When it comes to marketing your product, there’s nothing more important than understanding revenue attribution. So, first of all, what is revenue attribution? It’s the process of matching the advertising your company puts out with the sales you receive. In other words: how much of the money you spend actually converts to sales. This is important because you could be spending thousands (or hundreds of thousands) of dollars on ads and they’re actually bringing in zero sales. Attribution is both an art and science, and it requires a holistic approach to evaluating the buyer journey in combination with data to really feel confident in how your sales funnel is converting. Here is a really good introduction to marketing attribution - I will take a ‘one step back’ look at overall revenue attribution and how it can help inform your capital allocation/marketing mix/budgeting to provide a fresh perspective and not just regurgitate everything Alex said in her piece.
In our post iOS14.5 world, the waters have been muddied. Gone are the days of opening your Facebook Ads manager and understanding exactly what is going on. Facebook has gotten better over time and tools like Triple Whale help, but we live in a new world where creative and analytical marketers will win the race.
The above tweet is my opinion on how most digital advertising platforms fit into the funnel, specifically with regards to my experience with Mad Rabbit. This graphic lends itself to the idea of demand generation vs demand capture.
The top funnel activities can be considered in two buckets: paid and unpaid. Unpaid would be things like viral TikToks and press coverage, whereas paid advertising would include deals with influencers or paid media. These exercises generate tons of impressions, but do not necessarily cause an immediate jump in sales. They serve as one of the 8-ish touchpoints a consumer needs to ultimately convert on a sale. Being good at organic social media is so critical because if you can generate any of those eight touchpoints for free, you get the conversion with a significantly lower customer acquisition cost. So, in short, the top (and mid) part of the funnel are the beginning of the customer journey and serve as some of those early touch points in order to generate demand.
The bottom of the funnel is what I want to spend the beef of this article on: demand capture. A lot of the time, ecommerce operators are considering Google as their #1 way to capture demand for their products by bidding on either branded or non-branded keywords in order to show up first when a consumer is searching for their product. The biggest thing that gets lost on the ecosystem, especially investors, is that Amazon is the world’s second largest search engine - and it's actually the primary search engine for consumer product searches on the internet, ahead of Google.
So when you consider that Amazon’s gross merchandise value is marching towards $1 trillion and that it is the #1 search engine for consumer goods, in reality, Amazon is in some ways the #1 way to capture demand for your products.
Investors and operators get very fixated on the ideas of LTV:CAC (Lifetime Value: Customer Acquisition Cost) and channel profitability. The problem with these two concepts is that they remove Amazon from the equation. When we just established that Amazon is the #1 place to search for products to buy online, it’s asinine to refuse to attribute Amazon revenue to some of your non-Amazon marketing initiatives.
This also disproves the ideas of ‘channel profitability’ and “LTV:CAC”, at least to an extent. Today, it is still impossible to know for certain how many of your customers cross over from Shopify to Amazon, rendering LTV:CAC somewhat useless and completely obliterating the notion of channel profitability. If you were a merchant who spent half a million dollars on Facebook and TikTok and had all your products listed on Amazon, it's likely that a huge portion of your demand would convert on Amazon, even if you weren’t running ads there. If you wanted to create a P&L (Profit & Loss analysis) for Amazon, you would be showing zero advertising expense and would probably be pulling 40%+ EBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) margins. That’s just not realistic.
An exercise I have done and continue to do in order to understand cross-channel attribution is a very simple linear regression analysis to correlate channel ad spends and channel sales. Sounds daunting, but it’s really simple, I promise. All you need to do is export every day of ad spend for each channel and every day of revenue for wherever you sell, mostly Shopify and Amazon for purely online businesses. Then, follow the instructions in this video.
We only need to pay attention to one output once you complete those steps; r squared. R squared in a regression model is an indication of the level of variance in a dependent variable that can be attributed to the independent variable. In simple terms, r squared tells you how much two things are correlated. In this case, ad spend and channel sales.
Without getting too deep into it, when I do this for Mad Rabbit, R squared for Amazon sales and Facebook ad spend is a whopping 79% compared to just 54% for the actual in platform Amazon ads. So what this tells me is that my Facebook spend is likely responsible for more of the Amazon sales than the Amazon ads are.
This is a very simple exercise you can do to just get a quick gut check on correlations. It won’t replace the actual act of assessing multiple sources of attribution or using something like Triple Whale’s Total Attribution Model, especially when you are more in the weeds and in platform trying to decide where your dollars are best spent.
In closing, I encourage operators to be willing to attribute their Amazon revenue in the same way they do with their Shopify revenue, and to think about Amazon more like Google than like Shopify in terms of the end of the buyer journey. It is important to be open minded when attributing your platform revenue, especially post iOS14.5. Facebook provided us with almost a decade of unprecedented visibility into our marketing efforts, however the landscape today requires the same skills marketers developed in the pre-internet days. We have much more data to work with today, but we must take a step back in our approach to building funnels, much in the way pre-internet marketers had to build a holistic funnel using traditional paid media like billboards, commercials, and newspaper ads, without knowing exactly what was happening at each touchpoint. Use Triple Whale’s Total Attribution Model to get a little insight into where your customers are coming from.
Supercharge your growth with a purpose-built ecomOS for brands and agencies.