Every performance marketer is working towards increasing their ROAS daily. They have their own experiences and stories. There are a lot of variables which make each industry different and hence the advertising is also different. Here are the learnings from my experience of working with an Indian Direct to Consumer (D2C) FMCG brand.
1. Structure your conversion campaigns basis the customer buying journey or marketing funnel
As the first step towards improving ROAS, we created a marketing funnel and categorised our audience in 3 basic categories –
- Cold or Top of the funnel: This included people who have never heard about us. If the spends are high, then it’s better to have at least 2 cold campaigns. This is because when you must make changes across campaigns (like changing the ads) then you can spread that change over a couple of weeks. As a result, all your ads will not go into review and then learning at the same time. For example, in our case we had 2 cold campaigns – one had lookalike audience and the other had location-based audiences.
- Warm or Middle of the funnel: People who have heard about our brand or People who have visited our website but did not initiate a purchase. We created 2 warm campaigns, one was to target people who have engaged with us on Instagram and Facebook in the last 30 days. Another one to target people who have visited our website (product page) in the last 20 days minus people who have visited in the last 10 days. People who have visited the website product page in the last 10 days were included in the Hot audience
- Hot or Bottom of the funnel: People who have visited our website and added products to cart or initiated checkout. This campaign was basically our retargeting campaign. Here we had 2 ad sets for 2 different audiences. The first had people who abandoned the cart or checkout and the other was of people who have visited our product page in the last 10 days.
The 5 different conversion campaigns that we have are:
Use proper tracking parameters to track the performance of all the campaigns. Read the blog for a better understanding of the same – The Easiest and Best Way to Use UTM Parameters
2. Displayed multiple flavours or products in one carousel ads
We had very limited products, 3 to be specific which we wanted to advertise. Earlier we created different campaigns for different products so that we can control the spends on each product. But this was flawed in many ways. The customer was not exposed to multiple flavours or a variety of products that we had. Also, there are people who might like flavour ‘x’ and others might like flavour ‘y’. So, we really don’t know which flavour will make them click on the ad.
3. Have one audience per ad set
Multiple audiences in one ad set might give good results, but only for a short run. We realised that in the long run when we wanted to scale our revenue and spends it was difficult to understand which audience was working, which was not.
4. Excluding all other audiences from every ad set
If we have 10 ad sets with 10 different audiences, then all other 9 audiences must be excluded from every ad set. This helps in trimming the spends by reducing the overall frequency of the ad account. Which leads to a reduction in spends and hence increase in ROAS. For example, a person may be in one of my lookalike audiences and the same person may be in one of the location-based audiences also. Hence increasing the frequency.
5. Using Facebook Analytics to dig deep and refine the audience
Once we have structured the campaigns properly and the campaigns have run for at least a couple of weeks.
We should deep dive into Facebook analytics and find answers to questions like –
- Is there any specific device (Mobile/ Desktop etc) or OS which is giving us maximum conversions or ROAS?
- Is there any specific gender, age-group which converts more?
- Which placements are working better for us, by giving more ROAS?
Click here to understand Facebook analytics and how to use it.
6. Switching to Manual Placements from Automatic Placements
Selecting placements which give us maximum ROAS has also contributed to this journey. It made little difference, but it surely did make a difference. Automatic placements are good for traffic and brand awareness campaigns but my personal experience says that it doesn’t work for conversion campaigns.
7. Always keep Experimenting
Experimenting will give you a good understanding of what is working for your industry/ brand/ products. Every industry is different and hence, they see different consumer behaviour. So, define a hypothesis that you want to test and run an experiment. After collecting data for around 2 weeks analyse the result and implement changes. As digital marketers, we must always keep running experiments.
So, this is how we made our way to achieve an increase of 100% in ROAS. Please do share your thoughts and experiences in the comments below.