- I’m thinking about launching a course to help entrepreneurs and aspiring entrepreneurs! If you’re interested, please answer some questions in this quick survey.
- I’ve decided to change my email cadence to once a week every Sunday morning :)
Let’s get started:
Entrepreneurs talk about marketing endlessly but I am consistently shocked at how poorly they understand it.
I don’t consider myself a marketer. Prior to starting my last company Ampush, I worked at Goldman Sachs. Say what you will about Wall Street, but it does teach you to be insanely rigorous with very high standards.
As I was deciding what to build towards the end of my time at Goldman, I got attracted to digital marketing because it reminded me of a crude Wall Street. It was all numbers (or so it seemed) and you were trying to find "undervalued" things and buy them. In marketing those undervalued things might be keywords, audiences, etc.
Bootstrapping and Paid Media don't typically go together. But in my story they do. We used credit cards but got SUPER rigorous to actually bring our paid media to profitability.
Here are the 3 concepts I most often teach founders to take their marketing rigor up a level:
1) Funnel Stoichiometry
I hated chemistry in HS. My teacher was a super weird old guy who I buttered up to get an A. But he did teach me one critical concept: stoichiometry. For those who don't remember: this is where you use a bunch of fractions to get from a quantity in one unit to a quantity in another. I didn’t realize it until much later, but stoichiometry can be applied to media buying.
People talk CPA and ROAS constantly, but I'm always shocked when I ask: “well what's your CPV (cost per visit)?” And no one knows the answer. The math looks something like (check out the “Newbie Glossary” at the bottom of this email if you don’t know what some of these things mean in layman’s terms):
I won’t bore you with the simplification, but if you substitute Impressions in for Visits and Visits into the CPV equation, you’ll end up with:
While we’re at it (because I know how much you love algebra), let’s also show how CPA, Revenue, and ROAS are calculated:
Once you understand this math, you can start learning a TON about your campaigns (especially as you scale). For example Revenue/Impressions (how much revenue do you generate per impression you serve?). Do all channels have the same CPV? If not, why? Are the channels "right priced?" (meaning you're paying the right CPV relative to your CPA).
This is an easy optimization btw, low conversion rate DOES NOT equal bad. It just means you should pay less for the traffic. Rapidly navigating across these equations becomes a super power for a founder and/or growth lead.
One of the metrics we used a ton at Ampush was “APM” -> “Acquisitions per 10k impressions.”
It’s essentially a combination of CTR and CvR. It shows how much “yield” you generate per 10k impressions served. Yield meaning, how many new conversions/customers are you getting.
Why is this important?
FB charges on a CPM basis. So to be competitive in their auctions, you have to be willing to pay a higher CPM than anyone else. In order to do this, you need a great APM (yield). It’s like the FB impressions are plots of farmland. If you have a better tractor, you will get more out of the land than anyone else so you can pay more for it.
FB also adjusts CPM based on how "relevant" your ad is and importantly, how competitive the specific audience is. Therefore, a better APM probably means your engagement is also good (since CTR is part of the equation) and thus more relevant, leading to a lower CPM. Once you have a great APM, you can compete in more competitive auctions without sacrificing CPA.
By measuring APM, you can both improve it consistently (which lets you bid up CPMs and get more volume) and triage accounts effectively. If APM is down across the account, that probably means (1) there’s a FB issue, (2) your site is broken, or (3) there is a tracking issue. If APM is only down for some of your segments maybe your message isn’t resonating.
The main purpose of APM is to track overall improvement and more so to spot top-line trends. If you notice APM is down across the board, something is definitely wrong broadly but you still need to dig into the funnel to find out what it is. If APM is down in specific segments, everything should still be running fine but then you need to dig into the funnel again to figure out what you need to work on.
My CPA is $50 and I spend $50,000 a month. So I get 1,000 acquisitions per month. This is true ON AVERAGE. But try to remember your old stats class: there is ALWAYS a distribution.
What's your first decile CPA? What about your 10th decile? For a business with a $50 CPA, I'd often see the first decile be like $10 and 10th be $120+. A simple trick: shut off your worst spenders - your average CPA will automatically go down. Surely you can spend that money elsewhere or not at all.
Track and monitor these deciles over time. A great marketing program will make the distribution tighter (say from $40-60 vs $10-100) while making the average come down.
Why does shutting off your worst spenders make your average CPA go down?
- You learn your customer better so your tests win more often
- Your audiences get more dialed in
- And of course, you monitor/measure deciles so you avoid large distributions
Below is an example of a decile analysis. As you can see, this company spent $104,737 across all deciles of their marketing campaigns, and pretty evenly at around $10k per decile. However, the CPA for the top 2 deciles ($82.46 and $122.01) are outliers. If we cut those campaigns, our total media spend comes down by $20%, we only lose out on 5% of our total acquisitions, and our cumulative CPA goes down by 15%! The company can use those savings towards higher ROI activities!
Quick trick: the magic of “de-averaging”:
If you’re running search ads, try this. Instead of bidding for every search keyword the same, what happens if you bid down lower converting keywords and bid UP higher converting keywords? Your spend likely stays the same, but your CPA goes down and your deciles shift. MAGIC.
3) "Delta Value of Conversion Rate"
OK we are going next-level here but bear with me :)
It feels almost dumb to say this but lots of people don't realize: improving conversion rate is almost ALWAYS your best move in customer acquisition. Let's do some simple math. I get $100 in Revenue per Acquisition. My CPV is $0.80 and my CvR is 1%. So my CPA is CPV / CvR = $0.80 / 1% = $80. At 1000 customers, my revenue is $100 * 1,000 = $100k and my media spend is CPA * # transactions = $80 * 1,000 = $80k, so I make $100k rev - $80k media spend = $20k in profit.
Quiz: I want to maximize Revenue & Profit
Let’s say we start with the previous example: CPV = $0.80, CvR = 1%, $100 AOV
- Same economics, I spend $100k, so I get $125k Revenue - $100k media spend = $25K in profit. Not bad.
- Now, I spend $60k (75% of $80k), so I get $100k Revenue - $60k media spend = $40k in profit. SWEET. (Although my revenue didn't grow…)
- Here, 25% MORE visitors convert to customers for no more media spend. 1,250 customers * $100 AOV = $125k Revenue, media spend = $0.80 CPV / 1.25% CvR * 1,250 customers = $80k media spend, so profit = $45k. DING DING DING. Not only did I make more $, I also grew my business.
Now let’s take it to the next level: in bond trading, they have this concept called "delta value"... I won't go into the hairy details but when interest rates go up (quantified in basis points - 1 basis point = 1/100th of a percent), bonds become less valuable by some dollar amount, called the delta value. For example, the delta value of a bond may be $20 for every 1 basis point change in interest rates.
I apply this to conversion rates. For every 1% improvement in conversion rate (meaning 4% to 4.04%), I quantify how much profit that generates for my business. Once your team understands that number, they'll run the business very differently.
Creative is CRITICAL
Whew, as I wrote this my head filled with a million other quant concepts that are critical for excellence in digital marketing. I will share those in the future here or even put together a course. If you're interested, join our waitlist!
In closing, let me be clear: math is necessary but NOT sufficient. CREATIVE is still critical…especially for paid social. Great hooks, content and CTAs make the curves shift. But in my experience, without a strong quantitative underpinning and rigor with data, you will eventually hit a ceiling of scale and be in a world of trouble.
That’s all I’ve got!
BG team, one last request: Please share this email with 2 people who will get value! I'm building up the community for some amazing things in '24…
- CPA (cost per acquisition): how much it costs you to acquire a customer
- Impression: someone who saw your ad
- APM (acquisitions per 10k impressions): how many customers you get for every 10k ad impressions
- CPM (cost per mille): AKA cost per thousand impressions for paid advertising
- CPV (cost per visit): how much it costs you to turn someone who saw your ad into someone who landed on your website
- Conversion: someone who bought something from your store
- CvR (conversion rate): the percent of people who landed on your website that turned into customers
- ROAS (return on ad spend): how efficiently your ad spend turns into revenue $$