I run our Meta ad account 100% on math, and game theory.
A blind monkey could run it 🙈
And still, we're having our best month since peak Virus-Who-Must-Not-Be-Named.
- No 3rd party pixel.
- Zero daily maintenance.
- 100% focus on creating new ads.
Here's the secret sauce:
Today, I taught our new hires the framework I use when I write product descriptions.
It's not always applicable, but when it is, it converts 269% more comparing the CR of our most applicable products vs. our best-selling non-applicable products.
Here's the framework:
BREAKING:
@googleanalytics
4 is here for
@Shopify
🤩
All native. 100% plug-and-play. No coding. It's so easy! I literally just set it up.
Here's how to set up GA4 with your Shopify store, step-by-step:
I'm planning to build the most high-converting Shopify theme out there.
We don't need 1000's different themes.
We need one high-converting theme, based on CRO best practices, that we can use as the foundation to build custom sites on top off—or use out of the box.
I'm 29.
Three years ago, I quit my hedge fund job to sell clothes to +65-year old women with my mom.
I've never been happier.
Lesson? Don't chase money. Chase happiness. No amount of money is worth being miserable.
CRO tip: Don't show free shipping info in your announcement bar.
In a study, 27% of subjects overlooked free-shipping info in site-wide elements like the announcement bar due to banner blindness.
Instead, show it near the buy box on PDPs, e.g. above the add to cart button.
The more I work on my conversion-focused Shopify theme, the clearer it becomes that even the most popular themes don't know much about CRO.
I used to think their shortcomings were because of technical restrictions. But if I can build it, they should be able to too.
Examples:
Google Analytics 4 (GA4) is here. Here's why this is awesome, what you need to know about GA4 and what you need to do NOW as an eCommerce business/marketer👇🤓
Most Shopify themes are designed to be pretty, not to sell.
And that's your luck!
That makes optimizing your online store a competitive advantage. You can literally double your business by improving your site.
Here are 9 articles on optimizing your site, aka weekend reading:
Do you want to learn how to
1. set up Google Tag Manager?
2. set up Google Analytics 4?
3. set up Google Optimize?
4. run your first A/B test with Google Optimize?
As of today, all 4 guides are included in my "How to design high-converting product pages" 🔥
Coming soon 👀
Do you look at CTR to evaluate your Meta ads?
You probably shouldn't.
In a study of data across 262 brands, Facebook found no correlation between CTR and ROAs.
No correlation. R^2 = 0.0001.
I just finished an A/B test that ended with the craziest result I've ever seen for such a small change.
It's so simple, and yet it increased our revenue/session by 40%. I'm honestly embarrassed that I haven't thought about it before.
Time for a thread 👇🏻
Shopify SEO Tip:
Avoid duplicate products by pointing all your product links to "/products/product-name" instead of "collections/collection-name/product-name"
How? Replace {{ product.url | within: collection }} with {{ product.url }} in all relevant places in your theme files.
I'm really starting to question whether
@Shopify
is the best fit for us moving forward.
They just announced they will start charging a percentage fee to calculate taxes (0.25%-0.35%, with a maximum of 0.99€/order).
Apparently, calculating the 25% VAT on each order is so
Since iOS14, I've been struggling with the performance of our Facebook ads.
If you're also struggling, just know that you're not alone.
Maybe we’re just not hitting it right now. Maybe FB is struggling even more. We don't know.
The important part is that we don't quit.
I've never been able to get my Broad Dynamic Products Ads (BDPA) to perform. You know, the dynamic product ads aimed at a broad audience for acquisition rather than retargeting.
That is until I made a simple optimization. Now, I consistently see a ROAS > 3.5.
Here's what I did:
A high ROAS doesn't matter.
Your goal as a media buyer is not to find the most profitable ad or get a high ROAS.
The goal is to spend as much possible, profitably.
An ad that generates 1000 orders at ROAS 2 is a far better ad than one that generates 100 orders at ROAS 4.
Scaling too fast has caused us more bad than good:
- Lost sleep about how to meet ambitious sales goals every month
- Overdependence on Meta ads
- Cash flow issues from large inventory purchases
In hindsight, I would have grown deliberately slower YOY.
I’m starting to think that I've spent too much time in our ad accounts and too little time on CRO.
Try calculating how much marketing money you could have saved if you converted 30% more of your traffic.
And that’s pure bottom line. Pure profits.
Crazy, right?
THREAD: How I lifted our conversion rate by 60% by redesigning our navigation.
To put 60% in perspective, I would have generated 1.000.000 more in sales this year if I had implemented this January 1st instead of recently. That's a lot of money I missed.
Read this so you don't:
Your landing page builder is a waste of money.
Instead, develop 5-7 Shopify sections designed to convey information efficiently.
You save money, your site loads faster, it's easier to A/B test, and you're consistent.
I made these yesterday—100% customizable in Shopify.
You've heard all about how a slow site kills your sales.
Here's how you determine exactly what slows down a page:
Open Chrome -> Right-click, Inspect -> “Network” tab -> Disable cache -> Select “All” -> Load your site -> Sort by time.
(Images are the easiest to fix)
It's crazy how much photos can do to the perception of a product.
These two photos depict the same winter jacket from two different online stores.
Which one would you rather buy?
Can we please stop focusing on CPM’s?
Why do you care about how much it costs per 1000 impressions? What are you going to do about it if it's high? If it's low?
Focus on metrics you can control.
The DTC career ladder:
1. Build a company with a lot of revenue, but no profits
2. Start a newsletter or podcast
3. Post sponsored content for 36 companies you love and use every day but now suddenly just happen to be sponsoring you
4. Start 7 other companies because you're such
I'm not proud of this. But in my former quant life, I considered marketing a B-job. It certainly wasn't as important as what I was doing.
Now, I consider marketing the most important skill of any founder. FB ads is the reason we went from $60K to $3M/year in 3 years. No doubt.
Shopify just launched a "Pixel Helper" that lets you debug, e.g., your Meta pixel in real-time with all data included in the events.
Find it in your Settings under "Customer events" and click "Test" on the pixel you want to test.
Good job,
@Shopify
👏
Two observations:
1. You run more apps than our $3M store. You don't need post-purchase surveys, SMS, multiple channels, cashback, or paid theme sections before you even have a sustainable business. Yes, they might be worth their price in terms of revenue generated. But they
Full transparency mode (which is scary), here's my website monthly costs. I did audit this list in Jan and removed anything that's not bringing revenue or value.
Drip ESP ~135
Postscript ~85
KNO 16
Shopify (dtc) 80
Shopify (b2b) 40
junip 20
CleanSizeCharts 7
Checkoutlinks 9
I still don't get the
@trytriplewhale
Pixel—or any other pixel that claims to patch the data hole after iOS 14.5.
I would love someone to tell me where I'm wrong and explain what I don't understand.
Here are the 4 things that give me trust issues and makes me skeptical:
18% of all online stores fail to optimize their main menu for sales and conversions.
It's straightforward, logical, and easy to do. And yet, so many fail.
Here's how to drastically improve your conversion rate in 2 minutes:
Scale is overrated.
Last month, we did $100K in new customer revenue at roughly 0 profits. It looks great in the graphs, but is it really worth it?
I’m not sure anymore.
You can make a lot of money selling less than you think if you do it profitably and with low overhead.
Two years ago, I quit my job and plunged headfirst into the unknown world of eCommerce. In September, we hit our first 7 figure revenue month. Here's 9 random thoughts/learnings from our journey:
If you don't have a "Best Sellers" category, create one RIGHT NOW.
In testing, ~23% of subjects went to "Best Sellers" first when landing on a new site.
Why? Because i's social proof voted with purchases instead of reviews— a shortcut to which products will be most promising.
It took me six years of running Meta ads to learn this.
And I think most people misunderstand it, too.
The goal of your Meta ads is not to be efficient with ROAS, CPA, or whatever. The goal is to generate as much contribution margin (C.M.) as possible.
Example:
Hats off,
@Shopify
. That's a great looking store 👀👏
But why aren't you sending merch to Europe? I need that crewneck. And the hoodie. And the socks. And the hat 😭
Here’s a little sneak peek of my database FULL of CRO tips, tricks, and best practices 👀
Would you be interested in buying this if I packaged this in e.g. “How to build a high-converting product page” or some sort of CRO checklist?
Optimized email popups are CRIMINALLY underrated.
We're basically getting ~$3.5 email signups from our prospecting campaigns ON TOP of the revenue they generate.
🤯
I finally managed to get a double-digit converting e-mail pop-up 😎
It's currently at 11.59%, beating the shit out of our old 6.29% pop-up 🔥
And it's even getting MORE views - I found an error in our targeting that I think a lot are making 🤦🏻♂️
Do you want me to write a thread?
Recently, I wrote a thread about how you can optimize your dynamic product ads on FB by defining a google product category field for each of your products.
But did you know that you can actually take this optimization even further?
Here's how:
@SahilBloom
Do what gives you energy.
Context: Whatever gives you energy is an excellent proxy for what makes you happy, ridden from societal pressures.
Don't get caught in the yearly
@klaviyo
trap. Set a reminder to downgrade your plan right before your next renewal date.
Klaviyo conveniently auto-upgrades your plan during the BFCM surge, but innocently "forgets" to downgrade it afterwards.
Hot take on Meta ads:
Forget about landing pages, copywriting, creative, cost controls, and testing methods. Those are just the icing on the cake.
The real magic lies in driving down your unit economics so much that you break even at 1.3x ROAS.
Almost impossible to not win.
It's amazing what you can do with
@googleanalytics
.
One of my favorite things is determining which pages drive the most revenue. I call these "cash cows."
And implicitly, what pages need work. I call these "fixer-uppers."
Here's how to find your cash cows and fixer-uppers:
@Austen
This makes no sense. You hire an FTE for something that takes 1/8 of your time? And the thing you do is critical, and yet cheap and easy to hire for?
My biggest Meta epiphany was when I realized that we don't test different angles to find the best one.
We test different angles to unlock new customer cohorts and increase our total addressable market (TAM).
Meta ads are governed by something called a VCG auction.
When you do the math, using game theory, the optimal way to run ads in a VCG auction is to:
1. Use bid caps.
2. Set the bid to your break-even CPA after COGS, shipping, pick N pack, etc.
THIS IS THE SECRET SAUCE!
Just a friendly pre-BFCM reminder that revenue doesn't mean sh*it.
Who cares if you break BFCM revenue records if you're not making any profits.
Focus on profitability, not revenue growth.
You can't just look at revenue (ROAS) and gross profits when evaluating ads.
Consider the two scenarios:
A: $58k spent, 1.45 ROAS, $70 CPA, 828 Purchases
B: $25k spent, 2.0 ROAS, $50 CPA, 500 Purchases
A has more revenue and more gross profits, but B yields more net profits.
This one is for you,
@mrsharma
.
Here are 13 easy, low-code site optimizations I would do to give your Long Wknd acquisition the best start:
(Going from from the homepage -> collection pages -> product pages -> checkout)
I don't think people realize the importance of margins.
I have a Master’s in economics. I even applied for a Ph.D.
And even I didn't realize how much an increase in margins affects one’s business until we raised our margins from ~54% to 66%.
It makes everything SO much easier.
@EcomMiles
Yes, learn dropshipping so you can steal an image of a Ferrari from a towing company and fake it as a baller on the internet to get people to buy your dropshipping membership🤦♂️
I'm thinking about giving away my entire BFCM21 playbook for FREE.
I planned to sell it. I even made the Gumroad listing (image below).
It includes my thoughts, 8 key insights, and the exact 75 actions (!) I've planned for my brands.
Easy copy, paste, win 💸
Do you want it?
When doing A/B tests, optimizing for conversion rates makes intuitive sense.
But did you know that optimizing for conversion rate can actually make you less money and hurt your business?
Here's why and what to do about it:
Great news from Facebook Ads:
1. The 7-day click attribution window now includes modeled (estimated) conversions.
2. If you reprioritize your events in Events Manager, ads optimizing to an event already prioritized will no longer be paused.
I think 1. makes 7dc > 1dc for ~all.
I love Barry's first-principle thinking here. So much that I've launched a new experiment testing popup vs. no pop-up in all our stores.
Our pop-ups doesn't offer a discount. But it's always good to retest your assumptions. And I love questioning standard practices.
So far, no
@dtcpages
What if big email just programmed us to think that?
People are programmed to sign up for emails for discounts, but if you tell them there aren't discounts, then why does it matter, then maybe they'll just buy right away.
Here is the most controversial A/B test you will see in a long time.
More or less all landing pages (LPs) include buttons in their hero sections.
But in this test, removing the buttons on all our LPs increased revenue per session by 30%.
Here's why I think it makes sense:
Minimizing OPEX and improving your unit economics is a lot more important than anything you do on Meta. Anything.
You play ecom on easy mode if you break even at 1.16 ROAS (unit economics) and can spend 50% of your revenue on ad spend (low opex) and still be profitable.
American DTC brands, please realize how lucky you are to be American.
EU brands need to be in 24 of the 26 EU countries to reach the same market size as the US.
Most of the 24 countries have different languages and currencies, which makes it a pain in the ass.
On top of that,
We just smashed our daily sales record and did 70% of our total 2020 BFCM sales in 2 days.
Here are the 6 key insights that drive my approach to BFCM this year:
We spend 65% of our Meta ad spend on weekends (yes, really).
If you're not spending more on weekends, you're leaving A LOT of money on the table.
Here's why this is the case and how to exploit it:
I have a database full of best practices that maximize conversion rate and revenue for online stores.
Here are 14 website optimizations I believe would massively improve
@hint
's conversion rate and drive more sales:
Transparency time: I'm burned out.
I have 0 motivation to work and feel indifferent about most things.
I've built my company from 0 to 7 figures. And I'm going to grow it to 8 figures, too.
Just not right now. Not it's time to reflect and rekindle the flame.
And that’s okay.
@mkobach
Why is tipping even a thing? How did it start? Isn't it just an excuse for the employer to pay less?
It shouldn't be up to the customer to decide what anyone at a bar or restaurant takes home. Tipping should be a bonus for extraordinary service or experience.
Facebook is tracking ~75% of all our orders. This means that we can estimate the "true" ROAS, purchases, etc., by multiplying all metrics by 1/0.75 = 1.33.
Find your number by comparing no. of tracked purchases in your events manager with your actual Shopify orders.
I see smaller and smaller differences between Facebook's Ads Manager and
@triplewhale
.
If the difference keeps getting smaller, the only advantage of TW is getting the data up to 72 hours faster.
Is this the beginning of the end for 3rd party attribution? 👀
In
@mrsharma
's newsletter, he recommends using Clarity to heatmap your site.
But Nik. If you care about page speed, you should use
@luckyorange
.
It loads four times faster than Clarity and is a much better product.
It's a steal at $20/m for 3 sites and 60K pageviews.
In the last 30 days, ten of our Klaviyo flows generated $40K in revenue. That's 23% MORE than the total revenue of our first year of running our online store.
23% MORE?!?! 🤯
Here are the ten flows ranked by revenue:
We used to struggle a lot with
@PayPal
withholding our money for up to 21 days.
We recently set up Synctrack, a Shopify app, to sync tracking numbers to PayPal when an order gets fulfilled.
I haven't noticed any of our money being withheld since.
(No affiliation to Synctrack.)
@startupsmith
@Movingdeity
@lucasfms
@PierreDeWulf
So what, he shipped something new? It's not all about you. It's Sunday. He can do whatever he wants.
It blows my mind that you call yourself a mentor. It honestly sounds like you've never built anything with even moderate traction when you can treat a founder like this.
I've spent the last two years squeezing every drop out of Meta ads to hit our sales targets.
I should have just cut OPEX, lowered targets, and accepted slower inventory turnover.
I would have earned a lot more money — with way less stress.
@toddkovalsky
@soundslikecanoe
Please don't run awareness campaigns nor listen to the FB marketing team. No disrespect, but they don't know what they're doing.
TIL that you can change the color of Safari's top bar on iOS devices to match your brand colors 👏🏻
Just insert this meta property in your <head> and change the color property to your brand color:
<meta name="theme-color" content="
#EF4444
" />
Almost everyone on DTC X misinterpret statistical significance.
I have a Master's degree in quantitative economics. Here's what it is in layman's terms:
Samples
The first thing to understand is that any time you A/B test something, you do so with a sample, not the full data
With iOS14 making it difficult for Facebook, I predict Google Analytics will be even more important in the future.
I'm thinking about doing a blog series (5-10 parts) on GA. I’m thinking
- Optimal set up with Shopify
- Which analyses to do
- Next steps
- etc.
Should I do it?
Just rolled out 16 new ad tests.
Eight unique angles, all using the same static design, but with different headers (the angle).
Each angle tested in two versions: one directing to a PDP, the other to an LP with a custom hero section about the angle.
If something clicks, I'll
Do you want to work on your store's conversion rate? But you think it's difficult and doesn't know where to start?
Then I have an announcement I'm pretty hyped about.
I've partnered with
@itsOddit
, and together we'll soon release a 100-150 item "CRO Checklist" 👀
The idea:
Am I the only one that doesn't trust companies that don't just state a price on their site?
I'm really curious about
@siena_ai
. But I have to book a demo to see some action, and there's no pricing on the site.
Sooo I'm not booking a demo 🙃
I'm a big believer in cost controls, especially bid caps.
But I must admit: I can't find any explanations for Meta not spending on capped campaigns while the same exact ads are spending fine in non-capped campaigns — and get orders below the cap.
Example: Today, at 2PM, we've
ROAS is a horrible measure to evaluate your paid ads.
If you don't understand this, your ads might be bleeding you right into bankruptcy— as we speak.
I'm not kidding.
Let me show you why with a real-world example from one of my own ad accounts 👇
My new Shopify theme is coming along nicely.
I gave the hero section an iteration today. Love it 😍
The idea is simple. A theme where CRO best practices > flexibility. Basically, a theme where it's impossible to fuck up.
"What do you think about X Shopify theme?"
Don't buy it. It might be pretty, but that doesn't mean that it sells (converts).
Start with Shopify's Dawn theme. It's flexible, so you can build most with it on your own.
Then A/B test your way towards your own, customized site.
Do you want to save $7,552/year?👇🏻
I saved $5,140/yr AND went from 2000 to unlimited reviews/month by switching from
@Yotpo
to
@judgedotme
.
I saved $2,412/yr AND went from 2000 to unlimited tickets/month by switching from
@gorgiasio
to
@reamaze
.
0 problems/compromises so far.
My thoughts on when to use bid caps vs. target average cost, the two cost control options:
I used to be a big bid cap stan, as
@DaveRekuc
calls it, but I have to admit that I'm coming around to target average cost (cost caps).
Target average cost: For when revenue is more
@businessbarista
1. Go waaay broader. No reason to only target tailgaters. FB loves broad audiences.
2. Try bid capping it at, e.g., $4 or wherever your threshold is. That way, you can increase your budgets and get more upside on the "good FB days." If you can't get leads < $4, FB doesn't spend.
Your goal isn’t to find a single, best-performing ad.
Your goal is to create ads that sell your product to as many different types of people as possible.
Some buy your product to fight acne. Others appreciate how it softens their skin. For some, it’s a status symbol.
Some
CRO tip: Don't use sticky "Add to Cart" buttons.
It might sound weird, but visitors tend to overlook sticky ATC because of banner blindness.
In tests, a sticky ATC doesn't outperform a normal, attention-grabbing ATC button. So use the viewport for something else.
Meet Beep, my new Shopify app!
The last few years I've been focusing a lot on reducing the time spent on my phone. I don't have any SoMe apps installed, I only get the absolutely necessary notifications, and I use a VPN to block all sites I might visit to kill time between
Sneak peek of my upcoming CRO checklist 👀
The idea is a +100 item checklist you can go through and implement one by one.
When you’re done, you have a top-notch high-converting online store 💸
I'm less and less bullish on A/B tests and more and more on just implementing whatever
@itsOddit
or someone with a lot of tests under his/her belt says.
I love A/B tests in theory, but it simply takes too long to get results for most brands.
@dqmonn
Our first German ad account got banned because our customers reported it as a scam.
Why? Because our site lacked an "Impressum."
So we made this Impressum. It's literally just a page with company information that anyone could fake. Zero scam reports since.
Now, 10 months
Update: I just had a video call with
@AY_Orbach
and
@rabahrahil
from
@trytriplewhale
.
I'll do a write-up whenever possible. But tremendous respect for taking the call and being open to talking about my skepticism instead of just diminishing it. They're great guys, no doubt!