Unlocking the Potential of AR: A Cambodian Advertising Success Story

Our client was unique and not a run-of-the-mill one. Total Energies is an international oil and gas company, ranked 29th on the Forbes Global 2000 list for 2020. It's one of the top seven companies in the industry and falls under Big Oil. We collaborated with the company's branch in Cambodia, where Total Energies operates petrol stations and sells fuel in the B2C sector.

Total Energies sought opportunities to increase brand awareness for Café Bonjour, Wash, Excellium, and lubricants (Quartz, Hi-Perf) among millennials through real-time interactive games.

The overall mechanics of the gaming campaign were as follows: users clicked on a special link to access a chatbot => played one of the AR games and earned points => exchanged the points for branded prizes.

According to our calculations, the cost per lead acquisition amounted to $2.59.

Although my colleagues and I were accustomed to not placing too much trust in case studies (two identical campaigns under identical conditions could yield different results just because that's how Facebook operates), not examining publicly available sources of someone else's experience at the time would have been rather odd. Especially considering the foreign market, the non-trivial niche, and the atypical format.

Paradoxically, it turned out that AR was hardly used in advertising at all—there was only one reasonably decent article about the airBaltic advertising campaign using AR, posted on Facebook itself, effectively promoting the new format. However, the article didn't reveal anything fundamentally new, and we simply reaffirmed the indisputable truth:

The benefits of this approach are twofold: in the initial stage (traffic), we obtain relatively cheap ad clicks and help Facebook algorithms adapt to the brand and its audience. Additionally, through the pixel, we gather data for future Lookalike Audiences (LaL). In the second stage, we set conversion goals and reap the benefits of the groundwork from the first stage. Algorithms learn faster, leads become cheaper, and everyone wins.

Our situation had a dual nature. On one hand, we directed traffic to the chatbot, which meant that for effective optimization in the second stage, we needed to use the "Messages" objective. On the other hand, the chatbot game only opened when users clicked on a specific link, which couldn't have the pixel placed on it. So, while the actual objective was conversions, using conversion objectives was not feasible as Facebook wouldn't optimize properly. Therefore, we settled for the "Traffic" objective, and it turned out to be a wise decision. In our initial tests, we achieved a very low CPM and, consequently, significant reach that comfortably met our KPIs. In addition to our primary advertising, we boosted specific promotional posts from the company's own page to further increase audience reach and awareness of the campaign.

This is an important point for those who want to replicate our success or outdo us in terms of coolness (we'd be delighted for you). The Cambodian audience has some unique characteristics.

Firstly, local residents are highly active on Facebook, which is their primary social network (approximately 70% of the local population uses it), and Facebook Messenger is their primary communication tool. It's so essential that around half of the global voice message traffic in the messenger comes from Cambodia. While in Russia and the CIS, when talking about the Facebook/Instagram combo, we mostly refer to Instagram, in Cambodia, it's the opposite. Therefore, right from the start of the advertising campaign, we focused exclusively on Facebook and didn't spread our budget to Instagram (in other words, we abandoned the practice of auto-placement).

Secondly, private vehicle ownership and active use of both cars and motorcycles/mopeds are common among the region's population. Therefore, initially, our audience consisted of virtually all Cambodian residents who had reached the age of obtaining a driver's license.

During the optimization process, more so than just targeting, we discovered an interesting insight: users over 35 were more likely to click on ads but were less likely to engage with the chatbot. Even though we initially achieved impressive results with a 10% conversion rate from link click to full-fledged lead in our first tests, by adjusting the age targeting, we managed to further increase the conversion rate.

Based on the experience of our colleagues at Morozzi, residents of Cambodia aren't particularly inclined to "read between the lines," engage in abstract thinking, or seek hidden meanings. They interact with advertising using the principle of "what I see is what I understand." Understanding this specificity was crucial in creating creatives and texts: there were no plays on meaning, just extremely clear and straightforward instructions at every stage and in every element of the advertising.

Static Images: Either featuring popular regional bloggers or focusing on a specific prize.

Videos: Generally, we replicated the style and content of static images but added animation and music.

AR Creatives: Users could try out games directly from the ad and then transition into the chatbot.

Based on the test results, all three types of creatives showed roughly similar levels of engagement and interaction with the ads. Plus, the cost per click was lower!