Mitchell Keller

LeadGrow.ai

We help sales teams become predictably busy with programmatic outreach.

Co-Founder & Head of Growth

MACH4 Studios

Co-Founder

Ontario Power Generation

Nuclear Energy Worker

I'm actively building the most efficient system to go from product market fit to message market fit using outbound and inbound led outbound and I'm having a blast doing it. Basically my team will find the people you want to contact no matter how hard they are to find, and ideally ones that have a clear publicly available issue to pray o- i mean help them with. hahaha. 

It's simple, empower sales teams with top of funnel that drives revenue, acquisition value and fatter fundraising rounds. 

Here's some case studies: From Rought US market entry to 2m in popeline in 40 days:https://gamma.app/docs/From-Market-Entry-to-2M-Pipeline-How-LeadGrowai-Transformed-a-Top-gp4lui8pe6n30v4?mode=doc 

https://gamma.app/docs/From-1-meeting-a-quarter-to-320K-in-closed-revenue-12m-in-pending-j3zinck91npcgmv

Some more sweet stuff on my youtube: https://www.youtube.com/@MitchellKeller_ 

If you believe your sales team should spend more time crushing and less time data crunching then send me a message. 

P.S. We tend to work with extremely high ticket companies, companies with tiny hard to find markets and seed to series A funded looking to take their product market fit and find an assymetry that scales it to the moon.

Our Solutions  *Ask about our work for free performance guarantee ⚡️

✅ Organic Social: We drive quality traffic to your social channels, to generate exponential awareness to your brand, so you can draw in further investments, partners & have the community base for successful IDOs & INOs.

✅ Paid Social Advertising: We create grandslam offers & dynamic engagement-focused design to structure powerful campaigns that generate leads & increase conversions. 

✅ Web3 Consultations: We leverage our 60+ IDO's of experience to provide you endless value.


✅ Website Design & Copy: We research your brand & its competitors to generate compelling copywriting & design that is purpose-built to create the best customer journey possible. Through this we turn web traffic into valuable leads & conversions for you to increase ROI. 

✅ NFT Art: We conceptualize & create world-class 2D & 3D art that enhances your brand’s value to Web3 consumers & turn them into loyal fans that will buy your NFTs, while you sit back & focus on refining your business & utilities. 

✅ Pitch Deck Design & Copy: We do all the market research, copywriting & design for your pitch deck to be concise and tell your value story in a straightforward manner, so you can raise money effectively.


✅ Brand Design & Copy: We identify your brand voice & create value-focused copy that is empowered by dynamic design. Helping you connect with your target customer more deeply & generate lifetime value. 


✅ Video Production: We create premium 2d & 3d narrated promotional videos that provide a fantastic first impression of your business. Turning profile views into active followers for you. 

✅ Community Management: We manage & encourage participation in events to cultivate loyal communities. By gamifying user experiences we are able to provide daily value to community members & generate authentic discussions to turn members into fans. All of which creates the perfect community for you to sell anything too.

The legend 🦾Eric Nowoslawski posted about the cost savings of prompt caching last week and how it's actually an automatic process. 

Decided to do a quick test on a prompt we use a lot to quickly validate if someone still works at a company and collect basic up to date info on a prospect. 

On the left we have the inputs at the bottom of the prompt to promote more caching. Basically you want the input to stay the same for as long as possible without changes. 

my 100b token appreciation award from openai is gonna be about 10-15% cheaper. Across all my AI usage this should be about 500-700 in savings per month. 


PS here's the prompt AND the json output for those who want it: 

LinkedIn Employment Verification Prompt
<task> Verify if a contact still works at their last known company by checking their LinkedIn profile's experience section for current employment status and collect profile info. </task> <context> You are analyzing LinkedIn profiles to determine current employment status. The key indicator is whether their experience shows "to present" for the company we're checking. </context> <instructions> 1. Navigate to the LinkedIn profile at {{linkedin_url}} 2. If the page doesn't exist or is invalid, return status with null values for all data fields 3. Locate the Experience section on the profile 4. Find the most recent work experience entry 5. Check if this experience shows "to present" (indicating current employment) 6. Compare the company name in this current experience with {{company_name}}: - Look for exact matches or highly similar names (e.g., "Acme Corp" vs "Acme Corporation") - Account for common variations (Inc., LLC, Ltd., etc.) 7. Determine employment status and sub-status: - If the current experience (showing "to present") matches {{company_name}}: still_at_company = true, sub_status = "still_at_last_company" - If the current experience is a different company: still_at_company = false, sub_status = "changed_company" - If no current experience shows "to present": still_at_company = false, sub_status = "open_to_work" 8. Extract ALL required data regardless of employment status: - Full name from profile - Current job title (or last known if unemployed) - Current company name (or last known if unemployed) - Employment date range (e.g., "Jan 2022 to present" or "Jun 2021 to Dec 2023") - Company LinkedIn URL (search for company page) - Company domain (from company's LinkedIn page or profile) 9. Return all data in the specified JSON format with consistent field structure </instructions>
- LinkedIn URL:  - Last Known Company: </input>

B2B Growth ⊚ Turn strangers into customers with Situation-based outbound 👉 present like a technician not a maintenance man 🗓️over 2230 meetings booked in 2025.

We launched over 3600 total campaigns this year. 

Averaging 1 lead for every 175 contacts across all campaigns, losers and winners. 

This blows my mind. 

LeadGrow.ai has come a long way this year thanks to amazing early team member additions like: 

Charles Adriel Del Rosario
Eli Raterta
Rashi Agarwal
Nikolaos Malamis

And by deciding to be an open book with other awesome agency owners to grow together. 

Shoutout especially to Kellen Casebeer, Troy Aitken, MBA and Nick Konstantinidis

Super sick to see - and with claude code being a core part of our future work. 

I can't wait to 10x the number of campaigns next year.

EDIT: ALSO Nico Marino, The best the greatest, the moustacheiest - constant inspiration for myself and team and only wasn't added cause he wasn't an agency owner so would've need an inconvenient standalone line and I'm lazy

Building proposals -> drafting content -> creating SOPs and writing copy 

BRRRRRRRRRRRRRRRRRRRRRRRRRRRRR

No alternative text description for this image

This is what great consistent work looks like in outbound. 

We're getting about 1 lead per 74 contacts to strictly cold audiences for a K-12 edtech Seed-stage SaaS with no 'high intent' signals - and we're just asking for a signup or a call.

This is message-market fit. But it takes extreme hustle, creativity, and iteration.

We've launched and tested >45 campaigns across edtech. (shoutout to the impeccable work by Nikolaos Malamis). 

We had early winners with ridiculous 60% PR rates, but kept testing.

Recently we discoved an angle that hits 73% PR for certain markets. 

👉 How did we get here and how did we keep improving? 

It started with Regulatory changes, and vast organizational rollouts. 

These types of huge paradigm shifting events change the situation of your target customer - leverage them. 

The education sector is heavily regulated (good cause it forces change, and change usually requires help) and VERY public with data.

Additionally vast rollouts often bring with yet to be discovered issues (like getting a consult from a business owner a step ahead of you). 

We're using this knowledge as arbitrage to display extreme understanding of their current situation and offer up a simple fix. 

We wrapping the tools core features around the issue at hand. It sounds simple, but it takes a lot of work. 

here's what we're doing: 

We started with broad deep research across 8 different distinct markets. Each getting a campaign targeting: 
1. Bottom up (teachers, temp staff)
2. Middle out (IT, Curricular Leaders)
3. Top down (super intendendents, principals)

Then we layer in GTM intelligence:

🔧 Take everything that's working and download it to Claude Code
🔧Use Jacob Dietle's EmailBison MCP to extract maximum context on every campaign and split test.
🔧Pair this with our Slack MCP to pull context about sales calls from our client. 
🔧Run about 2,000 web searches covering every geography down to the state and province, and sometimes even deeper if the AI thinks it's worth it (based on our scoring system). 

What are these searches for? 

They discover the other markets we can apply the same winning formulas to, and our learnings from closed-won deals.

Within an hour we had our next 8 campaigns planned and ready, complete with a winning strategy to split test against in a new market and a brand new strategy based on that market's situation.

This is why GTM intelligence is important. It lets us do things in a day that used to take months.

If you're a Seed or Series A founder that's hands on with GTM, I would love to chat with you about it.

B2B Growth ⊚ Turn strangers into customers with Situation-based outbound 👉 Sell technicians not a maintenance men.

add /products.json to the end of any shopify store link and thank me later.

If you aren't this locked in, I aint watching your content

Shoutout: Jacob Dietle

How to fail your gtm as a founder

Don't run feedback campaigns to your target customers that engage with content related to your offer on social media

Don't record sales conversations

Don't use closed won recordings for lookalike campaigns and messaging changes

Don't post content that tells a story and defintiely don't reach out to engagers

Defintiely fall in love with the first draft of your offer and try to bend the world around it instead of listening to market feedback.

Do the opposite and win.

Stop looking for lookalikes. Start looking for outcomes.

Most agencies build target lists the same way:

"We closed a real estate company. Let's find more real estate companies."

That's lookalike targeting. And it leaves 90% of your market on the table.

Here's a better question: What pain did you solve?

A buddy of mine built a custom AI tool that helps sales teams practice calls. Like Hyperbound, but internal. His first client was a real estate media company planning to hire 50 reps in 2026.

His instinct? Go find more real estate companies.

My advice: Go find any company hiring 10+ sales reps this quarter.

Because the pain isn't "real estate." The pain is "we need to get new reps ramped faster."

That pain exists at every company with aggressive hiring plans. SaaS, manufacturing, financial services, doesn't matter.

How to find outcome-based targets:

Step 1: Identify the pain you solved, not the industry you served.
- "Ramp time for new sales hires"
- "Pipeline visibility for leadership"
- "Compliance documentation for audits"

Step 2: Find public signals that indicate that pain.
- Job postings with specific keywords
- Funding announcements (usually means hiring spree)
- Regulatory deadlines
- Technology migrations

Step 3: Build a Clay workflow to surface companies showing those signals.

For the sales training example:
- Pull companies with 10+ job postings matching "sales" in the last 90 days
- Enrich with company size and industry
- Filter for companies large enough to afford custom work

Why this works:

Normal cold outreach converts maybe 3% of your list. They happen to be in-market right now.

Outcome-based targeting can convert 30-50%. Because you're only reaching out to companies actively experiencing the problem you solve.

The signal IS the qualification.

You're not asking "might they need this?" You're finding proof they already do.

Build your list around the outcome. Not the industry. Not the company size. The pain.

That's where the meetings come from.

Less than 24 hours into sending and 2 meetings on the calendar.

Stop wasting $10K on conferences talking to students.

Most companies light money on fire at events. $5K for the booth, $3K for travel, 3 days of networking roulette.

Result? Business cards from vendors, conversations with students, maybe one warm lead if lucky.

Here's what we do instead:

Pre-book every meeting before the event starts.

We help a data center company attend AI conferences. They used to walk around hoping to meet the right people. Now they show up with 20-40 meetings already scheduled.

How we do it:

Step 1: Scrape the attendee list

Most events have portals or apps where attendees register. That data is scrapeable. We pull names, companies, job titles.

For events without public portals, monitor LinkedIn for the event hashtag. People post "excited to attend [conference]" for weeks before. That captures another 25% of attendees.

Step 2: Enrich and filter

Not every attendee is a buyer. Run them through LinkedIn to check experience level, decision-making authority, company fit.

Students, vendors, and competitors get filtered out. What remains: actual buyers attending the same event you are.

Step 3: Reach out with event context

"Saw you're attending [Event]. We'll be there too. Would love to grab coffee and discuss [relevant topic]. Does Tuesday at 2pm work?"

Events make cold outreach work because the context makes you real. You're not a random person in their inbox - you're someone they'll actually meet.

The numbers:

Normal cold email: 3-5% reply rate on a good day.

Event-contextualized outreach: 15-30% reply rates. Sometimes higher.

We've booked meetings with MongoDB, Unity, Asana, Databricks, Airbus using this approach. Companies that ignore cold email respond when there's a shared event.

One client got 48 meetings from a single conference. 37 of those were in-person conversations at the event itself.

The evaluation framework:

Before registering for any event, ask:
1. Can I access the attendee list?
2. Are decision-makers actually attending, or just practitioners?
3. Is the event big enough to justify the scraping effort?

If attendee data isn't accessible, reconsider attending.

The bottom line:

Events are only expensive if you show up unprepared.

With pre-booking, you turn a $10K gamble into a $10K pipeline machine.

Stop hoping for good conversations. Engineer them.

taplio