Which marketing task do you still refuse to automate?

MegaSmith

Regular Member
Jr. VIP
Joined
Jan 26, 2026
Messages
379
Reaction score
135
Automation tools are better than ever, and I use them for plenty of repetitive work. But there are still a few tasks I prefer to handle myself because they have the biggest impact on results.

For me things like reviewing campaign performance are still worth doing manually.
 
Two of these we still do by hand, and in both cases automating them quietly costs money.

The first is the opening message. We run outreach every day, and the gap between a message that gets a few percent replies and one that gets almost none is usually a single line of wording. Generated openers read well and perform badly, so a person writes it and we test it on a small batch before it goes wide.

The second is deciding when an account needs to slow down or rest. Our tooling tracks the counters automatically, but the call stays manual. The numbers tell you something is off. They don't tell you whether it's the account, the message, or the audience, and treating the wrong one costs you the account.

Everything measurable in between, we're happy to let run on its own.

Best,
Floqal
 
for me i prefer to optimize my GP myself for it not to get suspended even tho i can automate the listing part
 
Me too. It's awesome when automation helps to save time, but in many cases the decision making still requires humans. Campaign analysis, finding trends and changing strategies cannot be easily automated.
 
Automation tools are better than ever, and I use them for plenty of repetitive work. But there are still a few tasks I prefer to handle myself because they have the biggest impact on results.

For me things like reviewing campaign performance are still worth doing manually.
Ahahah, this got me thinking, to be honest. Design audits for sure, AI just can't do it properly now, and core strategic things. Rest - automated or mostly automated if possible. Recently, I decided to fetch 100 YouTube channels about crypto/business and get their contact Email, and AI was absolutely horrible in this task, so I had to manually get all these 100 emails, absolute pain xDD
 
Automation tools are better than ever, and I use them for plenty of repetitive work. But there are still a few tasks I prefer to handle myself because they have the biggest impact on results.

For me things like reviewing campaign performance are still worth doing manually.
Outreaching clients with email or DMing them. I usually use AI to create the email/dm but I humanize it myself in order not to appear generic.
 
Had keyword tools Keyword Revealer and BeyondMe. GPT has keywords with volume ranges, so I dont think I will ever pay out for AHREFS or such again at their pricing. My application coding and WP sites use AI though theres more to it! I have to watch, troubleshoot and give feedback to the AI (Chatbot), plus produce the ideas behind the feature(s).

I thought AI replies was too much, now Im looking at how to automate email replies. Even my phone system, that Im wanting to build, incorporates AI. If you can automate a system aka a funnel that brings in clicks/calls and nurture those closer to sale then great. So these two big systems; PBX and Email Outreach + Newsletter funnel I planning to build (again), this time with automated followups in the form of AI. Should say the PBX AI be a problem, its just setup because I want a 25 yr Perpetual license, so I can revert back to non-AI if I want or if policies change. Extra 10% tax now on software because of government idiot politicians in my state.

There is no auto-cart for my niche in high ticket services though, someone has to sign the contract either physically or digitally (like Docusign). I learned about using Zoom in the insurance industry, can effectively do Zoom discovery sessions (15 min), at 45 minutes the Free Zoom plan ends the session.

Also, someone has to generally setup the automations manually. Even AutoGPT which was taking a while to setup before, required some time. Some of the best software beyond my LTDs lately has been found on GitHub (such as no-cost email validator) or custom coded into application menu as a feature, harnessing Python code to power the features (40-50 when finished, on a menu with categories).

My next plan for AI beyond those 2 big B2B systems is a new RPA bot for real estate reports pulled from sources.
I have a step cousin who built a side project for Fractional board members and made tens of thousands already with his idea.
 
Back
Top