What's up, it's Zayd

Last Tuesday, I almost threw my laptop out the window.

I spent 8 hours consolidating every piece of customer data we've ever collected at Valley into one massive analysis; sales calls, churn records, expansion data, everything.

After a while it felt like a total time suck, but by the end I realized that the actual resource drain was…everything else I did that day.

83% of our revenue was coming from customer segments we weren't even targeting.

We'd been running our entire go-to-market strategy based on vibes and assumptions.

Our "ideal customer" was a fantasy we'd invented in our heads.

Forwarded this email? Get Enabled in your inbox each week.

Zayd’s Picks

My favorite finds of the week

• 3 golden rules for LinkedIn Cold DMs (link)

• How to generate demand on autopilot (link)

• Market segmentation insights for b2b startups (link)

• Why your cold emails are landing in spam (link)

• Your cold emails should make people feel good (link)

The Problem With How Everyone Defines Their ICP

Every founder has the same story.

(It’s me. I’m founder.)

You build a product. You think really hard about who should buy it. You write down some demographics. Maybe you add firmographics if you're feeling fancy.

"B2B SaaS companies, 50-500 employees, Series A to C, using LinkedIn for outbound."

Congrats. You just described 47,000 companies and told me absolutely nothing useful.

Here's what actually happened at Valley:

We assumed our best customers were "B2B businesses whose audience is on LinkedIn."

While that technically may be a true statement, it’s also completely useless.

That description is so broad it includes everyone from recruiting agencies to enterprise SaaS companies to consulting firms to my cousin's fractional CFO practice.

The 8-Hour Deep Dive That Changed Everything

It’s going to require 4 nitro cold brews and 3 David bars to get through, but here's what you need to pull together:

From your CRM and wherever else you're tracking things:

• Every sales call recording from the last 12 months

• List of every inbound signup (emails)

• Every customer who bought (emails + website data)

• Every customer who churned

• Every customer who expanded

• Customers with longest retention

The enrichment part:

Get the full picture beyond the contact info…

• Actual job profiles of buyers

• Their LinkedIn profiles

• What their company actually does (not just what their website says)

• Industry and category

• Revenue and funding stage

• Company age

I used Freckle.io for this, but use whatever enrichment tool actually works.

Then dump everything into Claude or a GPT.

What The Data Actually Revealed

I asked the AI to find patterns across five dimensions:

1. Which customers stayed longest? What was their ACV, LTV, industry, and what did they actually do day-to-day?

2. Which customers churned fastest? Same breakdown.

3. Which job profiles bought? Which ones expanded?

4. Conversion rates on sales calls—total, by channel, by user category.

5. What value propositions resonated with each segment?

The results were...humbling.

Our happiest, highest-paying, longest-retaining customers were:

1. Marketing and growth agencies

2. GTM and lead gen agencies

3. Consulting companies

4. Recruiting and staffing orgs

5. Tech startups

In that exact order.

Lead gen agencies specifically were spending 6x more on Valley than other segments.

Six times more. That is ~so~ many times…

The Math That Changed Our Strategy

If lead gen agencies spend 6x more than the average customer, I can justify spending 6x more to acquire them, right?

I guess technically that’s the math, but I never actually paid attention in math class, so who cares. What I didn’t understand is why we'd been treating them exactly the same as every other customer segment in our marketing.

Same messaging. Same campaigns. Same effort level.

We were leaving money on the table because we never bothered to look at who was actually paying us the most.

Enter: Carol (my new best friend)

I love you Carol.

There's this essay by Jason Cohen about "selling to Carol." 

The idea is simple: your best customer segment deserves a name.

Carol.

Because once Carol has a name, you can't help but think about her more specifically than just “mid-market B2B SaaS company.”

What does Carol care about? What are Carol's deal-breakers? What made Carol decide to buy in the first place?

For us, Carol runs a lead gen agency. She's got 15-30 employees. She's juggling multiple client accounts. Her team is manually researching prospects and writing outreach messages for hours every day.

Her biggest pain points:

• Time spent on manual research (3+ hours per day per rep)

• Inconsistent message quality across her team

• Client churn when results drop

• Difficulty scaling without hiring more people

Her deal-breakers:

• Tools that require technical setup

• Platforms that don't integrate with her existing stack

• Solutions that feel "too automated" (clients want white-glove service)

• Anything that risks LinkedIn account safety

The inciting events that make Carol buy:

• A client asks why response rates dropped

• She realizes she's spending more on SDR salaries than revenue from the account

• A team member leaves and takes their personalization approach with them

• She sees a competitor using better tools

Now we know exactly how to talk to Carol. And more importantly, where to find her.

The Tactical Playbook

Once you know who your Carol is, here's what changes:

1. Product Development: Your roadmap is now based on Carol's value points and deal-breakers. If Carol doesn't care about a feature, deprioritize it. If something's a deal-breaker for Carol, fix it immediately.
   

2. Marketing Copy: Stop writing for "B2B companies." Write for Carol specifically. Use her language. Reference her inciting events. Address her actual concerns.
   

3. Sales Strategy: Find churned customers who are actually Carols. Call them with a specific win-back offer addressing why they left. Build different pricing tiers based on Carol's willingness to pay versus other segments.
   

4. Customer Success: Onboarding flows customized for Carol's specific use case. Support documentation written for her problems. Success metrics aligned with what she actually cares about.

Don't do this if you're below $500K ARR.

You don't have enough data yet. You need at least 50-100 customers to find real patterns. Below that, you're just finding anecdotes, but once you hit that threshold, this becomes non-negotiable.

The Full Implementation Checklist

Week 1: Data Collection

• Export all sales call recordings

• Pull customer lists from CRM

• Get payment data from Stripe

• Compile all marketing copy and positioning docs

Week 2: Enrichment

• Run full enrichment on every customer email

• Get LinkedIn profiles, company details, firmographics

• Organize by cohort: retained, churned, expanded, new

Week 3: Analysis

• Feed everything into Claude or GPT

• Ask the five key questions above

• Document patterns across segments

• Calculate LTV and CAC by segment

Week 4: Implementation

• Name your best segment (Carol)

• Create Carol's profile: pain points, deal-breakers, inciting events

• Rebuild marketing messaging for Carol specifically

• Prioritize product roadmap based on Carol's needs

The Bonus Moves Nobody Talks About

Extract Objection Patterns:

Go through your sales call recordings. What objections come up repeatedly? Those become FAQ content, landing page copy, and product improvements.

For Valley, we kept hearing: "Will this get my LinkedIn account banned?"

Now it's addressed in our positioning, our demo, our website, and our product architecture.

Find Your Anti-Carol: Who churned fastest? Who was the worst fit? Create an anti-persona.

Build filters to keep these people out of your funnel entirely. Seriously. Some customers cost more to support than they'll ever pay you.

Map The Inciting Events: On every sales call, ask two questions:

1. "What made you book this call? What happened exactly?"

2. "Where did you first hear about us?"

The first tells you the inciting event and what made them solution-aware.

The second tells you how they found you specifically.

Track these in a simple spreadsheet. Patterns emerge fast.

For Valley, the top three inciting events:

• Received a Valley message from someone else (personalization proof)

• Had their LinkedIn account restricted (safety concern)

• Posted something that went viral and couldn't capitalize on the engagement (signal opportunity)

These became our entire marketing strategy.

How can we work together 🏔️

1. See more of Valley’s messaging examples, feel free to roast them: https://coolmessagebro.com/

2. Generate more demos for your company using LinkedIn: https://meetings.hubspot.com/zayd-from-valley/tryvalley
   

3. Become a Valley partner and get 20% recurring commission for every user you bring in: https://withvalley.notion.site/valley-affiliate-partner-program