You have a product (or an idea for one). You need customers.
Even in the idea stage? Yes. (And honestly, if you’ve built the product and have little to no customers, you’re still in the idea stage.)
It’s actually best that you get your customers during the idea stage rather than building first. But hold on. Let’s not “jump the shark.” Let’s start at the beginning.
A funnel is just a linear process with drop-off between steps.
People enter at the top. Some percentage exits at each stage. Fewer people reach the bottom than started at the top.
That’s it. Nothing fancy.
That’s what we’re building now - even/especially if you don’t have a product yet.
Why This Matters Now
Early founders typically don’t have a funnel problem. They have a “no funnel at all” problem.
They’re doing random acts of building and maybe marketing. Posting on social media. Maybe running some ads. Building according to their own ideas and experiences. If they’re lucky they’re getting occasional sales, but they don’t know why.
The fix isn’t optimizing conversion rates (yet). The fix is building just a basic funnel so you can see where people drop off, doing the activities you need to do, and then focusing your efforts on just ONE part of the funnel at a time. The stats will show you which part.
Here’s how I like to start doing this…
1) Create a Simple Funnel
Before you worry about AARRR metrics (“pirate metrics”) or refreshing your analytics again, start here:
Traffic Source
↓
Landing Page
↓
Capture (email, trial, booking)
↓
Nurture (emails, calls, demos)
↓
Purchase
That’s your first funnel. Five steps at most.
Write down what each step looks like for your business right now. Even if it’s ugly. Even if you’re doing it manually. Even if it barely exists.
At least you now have something to improve.
Now, to throw a wrench into this stereotypical funnel idea, when I did a 1:1 with a client, we customized his approach specifically to his outreach tasks. The idea is similar, just more nuanced.
His funnel looked like this:
Prospects
↓
Outreaches
↓
Responses
↓
Pitches
↓
Purchases
Notice, it’s still very much a funnel, but this one uses specific words for his particular phase of Customer Development.
He was just trying to figure out where his theoretical ideal customer lives online. So, his funnel had to do with finding prospects that fit his filters first. Next, he needed to reach out to them. If they responded, he’d pitch, and then we’d see how many purchases he got.
Simply building this funnel and tracking weekly numbers quickly showed that his biggest problem was in the prospecting area.
He switched between two modes - when he had hard filters on who he would consider a prospect, he would get a good response rate, but the flow was too low to get enough meaningful data to make a decision. But, when he loosened up on the prospect filters, he’d do the outreaches and the response rate would stink.
After doing this for a while, you can see that it gets pretty obvious that you can’t worry about the rest of the funnel if you can’t get enough of the right people into the top of it.
So, he spent some time delegating and automating the research and collection of potential prospects based on his hard filter criteria. Once he had enough flow of good people coming through the prospecting and outreach stages, the response rate got better. Then, and only then, did he move on to understanding whether his pitches were working.
And notice… we’re not even talking about the product yet!
But a simple funnel pointed out where the problems were. We could then ask “why”, diagnose, and fix them.
Info
The specific stage names are less important than having and using a funnel at all.
Demo-Sell-Build: What to Track
Most founders build first. Then demo. Then try to sell.
Do NOT do this unless you're B2B
Build → Demo → Sell
This is backwards.
You spend months or years building. Launch to crickets. Wonder what went wrong. The answer: you validated nothing before committing.
Flip it.
Do this instead
Demo → Sell → Build
Create a demo of the outcome (mockups, video, prototype, spreadsheet, manual service). Sell the demo. If people won’t pay for the promise, why build the product?
Ash Maurya, the Lean Canvas creator, puts it this way:
Quote
“If you can’t sell the demo, why build a product no one wants? Instead, ask why. Fix your demo, sell it, and then build exactly that.”
You can track a lot with a basic funnel - sales calls, customer discovery/learning, marketing ROI, and so much more. So, we start with the basic funnel and track the process of getting demos first.
When you’re first starting out, anything other than what I mentioned above is likely too much, because you’ll be changing your funnel(s) as you learn.
Pirate Metrics: AARRR
Once we have revenue, we shift to a higher-level funnel that uses the “Pirate Metrics” framework, mapping to these stages:
1. Acquisition (Finding) How do people discover you exist?
Traffic sources: SEO, social media, ads, referrals, partnerships, communities, content, cold outreach
Question to answer: Where do my ideal customers already hang out?
2. Activation (First Value) When does someone go from “visitor” to “engaged”?
For SaaS: signup, complete onboarding, hit their first “aha moment”. For services: book a call, reply to an email, attend a demo.
Question to answer: What’s the shortest path to “this is useful”?
3. Revenue (Paying) When do they give you money?
Could be: first purchase, subscription start, deposit, prepayment
Question to answer: What’s the specific action that converts them from free to paid?
4. Retention (Keeping) Do they stay? Do they use it again?
Metrics: churn rate, monthly active users, repeat purchases, renewal rate
Question to answer: Are people getting enough value to keep paying?
5. Referral (Spreading) Do they tell others?
Metrics: NPS, referral rate, word-of-mouth attribution
Question to answer: Would they recommend you unprompted?
Tip
Don’t optimize all five at once. Fix the leakiest bucket first.
2) Find Your Leakiest Stage
Look at each transition in your funnel. Where’s the biggest drop-off (compared to expected/best drop-off)?
The client I mentioned above, his simple sheet looked something like this (data made up):
| Stage | People (Agg) | Conv (Agg) | Exp (Agg) | Diff (Agg) | 2025-10-10 | Conv | 2025-10-17 | Conv |
|---|---|---|---|---|---|---|---|---|
| Prospects | 158 | 35 | 123 | |||||
| Outreaches | 64 | 41% | 100% | -59% | 35 | 100% | 29 | 24% |
| Responses | 7 | 11% | 50% | -39% | 4 | 11% | 3 | 10% |
| Pitches | 3 | 43% | 80% | -37% | 2 | 50% | 1 | 33% |
| Conversions | 0 | - | - | - | 0 | - | 0 | - |
Column explanation & suggestions
Freeze the first 5 columns. Create 2 columns for each week for a predetermined time (a quarter is recommended).
- People (Agg) = Aggregate of all the weeks
- Conv (Agg) = Conversion rate of all the weeks
- Exp (Agg) = Expected conversion rate
- Diff (Agg) = Difference between expected and actual conversion rate
Each week’s “Conv” rate column is just dividing into the stage above it. For example: 2025-10-10’s Responses rate is calculated by dividing 4 (# of responses that week) into 35 (# of outreaches made that week).
The most important column in here is the Diff (Agg) column! The highest negative number is where to focus your time.
Now, each week, you can fill in the raw numbers and the conversion rates and aggregate numbers should fill themselves in using formulas.
Important note
Yes, the more data you have, the more confident you can be in your decision. However, directionality and speed is what we’re after in an early startup.
If you’re wrong with the direction, the process of fast decision making will auto-correct you anyway.
But, if you take too long to make a decision, you could easily run out of money before anything else matters.
Thus, to make decisions quickly, 30-100 conversions at any stage, for any one test, where the actions taken afterward are reversible, will be our starting point for making decisions.
3) Diagnose the Leaks
When something’s not working, here’s where to look:
| Symptom | Likely Cause |
|---|---|
| Traffic but no signups | Landing page/offer mismatch, wrong audience, weak copy |
| Signups but no activation | Confusing onboarding, too much friction, unclear value |
| Activation but no payment | Price too high, value not proven, no urgency |
| Payment but high churn | Product underdelivers, wrong customer, competitor better |
| Paying customers but no referrals | Not delightful enough, no referral system, niche too narrow |
| Here’s some other thoughts. This is not comprehensive. Just a starting point. |
Low landing page → signup conversion
- Your promise doesn’t match what they expected
- You’re attracting the wrong traffic
- Your copy isn’t speaking to their pain
Low signup → activation
- Onboarding is confusing or too long
- Time to first value (time to aha) is too slow
- They don’t understand what to do next
Low activation → payment
- Pricing doesn’t match perceived value
- Free version is too good (or too bad)
- No urgency or reason to pay now
High churn (over 5% monthly for SaaS)
- Product isn’t sticky enough
- They solved the problem and don’t need you
- Competitors are eating your lunch
Low/no referrals
- They don’t love you enough to risk their reputation
- No referral mechanism exists
- They don’t know anyone else with the problem
Again, fix one leak at a time.
4) Weekly Review Habit
The thing is, even after setting all that up… tracking alone doesn’t help you. You need a review process.
Every week (Friday afternoons work well), spend 30 minutes asking:
- What changed? (numbers up, down, or flat)
- Why did it change? (what did I do differently this week? Have a theory.)
- What’s the bottleneck? (where are most people dropping off compared to expectations?)
- What’s ONE thing I’ll test next week? (not three things → ONE)
The funnel shows you where to look. The weekly review helps you decide where to spend your time.
TL;DR Action Steps
If you have no funnel:
- Draw your current customer journey (even if it’s ugly)
- Identify traffic source(s) → how they find you
- Identify capture method → how they give you their info
- Identify conversion → how they pay
If you have a funnel but no traction:
- Measure each step’s conversion rate
- Find the biggest drop-off
- Form a hypothesis about why
- Run one small test to improve that step
- Focus on messaging continuity (ads → landing pages → conversations → emails, etc), what you’re promising them, what outcomes they’ll get, how much they care about solving which problems, what they love/hate/fear, how they want their perfect solution to work, and what emotional words they say when they’re talking about it.
Use Demo-Sell-Build:
- Create a demo of your solution (doesn’t need to work)
- Book calls with 10 potential customers
- Try to sell the demo
- Track: how many conversations → how many yeses
- If no yeses, ask why and fix the offer before building anything else
Related Resources
- Customer Development - How to have the conversations that feed your funnel
- Demo-Build-Sell - The framework for validating before building (coming soon)
- Market Evaluation Framework - Is your market worth building for? (coming soon)
This resource is part of the Founder Labs Resources library. Questions? Drop them in Slack.