In November 2022, OpenAI released ChatGPT. By February 2023, sales leaders on LinkedIn were showing screenshots of cold emails with 40% reply rates. By summer, the same people were quietly deleting those posts. The truth sat somewhere in the middle, and for B2B teams it mattered more than the hype.
What actually changed in Q1 2023
Before ChatGPT, a decent cold email took an SDR five to seven minutes to personalize properly: read the prospect's LinkedIn, skim the company homepage, find an angle, write the opener. With GPT-3.5 connected to a data source, that dropped to under thirty seconds. For the first time, an SDR could hit a 150-prospect list with openers that actually referenced the prospect — not just Hi {{first_name}}, hope you're well.
The result: teams that adopted AI personalization early in 2023 reported a 2–4x lift in reply rates versus their own Q4 2022 baseline. Not because GPT writes better than a human — it doesn't — but because it lets humans personalize at volume.
What stopped working
Three things died fast in 2023:
- Template-only outreach. Prospects started recognizing the rhythm of a GPT-written opener ("I saw your recent post about…"). Generic AI tone is now a signal of spam, not effort.
- Pure volume plays. Sending 500 emails a day from a single mailbox became suicide once spam filters started training against AI text patterns.
- The "one prompt fits all" workflow. Teams that used the same prompt for every persona saw reply rates collapse within 8 weeks.
What works instead
The teams that are still winning in mid-2023 are doing three things:
- Feeding the model real signals — job changes, funding rounds, tech stack moves, hiring activity — instead of scraping the LinkedIn "About" section.
- Writing the human parts themselves. The opener and CTA stay human; AI handles the middle (context, transition, value prop tie-in).
- Rotating inboxes and domains. Deliverability is now more important than personalization.
"AI didn't replace the SDR. It replaced the part of the SDR's day that was typing the same sentence 80 times." — a VP Sales we talked to in June 2023.
The quiet shift underneath the hype
The bigger story of 2023 isn't GPT-written emails. It's that the bottleneck of outbound moved from writing to targeting. When every team can write personalized emails instantly, the differentiator is who you're writing to. That's why intent data and data quality became the real 2023 conversation.
What this means for 2024 planning
If you're budgeting for next year, don't buy "AI email writing" as a line item. Buy data, deliverability, and a sequencer that lets you plug AI in where it helps. The tools that win the next cycle won't be the ones that wrote the email — they'll be the ones that knew who to send it to.