How 1 Person Manages 15+ Social Media Accounts: Our Exact System

You’re probably thinking that having a 10 person social media team is necessary to manage the workload of your various social media accounts and make sure everything is running smoothly. And while that might be true in some circumstances, the answer in the majority of cases is almost certainly no. What you actually need is a good system in place for managing your content and social media marketing.

At GSI, we manage over 15 different social media accounts for GSI and various clients, with a staff member who works only part time. We have seen our overall social media engagement rise by 78% in the last year, which clearly shows our content is well-received. And we did it all without a team of ten people posting updates constantly all day. What we have is a system. A machine. If you’re looking for help implementing this, talk to our team.

Most social media “strategies” are really just bad habits masquerading as a workflow. You’ve got pointlessly fragmented tools, constant switching between apps, and a never-ending stream of mundane, soul-sucking tasks that are driving your team members to the brink of exhaustion. We chucked that whole approach in the garbage and started from scratch.

This isn’t theory. This is the actual system and process we use on a daily basis to create our income streams. We are laying bare and in great detail the tools we use, our methodology for work flow, the automation rules and scripts that we have created and the results we achieve with them.

TL;DR: How we manage 15+ accounts with 1 person

Content lives in one Airtable database, from the initial concept to the final URL of the published piece. No more digging through various Google Docs.

• Custom AI First Drafts: For a faster development pace, the adaptation of a custom AI is used to instantly produce platform-specific first drafts. This approach replaces the mundane hours-long process of initial scripting with a mere few minutes, significantly speeding up the application development process. We train the custom AI to sound more human and maintain a natural flow, ensuring that the output retains both language and coding effectiveness.

More Post Approval Fun =) Another very simple trigger based automation using Make.com. Airtable is used to keep the full history of a post and to hold the different versions (drafts). When the status of a post is changed to ‘Approved’ it is then automatically generated and scheduled in the marketing channel of the account. The actual automation work is done by Make.com (which connects our tools to each other). This robot does the very mundane work. A human is required to decide on the core elements to post, as well as review the output for any final adjustments needed.

• Integrated Analytics: Having a custom Looker Studio dashboard that pulls in data from across platforms allows us to get the full picture of what is working and where we should focus our efforts to amplify it versus making educated stabs in the dark.

The Problem: Drowning in 45 Tabs and Context-Switching Chaos

Social media management was a bit of a disaster before Argo. Does this sound familiar to you?

I have 15 tabs open at work every day. That includes LinkedIn, X, Instagram, Facebook, TikTok, YouTube and then I have a few tabs with copy that I need to copy and past into Google Docs, I also have Canva open to make images and a spreadsheet to do some tracking. In fact, I’m usually pretty much useless to my family before I can even manage to put the words down on the page.

Time to share my recent post on LinkedIn. Simple, copy and paste the content to Hootsuite to schedule it for later, shorten it for X (you have to keep it even shorter for X than for LinkedIn or Twitter), and then the perennial search for the perfect Instagram image, which inevitably involves going back into Canva to sort out my inexplicable lack of consistency on image sizes. Ugh.

Closing a tab that doesn’t want to be closed is literally the worst. This change is a major productivity killer and switching between apps constantly is going to disrupt focus and flow of work. Multitasking was reviewed in a large body of research by the American Psychological Association and found that mental switching between tasks can account for up to 40 percent of unproductive time. Two full days of work lost per week to switching between tabs.

All action, all activity leads to a single goal: the middle. You barely visit the blog, reproducing copied and pasted comments you find on other sites. You don’t reflect on whether they make any sense, you just continue and keep on posting. You spend more time copying and pasting the texts that you barely even read them, and less time on the social interactions they are meant to enhance or, at least, ponder upon the meaning of your online life. You’re just a Content Assembly Line Worker, a few bad mornings from a full-blown burnout.

Our Philosophy: Automate Everything, Systemize the Rest

At MSH, our core philosophy is straightforward – humans should think and machines should do the work. We believe that if a task is boring, or predictable, or non-cognitive (i.e. doesn’t require the use of our ‘higher-order’ thinking skills), it should be automated.

The purpose of automation is not to replace people. The purpose of automation is to assist people in order to be able to do better things. Paying a highly educated person in order to copy and paste information from one place in the web to another, is a criminal and ineffective waste of money and time. People deserve to be doing far more valuable things.

1. Systemize the Creative Work The creative work of coming up with ideas and spinning them into something fresh and new can’t be automated. Though it can be systemized. So what’s the system we use for generating, reviewing and editing creative ideas? Rather than being an “intern” we think of ourselves more as the “editor-in-chief”.

We decided to automate the production part of our work. Almost everything else was already automated — first draft revision, content formatting for every possible channel, scheduling, analytics reports, tracking links etc. We simply made our human role the quality assurance check right before we hand it over to the automated production line.

The main idea is to change the process of water management in relation to social media from a stressful, overwhelming and unmanageable task to a calm and controlled activity, transforming the monotonous and unmanageable task of bailing water from a sinking ship into the comfortable and manageable command of a submarine.

The GSI Tech Stack: Our 4 Core Tools for Solo Management

You don’t need a dozen costly specialty tools. You just need a few powerful tools that are well matched to the task. Our whole system is built around four core tools.

1. The Content Hub: Airtable

This is our command center. No more scattered notes in Google Docs or confusing spreadsheets. We store every piece of content in Airtable which acts as a database.

Base: Tables for ideas, drafts, platforms and campaigns.

Sheets: We use sheets to keep track of what’s happening with the content we generate. Spreadsheets were created to store and organize data, but they have their limits. For us, they are more functional than a traditional spreadsheet. We add custom fields to record additional information, like the current state of our content (Idea, AI Draft, Human Review, Approved), which platform it will be shared on, who is responsible for it, the publication date, what images are included, and what the finished URL will look like.

Killer Feature: Custom Views. This week we’ve got Custom Views. We have a view for Ideas Needing a Draft, a Kanban view to drag posts through the approval flow and a calendar view of our schedule.

2. The AI Writer: GSI Custom Model

We’re really excited about this project, and we feel like we’ve hit on something special with our custom AI assistant. We’re not talking about using the stock ChatGPT here or anything. Our custom chatbot is AI-driven but has been trained to be truly on brand. It has had access to all of our knowledge about the language and tone that is working for us online as well as to our key messaging pillars.

• What it is: A custom-trained language model accessed via an API.

What is Generic or Uneducated or Stupid AI? AI that generates AI-generated content in a generic way. Why would we use such a thing? Well, the answer is because the resulting content will be banal and unoriginal to the extreme. Our “smart” language engine knows better than to ever suggest the use of “leverage” in a context that even a baby would consider to be poorly written. It knows not to use our snarky language and not to over or under-explain, depending on whether it’s a LinkedIn post or a X thread.

Killer Feature: Airtable Integration. We’ve got a button in our Airtable base called “Generate Drafts” and when we click it, the concept and all the bullet points get sent to the AI and it generates the drafts for the sites we choose and also fills in the Airtable fields.

3. The Automation Engine: Make.com

We have Make.com (formerly Integromat) – the glue that holds everything together. Make.com (formerly Integromat) is visual workflow creator that connects all the tools we use. No code needed.

• What it is: A no-code integration platform.

Workflows: Most automation tasks can be handled with Zapier, but for really complex tasks, such as multi-step actions, we use Make. We then create “scenarios” within Make that continually watch our Airtable base for changes.

Killer Feature: “Approved” Trigger. We just implemented an “Approved” trigger for one of our scenarios that runs every 5 minutes to search for records in Airtable that have the status changed to “Approved” and then the scenario proceeds with pulling in the content needed for the social post, formatting it for the channel and sending it to our native scheduler or Buffer.

How do you keep all your social media accounts filled with content? Well, here at Make.com, we’ve got a pretty cool solution for you!

4. The Analytics Dashboard: Looker Studio

You can’t improve what you can’t measure. Logging into 8 analytics platforms to measure different metrics is a huge waste of time. Now we have all the data in one place.

• What it is: Google’s free data visualization tool.
• An API (Application Programming Interface) Connector is essentially a bridge that allows us to connect various platforms and bring their information into one place. Why we use it: It connects to almost everything. We use API connectors to pull data from LinkedIn, Instagram, X, etc. and bring all the information to one place.
• Killer Feature: Blended Data. We get an instant view of the topics we are publishing on that are killing it across all platforms, not just what has performed well on one specific channel which gives us a massive resource to help shape our content strategy. Check out our recent article all about How to Build a Custom Analytics Dashboard.

AI-Driven Content Creation Process | An Overview

The middle “C” in our Content Marketing Process is Content in Motion. This is where the magic occurs! The following is a minute by minute account of the steps necessary to generate an idea and then successfully publish the content.

Total Time: ~15 Minutes

1. Idea Capture (Ongoing)
   Ideas are captured from a variety of sources: client meetings, blog and online articles, team brainstorming and the like. These are then easily copied and pasted into the “Ideas” table in Airtable along with a few bullet points to describe the idea and this task takes no more than 60 seconds.
2. AI Drafting (3 minutes)
   Our social media manager will log into the system to start reviewing idea queue for content, will then select the social networks they’d like to generate content for and click “Generate Drafts” within our application. Off to the coffee machine they go! The custom-built application uses artificial intelligence (the “AI”) to start generating the content. A few minutes later, the content will be fully generated: — A long-form LinkedIn post — A short thread on X — 3 potential headlines. And because we are connecting this application to Airtable, the generated content will also be written as individual columns back into the associated Airtable record.
3. Human Polish & Review (8 Minutes)
   The AI has done about 80% of the job. The manager now needs to do the remaining 20% to ensure the output really impresses the reader. They will:
   • Fact-check any claims.
   • Add a personal anecdote or a timely reference.
   • Sharpen the hook.
   • Ensure the tone is perfect.

   When writing a blog post about Artificial Intelligence and Sales Teams, keep the following tips in mind:

   • Use an internal link to a relevant post, such as our AI marketing automation guide.
4. Asset Creation (3 minutes)
   Sometimes for a social post you just need a quick graphic. We most often use Canva templates for these types of posts. The graphic will be shared as an asset in the Canva template and then copied and pasted into the Asset field of the Airtable record.
5. Final Approval & Automation (1 minute)
   Manager checks the final time and says all is good. Manager then updates the status from “Human Review” to “Approved” in Airtable.

And that’s it. Their work is done.

I found a really neat way to turn this into an action in My Make.com automation. As soon as I change the status, it turns blue and does the magic that I needed. My Make.com automation can now see social media and access the approved image and lines for the post. It then schedules the post and finds a slot in my content calendar. Finally it changes the status and post date in Airtable.

Man spent 15 minutes on really important, valuable stuff. Machine spent 30 minutes just doing garbage work. That is how you scale something.

How We Adapt the System for 8 Different Platforms (TikTok vs. LinkedIn)

People say, “You can’t just post the same thing on every platform.” And they’re right. This isn’t being an amateur on social media. An automation system needs to be adaptable to the different platforms. We configure the Airtable workflow to accommodate each one.

Here’s how we handle the big three content types.

For Text-First Platforms (LinkedIn, X)

This is probably the most straightforward. The AI is trained for High Quality, engaging articles in LinkedIn and short, catchy posts for X, and you set the AI posts to “Approved” so they automatically get published. This is the famous 15 minute workflow we mentioned earlier.

For Visual-First Platforms (Instagram, Facebook, Pinterest)

For these platforms, the Airtable record has an extra field: “Visual Brief.”

• Step 1: The creation of the AI draft also automatically populates the visual brief with a suggested content for the infographic or photo concept, such as “Infographic showing the 3 steps” or “Photo of a messy desk with 45 tabs open”.
• Step 2: The human review step includes creating the actual visual based on the brief, using our Canva templates.
• Step 2: Add more content
• Step 3: The “Appr

This system has transformed our workflow, allowing one person to do the work of a team. If you’re ready to build your own content machine and ditch the chaos for good, contact our team to get started.