If you manage 500 TikTok accounts for clients, where do you log in to those accounts? And how do you make sure each account can watch videos and randomly follow users at the right time every day?
You might think of two options:
Build a phone farm with dozens, or even hundreds, of physical phones.
Use TikTok in a browser.
From a mobile-first perspective, a phone farm might seem like the obvious choice. After all, TikTok is built around the mobile app, and running accounts on real phones feels close to how normal users behave.
But there is another side to it: physical phone farms are expensive to build, hard to maintain, and not very beginner-friendly. Setting up devices, installing apps, managing proxies, and keeping everything organized can quickly become overwhelming.
So, how can you run random follow automation for TikTok account warming in a safer and more scalable way?
In this guide, we’ll walk through how to use GeeLark cloud phones and automation templates to run TikTok random follow tasks faster, more safely, and at scale.
Run TikTok in cloud phones
To run social media automation safely, you first need a better way to manage all your TikTok accounts. That means keeping each account organized, separated, and easy to access before any automation starts.
Use the native TikTok app
A browser can open TikTok, but it cannot fully recreate the native app experience. Physical phones can do that, but buying, configuring, and managing hundreds of devices is expensive and inefficient.
GeeLark uses a different setup: cloud phones.
Each GeeLark cloud phone is a cloud-hosted Android device backed by real mobile hardware, including ARM-based chips and phone boards. This allows each cloud phone to have device-level information such as phone brand, model, Android ID, IMEI, MAC address, and other mobile identifiers.
In practice, this means your TikTok accounts can run in the native TikTok app, just like they would on a physical Android phone. You can install TikTok, update it to the latest version, log in to your account, watch videos, and let the automation template perform actions inside the app.
For TikTok account warming, this matters. Watching videos and randomly following users inside the app creates a more natural mobile usage pattern than relying only on browser-based activity.
Keep accounts separated
When you automate random follows across many TikTok accounts, the account environment matters just as much as the action itself.
If too many accounts are logged in on the same device, or if many accounts share the same IP address, the setup can look unnatural. This becomes even more important when you manage accounts for different clients, niches, or regions.
A US beauty account, a UK gaming account, and a Southeast Asia shop account should not all share the same device storage, device identity, and network environment.
GeeLark cloud phones are built for account separation. Each cloud phone has its own Android environment, app storage, login state, and device profile. You can also assign different proxies to different cloud phones, so each TikTok account can run in a network environment that better matches its target region.
This follows the “one account, one phone” idea used in many multi-account workflows. It helps each account build its own browsing history, viewing behavior, and follow pattern over time.
It also gives you more room to test. For example, you can test a new warming routine on one account first without affecting the rest of your TikTok accounts. That said, the safer approach is still to keep your actions reasonable and stay within TikTok’s community rules.
Once you manage more than a few accounts, the problem is no longer just logging in. It is knowing which account is where, what stage it is in, and which task it should run next.
GeeLark gives you one dashboard to manage your cloud phones. You can name each cloud phone, add tags, write notes, organize accounts into groups, and quickly find the profiles you need.
This is especially useful when you are preparing to run automation tasks. For example, you can group TikTok accounts by client, region, niche, or warming stage. Then, when you want to run the Random follow on TikTok template, you can quickly select the right batch of cloud phones instead of checking devices one by one.
You can also handle setup tasks in bulk, such as creating cloud phones, assigning proxies, or preparing accounts for automation. Compared with buying physical phones and configuring each device manually, this can save a huge amount of time.
In short, cloud phones give you the infrastructure for large-scale TikTok account warming: native app usage, separated account environments, and centralized management.
Use automation template to auto follow
Once the infrastructure is ready, you can use GeeLark’s automation template to handle the repetitive task: watching videos and randomly following users in the TikTok app.
An automation template is a ready-made workflow. Instead of building your own script, you choose a template, add the cloud phones you want to run it on, adjust the settings, and let GeeLark run the task in the cloud.
The template works much like a real person using TikTok on a phone. It scrolls through videos, watches for a while, and occasionally taps Follow based on your settings.
And because the whole task runs in the cloud, it does not take over your computer screen. You can keep working, or even turn off your computer, while GeeLark continues running the task.
Set up the “Random follow on TikTok” template
Here’s how to set it up:
Download and install GeeLark. Sign up and choose a plan that fits your needs.
Create cloud phone profiles. It is recommended to set different proxies for your TikTok accounts, so different accounts can use different IP addresses.
Use “Applications” to install TikTok in bulk. This lets you install the TikTok app on multiple cloud phones at once.
Open each cloud phone and log in to TikTok.
Go to “Automation”. In “Marketplace”, find the “Random follow on TikTok” template.
Click the three-dot menu. Then select “Create regular task”.
Add cloud phone profiles. In the task setup page, select the cloud phone profiles that need to run the task.
Set the random follow probability and task time. I recommend setting different follow probabilities for different accounts to make the activity look more like real user behavior.
Save your settings. Then wait for the automation template to run in the cloud.
Go to “Logs”. You can check the task status there, or cancel the task if needed.
A quick note: start with a small batch of accounts first. This helps you check whether the watching time, follow probability, and schedule are reasonable before applying the task to a larger group.
