Godmode: Disabled

After 2 years of operations, we've decided to sunset Godmode and shut down the company. It's time to log off and restart.

Even with 3+ years of runway left, we're returning capital to our investors. Sometimes, the boldest move is knowing when to start your next chapter.

What happens next

You'll have access to Godmode until May 1, 2025. Export your spreadsheets by clicking the Export button at the bottom of your screen. We will be cancelling your subscriptions after this date.

Questions or just wanna chat? Our in-app chat is still open. We're here until the end.

Why we're shutting down

Two reasons drove this decision:

• First, we failed to crack enterprise sales. Replacing 50% of an AE's daily workflow created tension in our deal cycles. Nobody likes feeling replaceable - a lesson we learned the hard way.

• Second, we realised Godmode wouldn't grow into the venture-scale company we envisioned. Our technology was impressive and opinionated. Nobody asked us how we are different than competitor X after seeing a demo. Our customers found value. However we wanted to spend our time doing our life's work on something with greater impact.

The decision was one of the hardest I've had to make. However to quote Ben Horowitz: If you are gonna eat shit, don’t nibble.

The team

The most valuable asset Godmode had was its team, the incredible people who took a chance on building this with me:

• Our CTO and long-time friend Bertan Teberci: Atlas of our whole backend. Carried our devops, 30+ API integrations, AI pipelines - all by himself. We usually played catch-up to him. Features I designed would be integrated in less than a day.

• Frontend lead Gaston Che: Backbone of our frontend. Built the main features users loved and integrated never-seen-before AI capabilities. I'd write a message at 3am and get a response in 10 minutes.

• Frontend and design engineer Can Cotel: Incredibly conscious of every design choice. Even an off-putting word or bad button color bothered him. Why? Because our users deserved the best.

• GTM guy Hussein Abou Eita: The startup go-to-market guy. Always cheerful. Hard to admit, but customers liked talking to him more than they did to me. He harassed me daily to snatch customer calls I was meant to have.

What we built

Godmode's value was simple: help salespeople react prospects with the right messaging at the right time.

We built what we called a "reverse Google." When you search for COOs looking for new office space, Google shows you SEO garbage from WeWork competitors. You wanted actual people with actual needs. We found them by scouring the web for relevant information. We identified demand signals. Then we gave you tools to send personalized outreach based on AI research.

We sent a maximum of 50 emails and 25 LinkedIn invites per day per account. Not being fans of the cold email blast method, we designed Godmode as an agent that identifies companies with needs only you can solve.

Then makes the match.

This solved two problems:

1. It worked as an invisible marketplace connecting supply and demand in B2B software sales.

2. Salespeople could drop the volume game and focus on their highest quality leads every day.

Here's how it worked:

• You set your ideal customer profile along input with company info and the problems your product solves.

• You wake up the next day to 50 new leads. Emails verified. Profiles enriched. Account research done. Outreach messages written and ready to send.

• If you didn't abort or confirm, Godmode would send the messages both on email and Linkedin, just like an intern would. This kept the agent working without human input and time.

The challenges we faced

So why shutdown? After all we had options. Continue with existing capital for 3 more years. Pivot our approach. Change our messaging. Tweak our pricing.

Our sales cycles taught us hard lessons about market positioning:

Godmode operated in the same space as Clay and Apollo, arguably the two most known products in the space.

But there was a key difference: Godmode wasn't just a tool.

The cool thing is that in the last 6 months, we never received the dreaded question of: “How is this different than X?”. It was dead clear.

Clay and Apollo gave salespeople tools to do tasks such as account research, enrichment, write outreach, and send it.

Godmode actually did them on autopilot.

We completed those tasks end-to-end as a workflow. Daily. We delivered the work and just asked for reviews and edits. This is why we called our main feature "Associates."

However, with enterprise customers, we hit walls:

• Automating 50% of a salesperson's job created tension. Our product threatened existing roles rather than complementing them.

• If a human sales rep makes a mistake, it's considered human. Tolerated.

• When Godmode agent made a mistake? Not tolerated. This nuance was crucial and something we should have seen coming.

With SMB customers, we initially grew fast, only to hit a different wall:

• Our differentiation from traditional products in the space was completing outbound workflows on autopilot every day. Each email unique - the pinnacle of personalisation. Each message based on research so it didn't sound like slop.

• However, most demos ended with one positive and one negative remark:

  • Negative: "I will buy it after trial if it works."

  • Positive: Customers said yes to our pricing - which was 10-100x more expensive than Apollo or Clay.

    • This was good news! People were willing to pay way more for us than they did for other tools.

"If it works" rarely meant getting positive replies or meetings booked. It meant sales. Actual hard revenue. They needed prospects to turn into customers.

When they didn't, we were to blame. After all, we were automating the whole job presumably.

We realised our messaging made people expect a direct translation to revenue rather than automating a job they had to spend around 30 minutes and 3 different tools everyday. With SMBs, many of our initial customers themselves didn't have product-market fit. And it was virtually impossible for us to find PMF for someone else’s company.

Even after changing our messaging, demos kept happening this way due to the perception of the magical nature of the workflow of what we built. Prospects came to meetings with fixed expectations. At that point, it became damn difficult to convince them otherwise: To tell them that we cannot guarantee sales out of this but we can automate 90% of the outbound you do every day.

Which brings me back to delivering a product that works on autopilot:

The market narrative currently is that AI's opportunity lies in replacing services and consequently tapping into the labour budgets. Our experience is somewhat nuanced if not different:

1. For average users, AI is still software, not a service, regardless of outcomes:

   • Bad: People won't pay human-level prices.

   • Good: They'll pay significantly more than they do for simple tools. (Apollo: $7.90 → 1,000 leads; Clay: $31.40 → 1,000 leads; Godmode: $1,000 → 1,000 leads)

2. People expect software to deliver reliable, consistent results.

   1. Hence you have to make sure that your end product is something you can reliably deliver 100% of the time.

   2. Making money for another company is not one. If it were, honestly, I would never sell this. Instead I would use it as a secret weapon for my own.

A good way to illustrate this fact is the following:

Compare customer support vs. sales development

• Customer support AI works well because the job is clear: resolve the ticket or don't. Success is binary.

• Sales development AI struggles because the real job is closing sales. If conversations don't lead to revenue, users leave. Rightfully so.

  • Sales success depends on many factors beyond AI's control: relationships, deal cycle management, content quality, and most importantly, product-market fit.

For us this looked like this: Small companies expected miracles. Some wanted AI to solve all sales problems. Others expected us to find product-market fit for them. Many had never succeeded at outbound sales manually.

For larger customers, success rate was much better. However this time around, because the sales process took longer, they started to devalue the outbound process Godmode provided.

I know what you are thinking. The Why didn't you charge just for outcomes?

We tried charging $20 per positive reply with no base fee.

New problems emerged: When prospects didn't respond, clients blamed our targeting. When deals didn't close, they questioned lead quality rather than their messaging or product market fit.

The fundamental challenge became clear: In support AI, success is binary - ticket resolved or not. In sales, success means revenue. Many factors beyond outreach determine sales success.

The decision

Most startups fail.

I knew this and kept going. Why?

Because building a company is more about the chip on my shoulder than it is about anything else. It is what I will do until I can’t.

So we are shutting down, but only to build something bigger. Better.

What's next

I'm being very thoughtful about what to work on next. Building Godmode was harder than any of us expected it to be. Nevertheless, we loved it.

We're proud of what we built. Our technology worked. But ultimately, we believe we can create something with even greater impact.

Thank you to our investors who believed in us. Our customers who trusted us. Everyone who supported us through this journey.

So long, and thanks for all the fish,

Mert Deveci

Founder,

Godmode