Initially, the brand approached us with a twofold problem. While up until then it had used exclusively Google Ads, it now began searching for a company specializing in the Meta advertising ecosystem to increase its market share.
The challenge
Despite their initial independent attempts, the company continued to face issues with low returns and difficulties in properly measuring analytical data.
The client was also afraid of an aggressive entry into a completely different channel from Google, which was primarily supposed to be Facebook and Instagram. However, they were aware that the competition had long been running paid activities on Social Media, while at the same time, by using only simple search engine ads, their store was falling behind.
During our initial conversations, we established that there was no reason why we couldn’t start the store’s paid promotional activities with small budgets in the spirit of cautiously “testing the waters,” and then plan the next steps (scaling, additional larger promotional campaigns, launching subsequent marketing channels).
The goal was to verify the viability of Meta as an appropriate marketing channel for this type of business, and to achieve at least a several-fold ROAS within the first few months.
Analysis and preperations
In a situation involving data measurement, properly addressing the problem is crucial – because as an advertiser, we must ensure proper data measurement even before we show buyers the first published ads. Otherwise, we would not be able to utilize any conversion-based optimization goals for our advertising campaigns.
Unfortunately, the client’s e-commerce platform “merce” did not make our task easier. Initial problems with the Pixel turned out to be platform modifications requiring intervention in the code.
In this situation, due to the inability to base advertising activities on conversions, we decided to build brand awareness through organic communication on the client’s Social Media channels and to build the top of the marketing funnel by directing traffic (without measuring impressions) to tutorial articles on the brand’s company blog.
After the first month of operations, the client recorded an increase in sales, which – despite the inability to directly measure the impact of Facebook campaigns – was undoubtedly caused by an influx of high-quality traffic in the form of people interested in home and garden topics.
If, like our client, you are facing problems regarding analytics or Meta Pixel events, contact us. We will conduct an audit, prepare recommendations for fixes, or implement the corrections necessary for the proper functioning of the Pixel ourselves.
Even though we achieved good campaign results without detailed measurement of micro-conversions, we knew how much more we could achieve once we implemented it.
In the meantime, thanks to smooth communication with the other party, we resolved the issues with domain verification (i.e., the so-called meta tag verifying the company website within the Meta Business Manager) and the correct implementation of events and micro-conversions.
Now that we had the ability to fully track the actions taken by the audience on the store’s website, we could fully focus on building a data-driven funnel.
Testing the marketing funnel
It is sometimes said that marketing is a battle of hypotheses. Therefore, taking into account the current data of the client’s store, and also considering the results of the initial campaign without implemented measurement, we decided to test the hypotheses put forward by our project team.
To this end, while simultaneously waiting for the remarketing audiences to populate after the fresh implementation of the Meta Pixel, we approached testing various interest-based audience groups.
We mixed these groups with different advertising objectives (e.g., viewing a product page vs. adding a product to the cart) and ad creations (single products vs. product groups vs. product categories vs. dynamic carousels with single products).
By applying this technique, we were able to identify the winning ad creations and initial audience groups over the next few weeks, which over time became suitable for further optimization. In this way, we verified our initially proposed hypotheses and could develop the entire funnel in the right direction.
Optimizations and building the middle of the funnel
During the execution of advertising campaigns for this brand, we worked closely with the Social Media communication department.
Because the client entrusted us with both the management of paid campaigns and the organic communication on the brand’s channels, we had the opportunity to use organic channels in a complementary way, supported by paid activities at the very top of the marketing funnel.
Complementing the communication strategy with paid activities – somewhat incidentally – over a few summer months, we prepared three contests that reached a total of over 50,000 people and actively engaged more than 500 individuals.
In this way, we were able to provide the brand with a continuously replenished audience group, which we later utilized within campaigns responsible for the middle and bottom of the marketing funnel.
By segmenting audience groups based on their engagement with brand content and actions taken on the client’s channels and the store’s website, we prepared the ground for scaling the results.
Even at this stage, we could boast a fivefold return on ad spend. However, we knew that this result could be higher.
Let's talk about cooperation
If we had immediately focused on bottom-of-the-funnel campaigns, we would have quickly led to the saturation of remarketing groups. This would have actually resulted in greater profit (instead of a ROAS of 5, we could have had a ROAS of 10), but it would have been a short-term action, because rebuilding the most valuable remarketing groups usually takes several times longer than their instant utilization.
Therefore, keeping in mind the goal determined together with the client – starting with small advertising budget amounts (in the range of 3,000-5,000 PLN) and scaling it over time – we built the top of the marketing funnel described above, followed by the middle and bottom, where we placed both static and dynamic remarketing campaigns, constantly testing different timeframes and types of non-standard audiences.
Furthermore, we supplemented the continuously optimized campaigns with new tests, including those based on Meta’s latest artificial intelligence solutions, which select audiences matching the brand from more broadly configured (than before) target groups.
Data-driven e-commerce scaling
Seeing our work reflected in the growing store revenue and the data collected daily on user behavior at every stage of the funnel, the client agreed to begin scaling.
In the first few months, we began the process cautiously, monitoring the results and analyzing the trends on an ongoing basis. Achieving over 900% ROAS with a slight increase in ad spend was a sign to us that the prior optimizations had been executed well, and that we could continue to scale the campaigns.
When scaling, we speak of a process because, due to the specifics of how the Meta advertising ecosystem operates, a sudden and drastic increase in budget compared to the amount previously spent can bring an outcome opposite to the desired one.
That is why we planned the first scaling process over several months, flexibly, depending on the results achieved. When scaling an advertising campaign’s budget, it may happen that a 20% increase leads to a growth in the number of purchases, but it might turn out that this is not accompanied by an increase in ROAS, as it would be if the budget were increased by an initial 10% and then by another 10% over time.
It was similar in this case as well – doubling the campaign expenditures did not guarantee a linear growth in ROAS. The goal, however, was to maintain it at a minimum of the level achieved thus far. Therefore, in the following months, after minor optimizations, we were able to double the budget, bringing us closer to the client’s goal (i.e., a fivefold increase in ad spend).
Further scaling and results
Midway through the campaign scaling period, with expenditures exceeding 6,000 PLN, we continued to increase (not as spectacularly, but still with an upward trend) the return on advertising investment.
Ultimately, we achieved the set goal, conducting the entire campaign scaling process over two quarters – growing from expenditures in the range of 2,000-3,000 PLN to an amount five times higher, i.e., spending about 15,000 PLN each month.
Thanks to this approach, throughout our entire collaboration, the client’s e-commerce recorded over 1,000,000 PLN in revenue, calculated solely based on raw data from the Meta Ads Manager.
In reality, we noted that the total value of conversions made in the store was several times higher, but due to the measurement imperfections of the Meta advertising system (the portal’s official statement estimates underreporting of conversions in the ad panel by up to 15%), the ad panel recorded “only” slightly over a million PLN in revenue.
Such a result was a very satisfactory outcome for both the brand and us, resulting from extensive activities and trust in our professionalism. Thanks to such a comprehensively built funnel, we provided the client with basic resilience to market fluctuations, as well as a steady “influx” of fresh audiences.
Maintaining such a result in subsequent periods allowed the client to gain additional benefits in the form of a better negotiating position and the maintenance of relationships with manufacturers and suppliers of the brand’s key products, through reach campaigns concerning specific product lines.
The consistent ROAS achieved during the scaling process also allowed us to conduct broader image-building activities and reach campaigns, which, although they do not generate a return per se, position the brand higher in the awareness of future customers compared to market competitors.
Can such a success be replicated?
At Moonwise, we know what works and what doesn’t. However, we never rely 100% on our experience; every time we formulate a hypothesis, we juxtapose it with others and test various solutions so as to tailor them to a specific case.
The results described in the Case Study can certainly be repeated. It is probably also possible to achieve even better results. However, the success of such a campaign depends on many factors that you had the opportunity to learn about in this article. One question remains: do you want us to take on the challenge of generating such results for you as well?
If you are looking for an agency that will take care of your advertising campaigns – stop looking and contact us. We would be happy to discuss how we can help you.
