Alejandro

Most SaaS companies treat paid acquisition as a spend problem. It's a creative problem. Your CAC isn't high because the channel got expensive. It's high because you're running three ads against competitors running thirty. The winning motion is the same one performance marketers use. Test a lot of angles, kill the losers fast, put budget behind the two that actually lower CAC. Retention is what makes the payback math work. But you never reach payback if the top of funnel can't produce enough creative to find the right buyer cheaply. Distribution is a creative volume game now, not a budget one.

Here's the reporting workflow we deleted last month. Before: a strategist opened Meta, Google, and TikTok Ads Manager every Monday, pulled numbers into a sheet, wrote the same client summary by hand. Four accounts, half a day gone. Now: the data pulls itself across all three channels, flags which creatives are winning and losing, and drafts the summary. The strategist edits for 15 minutes and sends. Same output the client sees. Six hours of senior time back every week. The tool doing the pulling and flagging is Adle, but the real unlock was killing the manual step, not adding software.

Everyone's arguing about whether AI kills SaaS. Wrong fight. AI doesn't kill the tool. It kills the workflow the tool was built around. Same story in agencies. The threat was never AI. It's still billing hours for creative production and reporting that a machine now does in minutes. The ones who survive rebuild the workflow instead of defending it. We built Adle after watching too many good teams defend the old one too long.

@thejustinwelsh This is the whole game now. Building got cheap, selling didn't. The people winning aren't shipping better products, they're just the only ones who can get someone to care enough to buy.

Knowing how to sell is the biggest game changer right now. It's so easy to build shit, but 99.9% of people have no clue how to find leads, uncover pain, and sell the thing. Incredible salespeople are positioned to be 100x'ers.

Most startups don't die from a bad product. They die because nobody could find it. Distribution isn't a phase two problem. It's the whole game. The teams that win figure out how they'll reach people before they finish building the thing those people will reach for. Your first paid tests aren't about ROAS. They're the fastest way to learn who actually wants this and what makes them move.

Everyone using AI at work is asking it to do the thinking. That's backwards. The thinking was never the bottleneck. The grunt work was. The 15th creative variation. The same report pulled for the 8th client. One idea reformatted for every placement. Point it at the repetitive part and your best people finally get to be your best people. The teams winning with AI aren't replacing judgment. They're clearing out everything that stole time from it. We built Adle around exactly this, the work nobody should still be doing by hand.

@jasonfried The i3 is the perfect case of right product, wrong timing. Carbon fiber body, purpose built electric platform, and the market just wasn't ready. Being early and being wrong look the same until everyone catches up.

A friend just bought a 2020 BMW i3. I'd never really been up close and personal with one (an i3, not a friend... I have friends). So innovative, so impressed. The car was ahead of its time in so many ways. Would love to own one.

@ecomchasedimond Not on it yet. The part I'm watching is whether it keeps the whole loop on platform. Payments living where attention already is shortens the path from content to sale.

Just got access to X money. Anyone using it? If so, how's it been?

Every rule that makes a paid ad work makes a personal brand work. The hook is the post. Nobody reads line two if line one doesn't stop the scroll. Reps beat perfection. One banger a month loses to decent posts daily. The algorithm and the audience both reward volume. Creative fatigue applies to you too. The angle that blew up in March is invisible by June. Keep testing new ones. You're not building a feed. You're running a creative testing account for yourself.

Most early stage teams treat paid ads as a growth channel. Too early for that. At the start it's a research channel. $500 on Meta tells you in a week what 50 customer interviews take a month to say. Which pain makes people stop scrolling. Which promise gets the click. Scaling spend before the message is proven is how cash gets burned. Let cheap traffic pressure test the pitch first, then put real money behind what the data already picked.

Everyone's arguing about whether vibecoding is real work. The same debate is coming for ads. AI can generate 50 variations in an hour. That's not the hard part anymore. The hard part is the feedback loop. Knowing which of the 50 to kill by Friday and which one gets the budget. Volume without a system is slop. Volume with a system is how small teams beat big ones. That second part is basically why we built Adle.

Recording a workflow once beats writing a prompt a hundred times. Most ad ops is this exact shape, pull numbers, pause losers, rebuild the same report every Monday. We built Adle on the same bet for campaigns, show it the work once and let it run. The teams that hand off the repeatable stuff first scale without adding headcount.

Strategically speaking, this seemingly trivial feature is the strongest wedge into b2b workflows that I can imagine OpenAI having today. Especially when it looks like people tend be have Claude at work versus ChatGPT. x.com/hnshah/status/…

Record & Replay is wildly underrated. You teach ChatGPT a workflow by doing the work once. The recording becomes a map the system can understand, repeat, and eventually improve. OpenAI should go much harder on this.

Everyone's hyped about AI writing code. The real unlock is the loop. Code that ships, measures itself, and fixes what's broken with no human in the middle. Ad accounts run on the same principle. The account that wins isn't the one with the best media buyer. It's the one that reads live performance and cuts losers before you've had your coffee. Generation was never the bottleneck. The feedback loop is. We built Adle around that loop, kill what's dying and scale what's working in real time.

@ecomchasedimond The fake ones all sound the same. Round numbers, vague praise, no name attached. Real proof has friction in it. A specific number, a real outcome, an actual face. Specificity is what makes it believable.

95% of Social Proof Sounds Fake - Here’s How To Fix It:

TikTok and Meta punish creative on completely different clocks. On Meta a good creative runs for weeks. On TikTok the same ad is dead in days. Run TikTok like Meta and you fatigue before you scale. The winners aren't better targeters. They just feed the platform fresh creative faster than it burns them out. Build the pipeline for the fastest channel, not the slowest. We built Adle because keeping up with that pace by hand was impossible.

Your personal brand runs on the same rules as a paid ad account. The hook decides everything. If the first line doesn't stop the scroll, the other 200 words don't exist. Nobody reads post 2. Volume beats perfection. The people growing fastest aren't writing better posts. They're posting more and cutting the ones that flop faster. You don't have a taste problem. You have a sample size problem. Post more, study what actually landed, do more of that. The account that wins isn't the smartest. It's the one that treats every post like a test and lets the audience pick the winners.

Most brands using AI for ads are using it to make one polished creative faster. That's the wrong job. The value isn't a better single ad. It's 40 variations of the same concept by lunch so you can find the 3 that actually convert. AI didn't replace the creative. It replaced the reason you only tested 4 ideas a month. The output looks boring on paper. A folder of near identical hooks, angles, and formats. Then you run them and the winner is one you never would have greenlit by taste alone. Stop using AI to polish. Use it to test. We built Adle around that one shift.

Retargeting isn't dead. Your retargeting list is just a fraction of what it used to be. The pixel now catches a sliver of your site traffic. The pool you can bid on shrinks and gets more expensive every quarter. The fix isn't a better audience. It's better creative that does the retargeting for you. Sequence the ads. First touch earns attention. The next ones handle the objections. Let the creative walk people down the funnel instead of leaning on a shrinking list to find them again. The accounts still winning here moved the work from targeting to creative. We built Adle because that move is impossible when every new creative takes days to make.

@AlexHormozi Same energy as the $20 a day test budget. You didn't test anything, you just bought an excuse. Half committed spend gives you half the data and a clean story for why it flopped. Commit or don't run it.

Calling it a ‘side hustle’ is just a modern way to protect your ego if you fail.

@briannekimmel The market doesn't care how the ad got made, only if it works. Same with writing. Nobody bounces because it was AI-assisted, they bounce because it's boring. Slop is a quality problem, not a tool problem.

Maybe it’s just me, but I find the slop police to be more annoying than reading obviously AI-assisted writing.