions first.

I'm not saying this to gatekeep. I'm saying it because I've watched brilliant engineers burn months building the wrong thing beautifully.

You wouldn't architect a system without understanding the constraints. So don't architect a startup without understanding the customer.

And clarity doesn't come from thinking harder in your head. It comes from testing faster in the real world.

The difference between a fundable company and a side project:

Fundable companies solve urgent problems for Group 5 customers (who can actually pay) at the right time (something tangibly changed), in a market that's real (2-3 competitors) and big enough (at least $10B TAM).

Side projects are "cool ideas that Groups 3 and 4 say they'd maybe use someday."

Look, I don't really care how beautiful your architecture is. I don't care about your 200-person waitlist. I don't care how long it took you to build your prototype.

I care about whether someone will actually pay you money.

And these four questions tell me if they will.

Get clarity on these:

If any of these feel fuzzy, you’re not ready to build yet.

Next drop: How to validate these answers by getting people to commit money, including the customer interview script that surfaces real demand and how to turn those conversations into quantified LOIs.

— Christine

P.S. Reply to this email with your answers to these four questions. I read every response and I'll share the most common mistakes in next week's drop.