We built @ToltHQ to be the most AI-native partner platform out there.
Our MCP scored 98/100 on @manufact’s leaderboard.
Ready for Claude and ChatGPT out of the box.
Not all MCP servers are created equal, so we made a leaderboard.
Check your server, get a quality score out of 100, and see exactly what you need to improve.
Post a screenshot of your score and tag us @manufact for a free month on any standard plan.
Link below 👇
introducing @ToltHQ Agent: the AI affiliate manager
instead of clicking through ten tabs and running manual exports, you can:
→ run an entire 7-figure partner engine completely on autopilot
→ ask anything about your program in plain English
→ spot fraud, unpaid commissions, and partners going quiet
→ approve partners and adjust commissions
→ scale your creator program to thousands of partners without adding headcount
try it for yourself today
tolt.com
Can't wait to show everyone what we've been building lately! 😬
These past days I've been rehauling @ToltHQ's MCP and built our own agent 🤖
Here it is building a flow for me. Maybe in this day and age UI is slowly less used as MCP tools start to become more powerful...
.@heyreach_io's affiliate program runs on @ToltHQ and is now driving over $3M, so we get to see which of these plays actually work.
One I really like: every Black Friday they bump the commission rate for partners. Suddenly it's a no-brainer for agencies to push their clients toward an upgrade.
The agency makes more, the client gets a deal, HeyReach gets the revenue.
Simple, and almost nobody does it.
Congrats to the whole team on $17M! 🚀
How we bootstrapped HeyReach to $17M+ ARR in 3 years:
$139/mo → $1M ARR
AppSumo bought us time.
We focused on lead-gen agencies.
Unlimited LinkedIn senders for one fixed price.
$1M → $2M ARR
Built the API, webhooks and integrations with Clay, Instantly and Smartlead.
Agencies could add LinkedIn outreach to the workflows they already used.
$2M → $3M ARR
Rewrote the backend while keeping customers online.
Added weekend support and helped trial users launch campaigns.
Trial-to-paid conversion rose from 17.5% to 19.1%.
$3M → $6M ARR
Added quarterly and annual plans.
Ran LinkedIn outreach events with Clay across Europe.
Cut ads after two months of testing. Growth sped up.
$6M → $7M ARR
Shipped workspaces and user permissions.
Agencies could separate client accounts and control who had access.
Missing those features had been our biggest reason for lost deals and churn.
$7M → $8M ARR
Launched our creator program and LinkedIn MCP.
Creators taught the workflows. Customers could run outreach through AI agents.
Reached 2,600+ trials a month without ad spend.
$8M → $10M ARR
Gave creators higher commissions and sent them leads for their agencies.
Affiliate links accounted for $1M ARR.
We contacted old trial users who had started visiting our website again.
An $800 ad test brought in 17 customers.
$10M → $16M ARR
@Vick_Mucunski rebuilt onboarding and flagged customers who stopped using the product.
Revenue churn fell from 9% to 7%.
Recovered 30% of failed-payment revenue in four days.
Helped customers use more seats and new workflows. Some cancellation calls became upgrades.
$16M → $17M+ ARR
188 existing customers paid us another $100,865 in 30 days.
The offer: pay for 3 months, get 6.
Discounts on quarterly and annual plans when paid upfront.
What I've added above are some of the major growth plays and decisions at each milestone. Full breakdown below.
Found @andrejsshell on X, building @usekaneo, and knew we had to work together.
Now he’s rebuilding our entire API and SDK layer at @ToltHQ. Beast.
Over the last few days have been my most productive at @ToltHQ. I've rebuilt our public APIs, worked on SDKs, embeds, webhooks and what not.
Probably one of the cleanest code I've written in a LONG time! All I had to do is build out the initial structure, rules etc... and now my agent follows the patterns.
Feels surreal saying this, but one creator drove $3.5M in revenue and earned $624k in commissions through Tolt.
$624,000 paid to one person sounds crazy until you look at what they actually brought in.
No ad budget. No agency retainer. No upfront spend hoping something converts.
They brought customers in, the company made money, and they got their share. As simple as that.
We see payouts like this inside Tolt, and I still think people look at affiliate commissions the wrong way.
A $624k payout isn’t a cost you’re trying to minimize when there’s $3.5M sitting on the other side of it.
If someone brings you millions in revenue, pay them 🫰
.@praktika_ai has been following me around on Instagram for months trying to get me to learn Spanish 😅
Then they signed up for @ToltHQ.
They’ve built one of the fastest growing AI language learning apps, and apparently decided their next step was making sure I couldn’t escape them at work either.
I knew we had to get them properly set up, so @grupchev and Nick went all out.
Really excited to say their partner program is now powered by Tolt 🤝
Had a great catch-up with Eugene recently too, mostly around affiliate analytics and reporting. Gave some really good suggestions and a few of them are already making their way onto our roadmap.
They’re also launching with a pretty crazy offer for partners: sign up before the end of August and you can get 50% revenue share through the end of the year.
Big thanks to Eugene (CMO), @adampraktika (CEO) and the whole Praktika team for trusting us.
Check out their affiliate program here: praktika.ai/affiliate and learn more at praktika.ai.
This might not be the brag I think it is, but how many times can you get rejected by YC?
I started applying when I was 18. That was 9 years ago. Somewhere north of 12 rejections since then, and 2 interviews that went nowhere.
I always thought that was the path. Apply to YC, get in, raise, succeed. That never happened.
After the W23 rejection we joined Founder University by @Jason instead. No guaranteed funding, no batch, just sessions and workshops and building alongside other teams. Honestly it felt like the consolation prize at the time.
We kept building through the program. At the end their team flagged our progress and introduced us to the @LAUNCH Accelerator. We interviewed, and got into the LA30 cohort.
Nine years of applying to one place, and the thing that actually worked came from the door I walked through when that one stayed shut.
If I had to launch a partner program from scratch this week, I wouldn’t overthink it.
Most teams get stuck on the setup. Commission %, what others in the industry are offering, how the portal looks.
All of that matters, but it’s not what gets things going.
I’d just pick a number, say 30%, add our logo and favicon to the portal, and get it live.
Then start inviting people who already talk to our audience.
That’s it.
The difference between programs that work and the ones that don’t is usually just this.
Someone actually doing it vs someone still setting it up.
If you want to start, it’s here > tolt.com
We got tolt.com for 4k after 2 years of grinding it out with an .io domain.
Money well spent 🫰
This was our first “office” 3 years ago.
We were 3 people working out of an apartment, pushing Tolt out as fast as we could.
We outgrew it pretty quickly.
Now the team’s spread out, fully remote, and shipping a lot more than we were back then.
Funny part is, this setup was more than enough to get everything started.
We’ll probably set something up in Skopje next.
And outgrow that one just as fast.
2 years ago we tried to buy tolt.com for $20k. @GoDaddy's broker wanted $50k. We bought it for $4k.
Best decision we accidentally ever made.
We started on tolt.io, like every startup does. But as we grew, I really wanted the .com. So we made an offer, $20k. The broker came back demanding $50k and wouldn't budge. Too rich for us at the time, so we walked.
A few months later, an email lands in my inbox: the domain's going to auction.
Same domain. Now I can just… bid on it.
So we get in. A bidding war kicks off with one other person. I set our hard limit at $20k and told myself I wouldn't go a dollar over.
Didn't even get close.
We won tolt.com for $4,060.
Still have no clue how domain sellers/marketplaces/brokers actually work 🤷♂️













