Today I’d like to present a special case study from one of my favorite accounts. Fashion niche, Average Order Value (AOV) 150 RON. Before I take you through this entire journey, I need to mention a few important aspects regarding data analysis.
The account in question has had tracking issues and still does. Sometimes orders are attributed twice, so all calculations were made on unique orders, and the difference in percentages was calculated despite the figures in Ads Manager.
The business is wonderful. When I took it over, I knew about the existing problems in Ads Manager, and the client’s expectations were brought back down to earth. However, after 3 months of collaboration, we have an absolutely wonderful growth!
The client’s requirement was simple and had 3 well-defined options.
After 3 months of collaboration, we aimed to achieve one of the following:
- ROAS (Return on Ad Spend)
- Volume
- ROAS + Volume
The well-trained pixel and advertising history helped greatly in scaling the results comprehensively and improving performance.
Before I dug into the account, I took a look to see what was happening and noticed three major “bumps”:
- There were no exclusions for buyers
- There were no tests
- There was no scaling on profitable audiences
Being a business with impulse products and momentary stock, the mission seemed difficult initially, but after the first two weeks, I started to see the expected results.
For testing, I used the following variants:
- ABO – Macro Budgeting (+50 RON / ad set)
- ABO – Micro Budgeting (under 50 RON / ad set)
- ABO – Lowest Cost
- ABO – Cost Cap
- ABO – Bid Cap
- CBO – 3 audiences each
Since this was an account with history where the client knew the audience we were working with and the most successful creatives, there was no need to test creative variants. It was known from day one that the best results came from manually made Carousels, and we started directly with them.
After 15 days, out of all the tested variants, we were left with ABOs under 50 RON as the most stable. Coincidentally, Micro-budgeting corresponded with the instability in October and performed better than larger budgets. I’ve observed a similar trend in other accounts when there are stability issues. In 6 out of 10 accounts, Micro-budgeting manages to beat Macro-budgeting.
To organize myself better, I created a separate campaign which I called Ad Structure, and within it, I always create ads from scratch. Why do I do this? It helps me a lot with organization. If I know I have an ad that sells, I can always find it in that campaign that never runs, but which includes all the ads made with their names and POST IDs so I can easily take them for duplicates.
The initial daily budget was 500 RON, and scaling was to go up to almost 2,000 RON daily.
In the first 45 days, I couldn’t exceed 500 RON because the performance wasn’t showing good signs, but as soon as I felt there was room to scale, I started gradually adding more ads to the funnel. For a good period, a dynamic retargeting campaign was running, but I had to stop it because the number of duplicate orders was huge and it was no longer profitable.
From Dynamic Retargeting, we switched to 3 different types of retargeting.
- Manual Carousels
- Single Image Creatives
- Videos
Throughout the retargeting campaign, 20 different ads are constantly running. About 3-5 per ad set.
Once we identified winning audiences + winning products (around 100 different products were tested), we began scaling. Initially, scaling took place with a combo of ABO with a large budget + CBO with an even larger budget. The strategy wasn’t very profitable in the long term, but for 7-10 day bursts, if there are signs that campaigns are starting profitably, I highly recommend pushing budgets as much as possible.
Below you can see several screenshots attached.
June and July are before I took over the account. I didn’t note August because I took over in the last 10 days and it wasn’t worth displaying the results from there. September, October, and November are the first 3 complete months in which I worked on the account.
The statistics in Ads Manager are different from the real ones, but I did the calculations and it came out like this:
June
- 490 Total Sales
- 245 Unique Sales
- 50% Duplicate percentage
- REAL CPA: 53.2 RON
- REAL ROAS: 2.75
- REAL Sales Value: approx. 35,000 RON
July
- 457 Total Sales
- 250 Unique Sales
- 46.3% Duplicate percentage
- REAL CPA: 33 RON
- REAL ROAS: 3.2
- REAL Sales Value: 33,703 RON
SEPTEMBER – First complete month with us on the account
- 765 Total Sales
- 418 Unique Sales
- 45.36% Duplicate percentage
- REAL CPA: 35.39 RON
- REAL ROAS: 3.69
- REAL Sales Value: 68,863 RON
OCTOBER
- 788 Total Sales
- 316 Unique Sales
- 59.9% Duplicates
- REAL CPA: 30.49 RON
- REAL ROAS: 3.56
- REAL Sales Value: 53,512 RON
NOVEMBER
- 1221 Total Sales
- 956 Unique Sales
- 21.7% Duplicates
- REAL CPA: 31.73 RON
- REAL ROAS: 5.28
- REAL Sales Value: 168,349 RON
After 3 months of collaboration with this client, not only did we manage to double the number of orders, but we also raised the ROAS, and the AOV increased with the help of optimizations at the level of product package suggestions.
P.S. – The account benefited from a multi-channel strategy with Google, Email Marketing, Blog content, and affiliation with people who bring extra orders from outside the ads.
What conclusions can we draw from this?
- Budgets play a HUGE role in any business. The more you give, the more you get.
- Patience is VERY important. Many accounts that are just starting don’t have the luxury of historical data in the Pixel, which, in the first 60 days, discourages the business almost entirely. Unfortunately, in 2021, we’re heading towards this trend where if you don’t invest considerable amounts to teach the Pixel who your people are, you won’t be able to reach them.
- Patience again. In 2020, we often encountered potential clients who wanted results overnight or from one month to the next. Unfortunately, this is not possible, and whoever guarantees you results in 30 days is either fooling you or will ask for a huge advertising budget, which isn’t wrong. Yes, results can be brought after the first 30 days, but only with a budget of €10,000 and up spent in a month. It’s an amount that covers the market entirely and reaches the entire population of Romania, so the Pixel learns very quickly what and how. But it costs.
Don’t rush to stop the ads. We had two consecutive weeks on this account where the ads didn’t even reach breakeven ROAS. This was due to the election period + advertisers entering and leaving the market + other advertisers pumping even bigger budgets in the same period. We didn’t stop the ads completely, but we reduced the budgets to a minimum that wouldn’t hurt.
Throughout this process, don’t forget the key factor: constant communication with the client. Results can appear and be satisfactory or encouraging only when everyone “puts their shoulder to the wheel”. You do your job as a marketer to promote the business as efficiently as possible and to bring results, and the business owner ensures that they deliver impeccably to the end customer, so that everything goes like clockwork.
