A lightweight framework for founders finding traction without analysis paralysis
The Problem: Pre-PMF and early-stage founders waste time building complex attribution systems before they have enough signal to justify them. You're tracking 47 metrics when 6 would tell you if this is working.
The Solution: Track only what changes your next decision. For most pre-PMF companies: how people heard about you, what made them convert, whether they stay, and what they'd pay more for. That's it.
The Impact: Spend your time talking to customers and iterating product—not building dashboards. Make faster decisions with clearer signal. Know when you've actually found something worth scaling.
Implementation: Simple spreadsheet framework you can set up in 2 hours and maintain in 20 minutes per week.
You don't need attribution sophistication. You need clarity on what's working.
Most pre-PMF founders build the attribution system they think they'll need at $10M ARR—not the one that helps them get to $500K first. They spend weeks setting up Segment, building UTM taxonomies, connecting data warehouses, and creating Mixpanel dashboards before they have 50 customers.
The mistake: Treating attribution like a technical problem when it's actually a decision-making problem.
Attribution exists to answer one question: "Should we do more of this or less?" If you can't answer that with your current data, your attribution is broken. If you can answer it, your attribution is sufficient—even if it's a spreadsheet.
Before PMF, you're not optimizing. You're searching. Your attribution needs are completely different from a company with proven channels and predictable conversion rates. You're trying to figure out if anyone wants this thing, not optimize a 2.1% conversion rate to 2.3%.
Stop tracking everything. Start tracking what changes decisions.
Track: Single response when you ask "How did you hear about us?"
Why it matters: Tells you which channels have signal worth investigating
How to track it: Ask every customer. Literally. "How did you first hear about us?" Write it down.
Don't bother with: Multi-touch attribution, UTM parameters, journey mapping
Track: Specific moment/message that moved them from "maybe" to "I'll try this"
Why it matters: Your real value prop lives here, not in your positioning deck
How to track it: "What made you decide to try this today?" during onboarding call
Don't bother with: Analyzing every page view, session recording everything
Track: Points of confusion or near-abandonment before they converted
Why it matters: These are your biggest conversion leaks. Fix them first.
How to track it: "Was there anything that almost made you not sign up?"
Don't bother with: Heatmaps, scroll tracking, A/B testing with 12 visitors
Track: Day 1, Day 7, Day 30 usage. That's it.
Why it matters: If they don't come back, nothing else matters
How to track it: Simple cohort view: did people who signed up 7 days ago use it again?
Don't bother with: Weekly active users, engagement scores, complex retention curves
Track: The one feature that retained users touch. Not all features—the one.
Why it matters: Your product is not 17 features. It's the 1-2 things people can't stop using.
How to track it: What's the common behavior among people who came back 3+ times?
Don't bother with: Feature adoption dashboards, usage analytics for every button
Track: Verbatim responses to "What would make this worth $X more per month?"
Why it matters: Shows you what problem actually has budget attached
How to track it: Ask it. Out loud. To every customer. Write down exact words.
Don't bother with: Pricing surveys, conjoint analysis, feature valuation models
These are real things that matter—later. Not now.
| Ignore This | Why It Doesn't Matter Yet | When It Actually Matters |
|---|---|---|
| Multi-touch attribution | You don't have enough volume for statistical significance | After 1,000+ customers/month from multiple channels |
| Channel-specific CAC | Sample sizes too small to be meaningful | When you have 100+ customers per channel |
| Conversion funnel optimization | You need product-market fit, not conversion rate optimization | After consistent 20%+ monthly growth for 3+ months |
| Engagement scoring | You don't know what "good" engagement looks like yet | After 6+ months of retention data |
| Cohort LTV projections | You don't have cohorts old enough to project from | After your oldest customers hit 12+ months |
| Marketing mix modeling | Requires scale and multiple proven channels | After $5M+ ARR with 3+ mature channels |
The pattern: If it requires sophisticated math or large sample sizes, you don't need it yet. If it answers "Should we do more of this?" with your current customer count, you need it.
Everything you need to track in a single Google Sheet. No engineering required.
Tab 1: Customer Source Log
Tab 2: Weekly Retention Check
Tab 3: Value Signals
Maintenance: 20 minutes per week. That's it.
Every Monday morning: Update retention checks, add any new customer responses, look for patterns. If you're spending more than 20 minutes, you're overthinking it.
Data doesn't matter unless it changes what you do.
Look at: Customer source patterns
If you see: 8 of your last 10 customers came from Twitter
Decision: Spend 10 more hours on Twitter content this week
Don't: Build a content calendar, hire an agency, launch 5 channels
Look at: Conversion friction responses
If you see: 6 people said "I almost didn't sign up because I didn't understand the pricing"
Decision: Rewrite pricing page this week
Don't: Run pricing surveys, test 8 variations, build pricing calculator
Look at: Value signals and what retained users do
If you see: Everyone who stays uses Feature X, nobody uses Features Y and Z
Decision: Make Feature X 10x better. Kill Y and Z.
Don't: Add Features A, B, C because they seem cool
Look at: Retention + willingness to pay + referral behavior
If you see: 40%+ come back Day 7, multiple people asked to pay more, 3+ referred others
Decision: You might have something. Double down.
If you see: <30% Day 7 retention, nobody refers, lukewarm on paying
Decision: You don't have it yet. Keep searching or pivot.
There are real signals that you've outgrown the simple system.
You see: Same person mentioned two different sources—heard about you on Twitter but signed up via Google
What this means: You now have enough volume that single-touch doesn't tell the full story
What to do: Add simple last-touch tracking via UTM parameters
Don't jump to: Full multi-touch attribution platform ($50K/year)
You see: Getting 100+ signups/month from 3+ different sources
What this means: You have enough data to start calculating channel-specific metrics
What to do: Add simple CAC by channel (spend divided by customers)
Don't jump to: Marketing mix modeling or incrementality testing
You see: Your team can't agree on which channels are working
What this means: Spreadsheet is no longer single source of truth
What to do: Set up basic analytics tool (Google Analytics 4 or Amplitude free tier)
Don't jump to: Data warehouse, reverse ETL, customer data platform
General rule: Upgrade when the current system can't answer your most important question. Don't upgrade because you think you "should" have better attribution.
Before: Pre-PMF SaaS founder with 40 users, zero paying customers. Spent 6 weeks setting up Segment + Mixpanel + Google Analytics 4. Had beautiful dashboards showing... nothing useful. Couldn't answer basic questions like "which marketing is working?"
Founder was tracking 73 different events, spending 8 hours/week on analytics, and had zero clarity on whether the product had traction.
The lesson: Simple attribution that answers "what's working?" beats sophisticated attribution that shows you everything.
If you're stuck in analysis paralysis or can't tell if your early signal is real, let's talk.
We help pre-PMF and early-stage founders build the right measurement frameworks for their stage—lightweight systems that inform decisions without becoming a second job.
Stop building dashboards. Start finding customers who can't live without this.
Free Growth Session - Let's Talk