Your Next Customer’s Inbox Is Run by a Bot That Hates Your Templates

A sales rep from a leading public B2B software company emailed Jason Lemkin twice in 48 hours. The second email asked about his “end of 2024 goals.” In October 2026.

Lemkin’s response: “our 2024 goals?!!? seriously?”

The rep’s reply: “Hahaha fair enough, you got me there.”

And that was the end of the conversation. No pivot, no recovery, no attempt to actually sell. The rep just… gave up.

Lemkin tells this story in his piece on SaaStr, and it’s easy to laugh at the obvious fail. But there’s a darker layer: it wasn’t the rep’s fault. The system around him was broken. And that broken system is exactly what’s about to get replaced—by agents that won’t just send better emails, but will also screen them.

The real story isn’t the stale date. It’s the gatekeeper.

The sequence step was written in 2024. It survived two annual planning cycles. It was loaded against Lemkin’s name. And an AE sent it—or let it auto-send—without reading it.

Lemkin’s point: at volume, humans stop reading what they send. The tooling assumes they still do. That’s why he’s moved 90% of SaaStr’s outbound to AI agents—Artisan, Monaco, Agentforce—with three humans managing the relationships that close.

Here’s the part that deserves more attention: buyers are changing too. Lemkin writes:

Agents are now a top-5 source of SaaStr’s web traffic, and 10K (his AI VP Revenue) evaluates vendors for us. It shortlisted a vendor, ran the numbers, and dropped it within 12 hours: premium pricing with no conversion data, multi-month minimums, and scarcity framing. It moved on to other options without being asked.

Read that again. An AI agent is screening vendors on the buy side. And it’s not just SaaStr. PainSignal data shows this is a spreading pattern. We track problems where prospects’ AI assistants filter out emails before a human ever sees them. The severity is high because the old playbook—spray and pray—doesn’t just underperform; it’s now actively blocked.

Your outbound is already being judged by robots

If you’re an indie hacker or agency dev building sales tools, this changes the game. It’s no longer enough to write a great cold email. Your email has to pass through a bot that’s been trained to ignore generic pitches.

And the data backs it up. PainSignal tracks 41 problems in the Communication category alone. “Generic email outreach,” “lack of personalization,” and “cold emails that ignore context” are top complaints, with several rated 4/5 severity or higher. One specific problem: “Sales emails are too generic and don't reference our company or usage” has a severity of 4.2/5.

But this isn’t just a B2B SaaS problem. The author’s focus on a public tech company might make you think agents are an enterprise luxury. That’s wrong.

PainSignal data shows 18 problems in Real Estate, 12 in Insurance, and 15 in Home Services that all complain about cold outreach that doesn’t acknowledge specific situations. Severity up to 4.5/5. These are high-touch, long-sales-cycle businesses where personalization matters even more. And they’re drowning in generic, sequence-driven communication.

The opportunity: build the “SaaStr-style” agent stack for the long tail. Small and mid-sized businesses in these verticals are AI-curious but overwhelmed. They tried tools that were too expensive or complex. They don’t need Agentforce. They need a simple agent that reads the CRM record, pulls public info, and writes a personalized email—for $50/month.

Don’t just replace the human. Replace the sequence.

Lemkin’s four recommendations for fixing that failed sales org are gold:

1. Have an agent review every outbound email, including ones humans write, to block stale dates or generic pitches.
2. Feed the agent the prospect’s public footprint.
3. Replace the six-item menu with one pain, one number, one ask.
4. Keep humans on deals that need them.

But there’s a twist he doesn’t address: the agent reviewing your outbound is not on your side. The buyer’s agent is looking for a reason to archive you. So your outbound agent needs to be trained not just to avoid mistakes, but to pass the buyer’s AI gatekeeper.

What does that look like? Personalization is table stakes. But you also need to signal relevance in a way that a bot can parse: specific data points, clear context, and no fluff. It’s the same principle as SEO for search engines—you’re optimizing for the algorithm, except now the algorithm is deciding if your email gets read.

The data you need

If you’re building in this space, here’s what to track:

• Problems around sequence staleness are on the rise. We see an average severity of 3.9/5 for problems about outdated email sequences.
• AI screening is an emerging pain. Sellers are reporting that prospects’ assistants filter them out. This will explode.
• The long tail is underserved. Real estate, insurance, and home services are crying out for better personalization, but they don’t have the tech chops to build their own ABM tool like Lemkin did.

What to build next

1. An email review agent for outbound that flags generic content and stale dates before sending. Think Grammarly, but for sales context.
2. A personalization engine for small businesses that pulls from public data (social, news, reviews) and generates a unique email for each prospect—without requiring 10K’s data warehouse.
3. A “robo-gatekeeper” simulator that scores your outbound emails based on likely AI filtering. If it gets past the bot, it gets sent.

Lemkin’s story is a funny anecdote about a 2024 pitch in 2026. But the real punchline is that the next time you send a cold email, it might not be a human who laughs—it might be an agent that simply deletes it.

The reps who survive will be the ones who learn to sell to the bots first.