Running an experiment on social media: TikTok
One thing we love here is experimentation.
Growing a business isn’t just about developing a good idea, a solid strategy, or even an incredible marketing campaign. The most successful clients I’ve worked with are experimenters. They’re willing to fail learn fast by testing things to know if they work or until something does. I love the quote below from Naval Ravikant, co-founder, chairman, and CEO of AngelList.
“Solve via iteration, then get paid via repetition.”
Test, test and test til you can scale. Here’s an experiment I ran with a client.
The Brief
A business-to-consumer (B2C) client wanted to expand its social media channels. In terms of marketing structure, it had a website, an Etsy store, and an Instagram account. Sales were growing but quite slowly. Most of its sales were through searches on Etsy or directly on the website.
The challenge for this client was ‘how to market their products’. Their flagship products contain the c-word, and there are often advertising rules against this on social media platforms. Surprisingly or not, they found that their stationery products appealed to women, mainly mums, who would buy these gifts for their significant others or adult children.
Instagram was working to build brand awareness and connect with the desired audience, but it was slow. The virality potential of TikTok as a platform appealed to the client, so we decided to test it. Here’s our experiment.
The Experiment
Experiment Title: From 0 to 1000 in 30 days
Objective: To grow a TikTok audience from 0 to 1000 followers within a month while targeting females aged 28 and above.
Hypothesis: If we use various content types, engage with the audience, and invest in advertising, we can successfully grow a TikTok audience of 1000 plus followers within a month.
Experimental Variables:
Independent Variables:
1. Content Types (various formats, visuals, music)
2. Engagement Strategies with Audience
3. Advertising Spend ($200)
Dependent Variables:
1. Number of TikTok Followers
2. Audience Demographics (age, gender, location)
Experimental Design: Create and post diverse content on TikTok, engaging with the audience through comments, likes, and shares. Allocate an AUD$200 advertising budget to boost visibility to reach this goal faster.
Timeline: 30 days
Target: Achieve 1000 followers. Typically, we focus on engagement rather than followers. However, for this exercise, we wanted to learn about the audience and what works in terms of content types, so we decided to stick with follower count as a base metric.
Implementation Plan:
Step 1: Create a Content Calendar with Various Content: products, how-to’s, behind-the-scenes, playful videos
Step 2: Engage with the Audience through Comments and Direct Messaging
Step 3: Allocate the Advertising Budget and Monitor
Data Collection and Analysis:
Utilise TikTok analytics and tracking tools to collect data on follower growth, engagement metrics, and audience demographics.
Analyse the demographic breakdown of the audience, engagement rates, and the effectiveness of different content types. Compare these metrics against the success criteria.
Expected Outcome & Success Criteria:
The experiment will succeed if we achieve 1000 followers within the month and most of our audience aligns with our desired demographic profile.
Results & Conclusion
To be honest, the experiment failed to engage with our desired audience of females aged 28 and above. However, we successfully achieved the goal of reaching 1000+ followers on TikTok within a month, including high engagement with the audience we did attract.
What worked
The highest-performing piece of content was a video based on a submission from an audience member. That video now has over 115k views. Based on our analysis, the reach and success of that video was due to several reasons:
It was user-suggested, so that particular audience member shared it with their friends
We actively engaged with the audience on that piece of content through comments and likes
It was shareable. The content had a clear call to action, asking viewers to share and tag their friends.
What didn’t work
Due to the comedic yet rude nature of the products, despite utilising various content types and engagement strategies, including audience filtering for location and age, we found that our audience consisted mainly of <18-year-old males and females from the US rather than our desired audience in Australia.
Here’s a breakdown of the audience reached
We reached a predominantly female audience, however, not the right age group who would convert to buyers of our client’s product.
What could have been better?
Relying on TikTok's targeting and filtering isn’t enough to find the desired audience. We’ll reassess our targeting strategies and content approach to better align with our target demographic.
We’ll explore the analytics to understand engagement and how to build more conversations with the audience.
What else did we learn?
The first piece of content created on TikTok will generate a reasonable amount of organic reach. This is due to the algorithm manufacturing a ‘sugar hit’ for the user so that you come back for more. From there, videos will drop significantly until you start posting regularly and boosting with advertising spend.
Although the platform allows filters based on age, location, etc., the audience attracted during this campaign was well outside the desired demographic. Our assumption here is that users on TikTok will have multiple accounts, enabling them to access content outside their age bracket.
What’s next?
We plan to explore alternative platforms and tactics, such as user-generated content with creators, to effectively reach our desired audience.
