Ever wonder how big Black Friday can be for a newly established brand? Here is a an awesome case study that was published by Tomasz Babiec, a Facebook marketing expert specializing in ecommerce. There are many people struggling to get sales for new brands, but this case study shows that with a great product, good website and the right approach, you can sell a shitload of it on Black Friday by giving an amazing deal that’s better than all the deals throughout the year. He talks about strategy, he talks about targeting, and he talks about optimization and what I love the most, what he would’ve done different. Thanks Tomasz for letting me share this with all my readers!!
⬛️💲Black Friday Case Study: How We Made $51,167.60 For A Client On $7,240.42 Ad Spend (707% ROAS!)💲⬛️
Hey everyone, hope you guys crushed Black Friday! I’m writing this post to share a case study around this year’s Black Friday campaign we did for a client. This is raw and fresh information that you can use in your own campaigns!
Quick background story:
▪️This campaign was for a premium brand that is not available anywhere else other than on the client’s website, so it may not work for affiliate offers.
▪️A normal day of sales on this store is about $5k on average, sometimes reaching highs of $15k and lows of $3k.
▪️95.7% of the ad spend ($6,930.35) was spent on Facebook, the rest was Google Ads spent across brand keywords and several niche ones.
💊💊 The recipe:
▪️You really need to have a great product to confidently push it. If your product is average, you risk getting negative feedback on your ads (measured in comments and in the results of surveys Facebook sends to people) and dealing with refunds/chargebacks.
▪️Prior to big events like this, you NEED to upload and publish your campaigns in advance (that means setting up campaigns, targeting them, giving them initial budgets, and scheduling them). Many people here have posted about this and it is absolutely crucial.
▪️Your offer should ideally be something that is not normally available for people during the year. A great offer was what made this campaign. If you give 30% discounts all year round, doing that again on Black Friday may work better than on other occasions, but most probably you aren’t going to make you a killing. We did a buy 1 get 1 free offer to drive up AOV and don’t kill margins too much.
▪️As others have suggested (shoutout to Nick Shackelford!), you need to drive traffic to your site several weeks before the offer goes live to warm people up so you can later retarget them with your offer. Your short-term ROAS will decline, but it will pay off. We didn’t say anything about the offer until 12:00am Black Friday, so people weren’t sure if we were going to do it.
Needless to say, on days like this, you don’t go anywhere without your laptop, 100% battery, and good internet!
💊💊 The setup:
❄️ Cold audiences: 3-10% LLAs (mostly 6% and 8%, apart from that I had 1×3% and 2×10%) of audiences that were historically performing very well, that is: 30 days FB engagement, customer LTV lists, Facebook’s new “value-based audience using pixel data”, 4% top pageviewers (shoutout to Tyler Garner who posted a video with this a while ago. You create this audience through FB analytics by checking how many times the top 4% purchasers trigger the PV event).
💰 I had 6 cold ad sets total, each with initial budget at $1000 and autobid. I later switched to manual, more info below. I admit I was a little too scared to go straight for Tim’s recommended shotgun and surfing setup and do $16k/day 😃
Literally anything you can retarget. This is where you make money from the traffic you were driving in the past few weeks. Target FB engagement, insta engagement, VCs, ATCs, past customers, newsletters, all kinds of website events, 75% VVs, literally anything that is relatively recent or that worked in the past (for example I wouldn’t target FB engagement that’s 180 days old, IMO that’s a little too broad for something like clicking your post). Depending on the size of your audiences, it may be better to put them together in one ad set, I had 2 ad sets.
💰 Budgets: $400/ad set for the start.
VC/ATC: past 3 days, past 7 days, past 14 days (this one did bad compared to the other DPA ad sets), past 30 days, and cross selling to people who bought in the past 60 days.
💰 Budgets: $200 per ad set for the start.
🤖 Rules (Revealbot) – scheduled every 15 minutes. Target CPA was $130. I attached a screenshot of the rules I was running on the cold campaign, I also had rules I was running on retargeting and DPAs.
📧 We had a BF email that went out early in the morning on Black Friday.
💊💊 Campaign execution:
I checked stats every 1-2h, I didn’t want to do it too often because for me personally it does more bad than good. Patience is a virtue!
I really wanted to test Tim’s shotgun and surfing for the first time and my aim was to start bumping budgets as fast as possible, as soon as I started seeing initial proof that things were working. As I said earlier, I was scared to do a $16k per day setup, but conversions started flying in so quickly that I didn’t even wait for my reveal rules to kick in to increase budgets. I knew this was black Friday so the odds were kinda in my favor, so as soon as ad sets got 3 purchases, I doubled their budgets.
It quickly turned out that Facebook wasn’t spending quickly enough, so I went from auto bidding on all campaigns to manual. For cold ads, I first started with a bid cap of 3xCPA goal ($390), for retargeting I went for $450, and dynamic something like $600.
After letting that run for 1-2 hours and seeing the hourly breakdown of spend, I decided to give the best cold ad sets $10k/day budget and bump the other ones to $4-5k/day. I set DPA budgets at $1000 each and retargeting $2000 each. It still wasn’t spending quickly enough so I maxed out bids on dynamic ads and retargeting ($1000 USD bid cap) and set them to accelerated delivery. I did the same thing with cold ad sets, the best ones had accelerated delivery, $800 bid caps (not sure why I didn’t go for $1k) and $2-3k budgets.
I admit this was a little crazy as some ad sets started spending $500 per hour so I had to decrease their budgets from the $10k I set a few hours ago. They ended up at 2-3x the spend they had when I set them. My logic was that the daily budget was acting as a stop loss. If a rule I had set up wouldn’t kill it quick enough, the daily budget would run out. I was increasing budgets on accelerated ad sets and carefully monitoring performance. Surf with the waves!
My total daily budget at one point was probably $25-30k, though not all of that was accelerated.
Sales were flying in like crazy all the time. First I wanted to crack $30k, then 40, 50, and 60. For a few final hours of the day I was watching the account very closely (anyone know that adrenaline rush when you set a lot of ad sets on accelerated delivery and crazy bid caps? 😀) and 30 minutes before midnight ad account time, it completely stopped spending. I panicked thinking it’s a manual review that would shut us down for the rest of the BFCM weekend.
I still don’t know why but it only started spending at 12:00am Saturday ad account timezone. All credit card charges were successful, there was plenty of budget left, all important ad sets were live at that moment, yet it stopped spending. I admit I was disappointed because I wanted to get closer to $60k, but my hands were tied.
Conclusion: MAN, I exceeded my most ambitious expectations!
Shotgun and surfing works MIRACLES if you do it right! It wasn’t ideal (too much manual work on my end), but it worked really really well. I highly recommend this to anyone for events like this, I’ll be looking forward to testing it on another occasion without a sale really soon!
Now, some thoughts after analyzing everything.
What I would have done differently looking at it now:
💵 Higher budgets (definitely would start with $16k/day)
💵 Manual bids all the way from the start
💵 HIGHER BIDS! Bully them!
💵 Spend more budget on retargeting and DPAS
🎯 Broader targeting!
▶️ I would use the “Black Friday Cyber Monday” angle from the beginning. Doing Black Friday only limited the offer too much.
🤖 Set more automated rules!
Thank you for making it this far! It means a lot for me that you read the whole post so thank you once again 🙏🏻
I hope you found it valuable.
Do you have any questions? Let me know in the comments 👇🏼