#181 · What GPT-5 means for AI agencies
The automation stops because of a timeout or error around 150-160 rows. Instead of just increasing the timeout (which is a temporary fix), split the automation into a subworkflow that returns instantly, handling queuing and waiting logic in the subworkflow so the parent never worries about timeouts and can process thousands of records.
The length of a cold email campaign depends on your email volume and number of leads. For example, sending 100 emails per day to 1,000 leads lasts about 10 days (or roughly six days a week adjusted). Campaigns can run as long as you want—I’ve run some on and off for about a year until the lead pool was depleted, so there’s no fixed rule. To minimize operations, note that the process scales linearly with the number of leads (O(N)). With 1,000 leads you’ll perform roughly a multiple of 1,000 operations. You can reduce ops by using Google Sheet scripts or App Scripts to handle personalization automatically, or by leveraging community hacks and tools such as bulk‑adding leads to a sheet or the OpenAI bulk endpoint (which takes about 24 hours). While custom solutions exist, I find the 80/20 approach with Make.com sufficient, spending only $10‑$15. To avoid Make.com’s 45‑minute timeout, split the workflow: instead of running the OpenAI call inside the main scenario, send the data via a webhook to a second scenario that processes it and returns the result. This way each scenario handles one lead at a time, keeping execution time low (e.g., ~45 seconds for 1,000 leads at 10 seconds each) rather than hitting the timeout.
First, don't overreact to a single day of zero replies when sending at a small scale like 100 to 150 net-new leads per day. That is well within normal statistical noise—one day could be 0% and the next could be 6%. However, if your deliverability score has genuinely tanked below 80% on platforms like Instantly or SmartLead, warm-up usually will not fix it. The best move is to replace those accounts with fresh ones. To prevent getting stuck, you should always maintain a buffer of extra mailboxes on continuous warm-up—roughly 20% above your active sending capacity. Treat this extra 1.2x cost as an insurance policy so you never have to pause sending entirely while paying for software, domains, and leads.
Sure. Let me walk you through the solution. Your current scenario loops over an input—say a thousand records—using a built‑in iterator and then feeds each record into an AI module that takes several seconds to run. At 5 seconds per execution, you’d hit the 45‑minute timeout after roughly 600 runs, which makes the whole process take a very long time. The fix is to split the scenario: instead of processing everything in one long loop, have the iterator immediately send each record to an external service via an HTTP webhook. The webhook executes almost instantly (around 0.001 seconds), so you can process millions of records within the same timeout window. In practice, you can create a small JSON array, iterate over it, and for each item make a webhook call that triggers the AI module on a separate scenario. This way the sender scenario finishes quickly, and the receiver scenario handles the AI work without ever hitting the timeout. You can also use Make’s sub‑scenario feature or a dedicated Make.com module, but webhooks are the simplest and most reliable method. Just ensure your Perplexity search and ChatGPT modules are placed in the receiving scenario, and you’ll avoid the timeout entirely.
You're probably doing something wrong — if that's 200 new leads a day (1,000 total over 5 days) with literally zero replies, positive or negative, something's broken, and it's overwhelmingly likely to be the copy. Start by adding yourself to the sequence and sending it to yourself to catch anything obviously wrong (broken HTML, stripped line breaks, ugly formatting). Then dig into what's putting people off in the subject line, teaser, and body — I have a full copywriting walkthrough in my 'watch me start and sell an AI service' video, including a 4-step email framework (essentially: not spam, who are you, why should they care, and what happens next). Work through that, make changes, send the next thousand, and see if the reply rate moves before assuming patience alone will fix it.