Oh boy, Legend, our mascot, is working so hard during the @keeponboard customer onboarding stage 😓 He has to make sure our agents are connected and that everything is synced. Can't wait to share our onboarding journey with you once it's done! WDYT? Happy Wednesday, gang 🌞

Meet the last team member of the new @keeponboard comic strip crew! Captain Paige Turner. ⛵️ Paige is hands-down hands-on, unlike Chad. The minute a new visitor lands, she hands them swag that 'successfully' onboards them and immediately turns to the next visitor! Yes. Paige **leaves** in the now. 😌 Who needs retention if the top of the funnel is a waterfall?

The @keeponboard comic strip is picking up speed! And it's all because we have a Dr. on board now. Introducing: Dr. Carl Knewthat. - If someone died? He knew why... - He also knew that you could have saved him if you followed his instructions! His amazing hindsight capabilities are unparalleled. We. You. All of us. Are lucky to have him.

Our product is not the only thing we're building in public 😉 We're also building @keeponboard's Customer Service comic strip! We already introduced: - Legend, our mascot and ship Beaver. - Captain Chad Tablethis, our fearless bleeder! - Sam Doesitall, because he does. - And First Sailor, Jewels Backflips. And today, we're introducing a new character who believes that if you're gonna ship, at least do it with some style. Meet: Christof "Laidback" Knewfirst, the ship's bartender. He knows everything about everyone on the ship because, well, what happens in the bar stays in the bar. So let's help him keep the bar high with a round of applause 👏👏👏

Developers: "Building a headless product is such a great idea, hell, UI is redundant!" 😎 Also developers: "Hi, someone copied me, they're doing $10M ARR, they have UI... Why are they succeeding with this?" 🙄 Well fella, this experiment already ran. It was called Alexa. Conversational interfaces hide functions the way command lines did. GUIs were invented to surface them! It's called the articulation barrier. Users can't ask for what they don't know exists. (APIs for agents? Fine naked. I mean products humans use.) And when your product lives in someone else's chat, the blindness goes both ways. 1. The user can't see your capabilities. 2. You can't see your users. The hesitation before churn, the feature never found. All of it lives in the host's logs now. A headless product can't watch its own customers leaving. A good UI is a map of your platform's capabilities, and when users recognize things, they don't need to recall them. UI isn't here to stay. It's here to win.

The @keeponboard CSM comic strip grows! We're adding some girl power! Introducing "First Sailor Jewels Backflips!" She literally performs acrobatics to make sure everything stays on track and everyone's happy, doing backflips with too much to do and no resources. But somehow, she wings it. Will she be able to keep everyone on board?

Remember Legend, our Maskot? 🦫 We've been developing the @keeponboard customer success comic strip further, and we're happy to introduce 2 additional characters. - Captain Chad Tablethis. A fearless leader, at least until it comes to making decisions, and... well, leading. - Sam Doesitall. Wears every hat on the ship, but alas, he knows that fires will be fires, always the solution, always gets dilution. What do you think they will do first?

4 rules to succeed with any product you're building. Have your: - Ego = 0 - Research > your last assumption - Effort invested > your competitors' - Knowledge of the subject > your customers' Or: Success = (Research × Effort × Knowledge) ÷ Ego Yes, that's division by zero. That's the point.

World-building your brand is awesome! We have a mascot! 🦫 Meet "Legend," the @keeponboard beaver. He's plugging leaks in everyone's customer journey, keeping the boat afloat behind the scenes.

We need your help. We want to tell the @keeponboard story. We previously tested a Viking line. And now? We went full-on cruise line with the Love Boat ♥️ Be honest. It will help. WDYT?

If you're honest, at this point in time, you need to admit that you don't know who your ICP is. Not until you have at least 100 customers under your belt, and even then, hold it loosely because the times they are changin'. Product-market fit moves so fast today that your ICP can change within weeks, even if you have 10K customers. It happened to Shopify, HubSpot, and even Canva. So ya, it can happen to you. Speaking of you? Well, it sucks for you… And I'm not here to bring you down, I'm in the same boat, rowing. Building your SaaS on conversations with 10 people who loved the idea? Scary as shit. Even after you ran a campaign or two to "validate your product." Please, don't try to convince yourself or me. It is. When building your dream product, you don't yet know your ICP. However, you DO get to choose the axis you'll learn it on. Choose wrong, and a hundred customers will teach you nothing. You'll be reading noise, sorted neatly. I nearly chose wrong. I almost segmented our market into three revenue bands. $500K to $1.5M, $1.5M to $5M, and $5M + ARR. It's clean, presentable, and totally wrong. SaaS Capital surveys 1,500+ private SaaS companies every year. Their conclusion, in writing: for retention, segment by ACV. Not company age, industry, or revenue. Why? Because two companies at the same revenue can have nothing in common. But two companies with the same contract size? They're practically siblings. Same organizational structure, go-to-market approach, and way of supporting customers. A company selling a $19.99/month product and one selling a $250K/year contract are different species. Even at identical ARR. And the data tells it beautifully: - Under $25K ACV: the median NRR is around 97%. - At $25K to $100K: It's around 108%. - And above $100K? Around 118%. That's a 21-point spread, decided by contract size. Now split the same companies by revenue stage instead? Above $1M ARR, the median barely moves. One axis predicts how likely you are to churn. The other just tells you how big you are. Willingness to pay is not an ICP. Three price levels don't mean three customer profiles. I was about to build a strategy on an axis with no signal, and then wait for customers to "teach me" on it. They'd have taught me nothing. Whoever you are in this story, the question remains the same. - Doing CS yourself at $800K? The axis decides which customers you learn from first. - Just hired your first CSM? The axis decides which accounts land on their desk. - You ARE that CSM? Staring at a book of 200? The axis decides which 20 actually need you this week. Pick the axis that predicts behavior. Your slides can survive being ugly. Your retention can't survive being blind.

Keeping customers happy is pretty simple, and the "How" sounds like the complete opposite of what you should do.

You don't need to do unnatural things to keep your customers happy. I added my input into this list of 30 things you can do to keep your clients happy with no phony fakery. A wonderful weekend read that you can easily apply next week. (Maybe I'm biased 😆) I'M NOT!

My little dog is one unsocial beast. 🦮 He loves humans, don't get me wrong. But other dogs? Not so much. I tried: - Exposure to more dogs over time? Makes it worse. - Fed him treats to stop barking? Eats and runs. - People even suggested I neuter him. (We didn't.) Agents can be pricks too. 🤖 - Swarm? They break everything w/o supervision. - Feed them good data, costs go up, results might not. - Put guardrails, all the types, not enough. That failure I described in my previous post? Agents marking failed communications as resolved, closed, or "handled." On all possible edge cases! And when you deal with onboarding and retention of customers, edge cases are all the cases... Both dogs and agents jump over fences with a mission. At times, it's the mission you gave them originally. That's how they escape🦮 + 🤖 = 📉 So the fix was never a taller fence. Audit the mission, not the fence. The escape was in the briefing, not the barrier. What did you put in your agent's mission that it's going to obey too well? The only thing that's ever worked for my beast: I stay on the walk with him. Same fix, both species. No need for ✂️

Let's get nostalgic. 💻 Do you remember gating a PDF behind 11 fields? 💾 And the guy who wanted a 12th? My god, how many arguments did we have between product, marketing, sales and dev about how many fields a form should have? 😅 Today, almost no one talks about forms, it's like they died out. Well, they didn't die exactly, they Wolverined into a better version. Did you ever count the number of form fields on ChatGPT? Zero. There's a box. You type in it. That's how onboarding starts. Same with @claudeai. Same with @cursor_ai. Same with @Lovable. The most-adopted software of the last three years asks you almost nothing before it lets you in. Meanwhile most SaaS still greets a new user with a nine-field questionnaire. - Company size. - Role. - Team size. - How did you hear about us. And my favorite, "what's your goal," pick one from this dropdown. But, what could we do? Forms made sense. You couldn't sit down with every new user and actually ask them what they were trying to do. That didn't scale, so we built forms and made people translate their pains and needs into dropdowns, and called it onboarding. Now, that constraint is gone. So why are we still running the workaround? And look at what the dropdown actually costs you. Someone shows up thinking "I just lost my biggest customer and I need to know why before the board call on Thursday." Your form turns that into: Role: PM. Size: 200. Goal: Other. You had the whole story. You kept the checkbox. So forms aren't dying because forms are bad. They're dying because the reason we put up with them expired. A form records what someone said about themselves. The product records what they actually did. Only one of those tells you if they'll still be here in 90 days. Stop playing gatekeeper. Let them in, and watch. Okay Bub?

The inbox is an unforgiving... box. We've run thousands of email sequences for SaaS, apps, and e-commerce. onboarding, referrals, launches, promos, you name it. One thing decides whether a sequence pays, and It's dead simple. Say the value in the subject line! No games. No curiosity gaps. No bait. 😌 Put a number in the subject line. Around 15% more opens. 🤓 Reference something users did. Up to 45% more. 😵 Reach for bait words instead? That raises your odds of landing in the junk folder by 16%. Hell, after Gmail's Nov 2025 update, it will get that sequence thrown down the drain faster than you can say... Well, you won't be able to say anything, would you?

Founders, did you ever check If your emails land in the junk folder? Oh, you did, and you asked your users to white label them, right? Big mistake. Your users should never visit the spam folder, don't ask them. You work so hard on your product, don't neglect your emails. Instead, make sure you inbox properly each and every time. I was in the email marketing industry for 15 years. Here are my 22 rules to do just that. Bookmark this list for later, and spread it for others to know: 1. If you're using a shared IP, there's a high chance you'll land in the junk folder. Own your reputation with a dedicated IP. 2. Authenticate your domain and sending IPs with SPF, DKIM, DMARC. 3. Use a reliable and reputable email service provider (ESP). 4. Validate your subscription list and keep your contact information updated. 5.  Use a consent management and preference center to give your subscribers control over their inbox and data. 6. Keep your sender reputation clean, make sure your spam complaint never passes the 0.3% spam-complaint ceiling (0.1% is the real target). 7. Register for postmaster tools and feedback loops (FBL) to keep an eye on your performance with ESPs. 8. Comply with privacy and anti-spam laws and protect your subscribers’ data, use one-click unsubscribes. 9. Get an explicit consent before sending and include a clear unsubscribe option in your emails. 10. Develop engagement from the start to demonstrate your emails are welcomed by subscribers. 11. Segment your email list to send effective, tailored campaigns. 12. Create compelling email subject lines and CTAs to boost subscriber interactions. 13. Use dynamic personalization and customized content to delight subscribers. 14. Sunset disengaged subscribers for  lean, clean list. 15. Optimize your messages for mobile and screen readers. 16. Keep your email files right-sized and rendering correctly. 17. Spam test and QA your emails before you send. 18. Manage the frequency of your email sends and use individualized send time optimization. 19. Avoid including elements that trigger spam filters in your marketing emails. 20. Get deliverability reports from email deliverability tools and services. 21. Monitor your email metrics for signs of deliverability drains. 22. But be careful, since Nov 2025 Gmail permanently rejects non-compliant mail, and Microsoft bounces it outright. It doesn't go to junk anymore. It never arrives. Cool? Cool.

The most expensive moment in SaaS is the one nobody owns. Sales or self-serve learns why a customer showed up and what they were trying to fix. Then activation starts, and almost none of it crosses over. The people meant to keep that customer begin from zero, working on an account they were never introduced to. Nobody's job says "own the handoff." So the money leaks at the seam.

Why do accounts churn even if the account's data showed it's healthy? That question, if answered, will save the SaaS and agentic industries billions of $$'s. "But, but, it logged in daily for a year and still cancelled." "Usage looked fine..." What no dashboard could see? The champion who set it up had quietly left in April.