Target advertising: Wood porn & US subscribers for 10 cents

The Woodclans project came to us and requested assistance in attracting a target audience to their social networks. With the help of target advertising, we have brought 26,000 people on Facebook from the ground up and reached a price of 10 cents per subscriber. In this article, we are going to tell you how exactly we get the job done.

The objective

Woodclans is a community of woodworkers. The project also has an Amazon store where they sell furniture, equipment, and accessories. Woodclans contacted us at the time of their entry into the market. Our objective was to create a high-quality audience in social media among US citizens and bring the price for one subscription to $0.1.  

Choosing the platform 

Initially, the project has been being developed for the US and Canadian markets. Therefore, we created accounts on Facebook and Instagram, filled in bio information, and published about 10 posts on the topic.

Identifying the audience segments

Together with the client, we determined the target audience and relayed this information on the mind map. We ended up with 4 segments: carpenters, designers, craftsmen, and brands. 

We have singled out the segments we chose to work first with - carpenters and craftsmen.

Developing the designs

We decided that the visuals should represent a handmade wood product and a short message calling to action.

The sources of inspiration and ready-made images for the designs were Pinterest, Tumblr, Instagram, and theme groups on Facebook.

The first designs looked like this:

Launching the advertising campaign

We launched target advertising on Facebook in all regions of the United States. The audience was made up of carpenters and craftsmen. The first results ranged from $0.2 to $0.33 per subscription.

Generally speaking, it was not so bad for such a specific audience- but our goal was 10 cents. So, we started to think about how to improve performance. 

If we hadn't had any reaction, even a negative, to the advertisement, it would mean that we had made a mistake. But since the overall feedback was positive, it wasn't just the scale we wanted. So, we decided that the audience was right, but the designs needed further optimization.

We changed the design, made the slogan more visible, added a "join us" message with an arrow to the Subscribe button. 

Also, while in the previous case the ads were displayed throughout the U.S., this time we segmented the audience by region. We chose the most active states based on the results of the previous campaign.

The outcome was controversial. Geolocation optimization did not yield results. The cost for a subscriber didn't change. But we knew that we are operating within the right audience. The people were chosen correctly, it’s us who showed them the wrong things.

Advertising campaign statistics screenshot

Launching the video advertising campaign

In August 2017, Facebook launched its Watch video platform and began to actively promote it. At that time, the priority was given to video advertising: 1,000 video views were cheaper than those with static visuals. We decided to try this format. In addition, it should have been a better way to involve the viewer. 

That was exactly when the client had finished editing the first videos for his YouTube channel. We took those as a reference. 

The first videos were about how to make a wooden guitar wall hanger. We decided to experiment with the first 3 seconds: we made 8 different intro versions, came up with video backdrops and put everything on the target. The results we got were about +/- 10 cents depending on the first 3 seconds of the video. Thus, we were able to see which videos showed the best conversion.

A subscriber price after we've launched the videos

So, we had $0.36 per subscriber. We still had a lot of work to do before we reached our goal of 10 cents. We decided to follow the same course, but to pay even more attention to the videos. When we were thinking about the ways of content presentation, the idea of food and car porn came to mind, and we were like, ‘What if we try wood porn’?

The perfect solution seemed to be GIFs showing people drilling, cutting, whittling, varnishing, etc.

Back then, the client was still actively making those videos for YouTube. So, we briefed the masters about exactly the kind of content we needed.

That's what we were looking for:



We put the GIF on targeting with a small budget and what we saw was the positive dynamics: $0.27 per subscriber.

The best versions by conversion

Then we added a call to action to the vertical GIF, and here's what we got.



The prices in the best segments:

By comparison, the static visuals cost us more than 30 cents in the same regions.
Design examples for one group of ads
Examples of groups which were working

As a result

Along with the selection of designs, we never stopped analyzing the results and segmenting the audience in order to show the right visuals to the right people. After 1.5 months of experiments, we came up with a formula which allowed us to get a subscriber at the right price - for an average of 10. As a result, Woodclans now has a community of 26,000  people on Facebook. 

It is also worth mentioning the quality of this audience, which we checked every time we ran another campaign. When there was an influx of new subscribers, we manually looked at their profiles, approximately one-third of them. That's how we knew if the audience we were attracting was right. The results were good: it was almost always clear from a person's account that he or she has something to do with wood. We have formed a very specialized and involved community.

After that, based on the users' interaction with our advertisement and Facebook profile, we segmented the audience once again for retargeting. We launched look-a-like for the entire USA. This way, our ads were only shown to the users similar to our subscribers.

After the successful formula to attract the right audience in the U.S. was developed, we applied the same scheme for the Russian and European markets, where this method has also succeeded.