The Outlook
With performance marketing, also known as paid marketing, we're talking about doing some type of paid media with the end goal of driving someone to transact.
One of the things that's really great about this is that over the last decade or so, this has gone from being out of home and TV to now being very, very measurable through an awful lot of platforms. The main ones that we see throughout our user base in terms of channels, where you might be doing performance marketing are across social and SEM (search engine marketing).
What is the customer experience when they move from an ad directly clicking through to whatever digital product or branded commerce that we're sending them to? How do we get in front of people? Or, how do we stay top of mind for people that we’ve never had any touch points with? The answer? The right metrics.
A few metrics come to mind pretty immediately. I always like to think about all of these metrics first on a blended perspective. How do we take everything that we're doing from a paid media performance marketing performance to multiple different platforms like Instagram, Snapchat, Pinterest, Twitter, Facebook, Google, even getting ads in Gmail? You're going to be seeing multiple different ads.
That's why I think blended is really the way to go in terms of the top level things that we should be thinking about when it comes to performance marketing, because by fully blending things, we just look at the raw data and see what that tells us about the efficiency of what we're doing from a marketing perspective holistically.
Now for the metrics. Let’s start with the budget of everything that we're doing, such as total paid spend, which we can get by adding spend from across each channel. We might also want to think about it in terms of the amount of discounts that we're giving. We have some customers that we work with where it's really helpful to understand how much cost of discount goes into it as well, especially if you're lower margin. It's also important to look at a fully blended ROAS (return on ad spend). A lot of people will look at a blended ROAS broken out by channel, which I think is okay to try to understand single channel efficiency. However, I would highly recommend that people start with a fully blended ROAS.
I personally think the best way to look at that is to first take net sales, which should be your gross merchandising value minus discounts. If you factor in shipping revenue, you can include shipping revenue. I would exclude returns over time, if possible, and exclude taxes since that's venture revenue. By looking by those net sales, whether that's single channel (like Shopify) or multi-channel (like total net sales on Shopify plus total net sales on Amazon), and then if we divide that by our total spend (if we're decreasing discounts, we don't need to include discounts there and that denominator, but we might want to include influencer spend there and all of our channel spends there), that's going to give us what I think about as a true blended ROAS. It's going to give us a really good insight into what the overall efficiency of our marketing is.
Generally when we work with customers, we'll view marketing efficiency either as a percentage or a multiplier. If you have access to your cost of goods sold, you can bring that in and even create what we think about as a PROAS, which is a profit return on ad spend. This tells you directly how many dollars you're putting into your bottom line for every $1 that you're spending on paid media. Starting with a blended ROAS where you're understanding the top line, which is how much revenue is being generated for every dollar spent, is going to be an incredibly helpful place to start.
For instance, if you're a 50% margin business, we can think about if we've got a 10x blended ROAS from across all media channels, influencers, etc, then we're generating roughly $5 for every $1 we spend, which is really nice.
Building out those metrics and understanding the ranges based on the type of business you have, as well as your need to be profitable on things like first order, is a helpful place to start in understanding overall performance marketing efficiency.
If you're in budget and you're within some type of range of blended return on ad spend, almost nothing else matters. But, those tend to be fairly lagging indicators in terms of metrics. Usually, you have to wait at least a week or so to post something new, to see how it's impacting, and make sure we have a statistically significant sample size to make any really good decisions.
The next thing looking holistically as a business is drilling into the order economics of what we're doing. For me, that all starts with our blended cost per order. A traditionally big metric that people look at in DTC (direct-to-consumer) and CPG (consumer packaged goods) is blended CAC (customer acquisition cost). I think that metric is starting to matter less and less, unless you are a subscription business, because whether you're trying to raise capital or if you're trying to be a profitable business, profitability matters significantly more than an LTV (lifetime value) or CAC ratio.
Touching on order economics: let’s discuss blended cost per order. Similar to blended ROAS, in the numerator take all of your spend across all channels the same way that you would have done for that past metric that we talked about, being inclusive of things like discounts, in this case, and then having your orders in the bottom. I think it's great if you can filter out things like returns or orders with a non-zero value. Looking at what it actually costs on a performance marketing, paid media basis in order to generate a single transaction is going to paint a really good picture of marketing efficiency.
Where we see that being helpful inside Yaguara’s platform is layering that cost on top of metrics that are good indicators of the quality of customers we're bringing in. If we're transacting someone at a $10 blended cost per order, but our average order value is $20 and we have a 50% margin that we're only breaking even on in every order, that's where it's great to think about the additional context that we can layer in there on top of what's coming from our paid marketing channels.
Two things that I love to think about: First, average order value (AOV). If we're a 50% margin business and we might want to generate $10 of profit per order, that means we want to look at AOV of at least $30 or $40, and probably even more than that. I think those things can start to give us a good idea of creating these ranges, especially in our platform, putting them next to each other and then understanding trends.
Say we're in the middle of week three of Q4, and our blended cost per order is significantly lower than we thought it might be. Say our range was $10 to $15, and our AOV is still in that $40 to $50 range. Maybe we should be spending a little bit more this week because we seem to be spending it really efficiently for one reason or another.
Another thing I would think about, which is dependent on the type of company you have, is units per transaction, which can be another great leading indicator to the quality of that traffic. If you're a lower price point, that conversion rate can be another one. If you use something like average order value, units per transaction, and conversion rate, and we coupled that with blended cost per order, then we're able to see not only transaction in an efficient manner in terms of costs, but are we also transact customers who have the potential, who are currently high quality order transactions and have a chance to be a high quality long-term customer.
We talked about context in our last conversation. I think particularly if you have high value products, say a $5,000 sofa, or a $1000 leather jacket or something of that nature, we can look at what actions, including the cost per action, that might show intent. I like to understand what it costs for us to drive high value or high intent customers to their site.
The first way of doing that is just to blend everything and ask, ‘Hey, what are we spending similar to blended cost per order, blended cost per visit? What does it cost for us to generate a single visitor?’ We can try to do one-to-one and just do total costs plus clicks from all of our different channels. I like totally blended because I think a lot of times, if I get marketed to on Instagram, I might love your product, and I might tell my wife about it, and then she might visit– so, I think it's really helpful to understand what that blended cost per visit is. Even for folks who probably aren't going to transact on an initial visit, I still think it's really helpful to understand what that blended cost per visit is.
Secondarily, it's really helpful to understand that this is taking data from Google Analytics or Segment, and it’s mixing in with our performance marketing data, in turn understanding our blended cost for a PDP (product detail page) view. We would generally say, particularly if you're sending someone to a homepage, a landing page from an ad would be some high level of intent. For a user, they would come and actually view the product versus bouncing. That’s the next step in understanding the quality of the traffic that we're driving.
With the third piece, similar to the cost of add to cart rate (particularly for products of higher value with a longer consideration time), one thing that a lot of customers might do is use the basket a bit like a wishlist. Maybe when they're going to look for external reviews, ask a friend who might have it, etc, in order to validate making a larger purchase. If we can see the cost for someone to add to cart, that can be another really great way for us to understand if we're having high value customers.
The last piece I would think about, and this can be true across anyone, particularly if you're someone who does retention marketing really well, is what's our cost to add to, say, a newsletter, or just to opt into marketing. That can be a great conversion as well for us to think about.
Those are a lot of different metrics. If we're a more traditional, direct-to-consumer business where we might have an opportunity to convert someone at a first visit because we have a lower consideration period, then we can look at a lot of these great leading indicators, like blended cost per order with AOV, with something like units per transaction and conversion rate as secondary leading indicators.
On the other side, if we have that higher product value, we still look at those metrics that we talked about before, but with the understanding that those are going to be lagging indicators, and then layer those in with some good leading indicators of cost per visit, cost per PDP view, and then cost per add to cart rate. If we layer all that data on top of each other, that should give us a really good understanding of the efficiency of our marketing, as well as our ability to transact profitably hopefully on every single order.
I think the first question is, what are you trying to achieve with performance marketing?
Are we thinking about acquiring new customers? Thinking about reactivating existing customers is always a good question for us to think about. What’s the ultimate goal? If we're just looking at driving direct transactions and having a really high blended ROAS, then that can be a really important North Star for us. If you’re thinking about brand building, you might want to think about things a little bit differently: consider a blended CPM (cost per thousand impressions) and thinking more about cost per visit or cost per impression across all platforms, not just impressions or direct visits to your site.
Then, who is your end customer? I think that can change how you look at scaling performance marketing in a really meaningful way. Also, how do you choose? There's a million different channels, and new channels are popping up all the time. Who are you marketing to and where should you be talking to them? If you listen to the podcast we did with Savannah Sanchez, she talks about this really poignantly: if you're marketing to young millennials or Gen Z rather than marketing on Facebook, you would probably want to look solely at Instagram in terms of where you would utilize an ad manager. Then, something like TikTok, particularly if you have a product that's a lower product value where people might transact very, very quickly.
That's where I think blended metrics can be super helpful, because let's say we're selling some cool streetwear that the user's wearing these days. We're trying to market to 18 to 25 year olds, and we've only been doing performance marketing on Instagram. If we add something like Snapchat, for instance, blended is super helpful. It doesn't just tell us if people are transacting directly through Snapchat, because maybe they're not. Maybe they just see the ad and they move forward because of the way they're consuming content and Snapchat. But, if we fully blend that using something like our platform, we can annotate something like, ‘Hey, this is a day we launched Snapchat. This is when we started spending on a new channel.’ With that, we can see if it impacts the overall efficiency of what we're doing from a performance marketing perspective. That's where I think blending all of that data not only removes false data, in terms of multiple attributions being counted and needing to dedupe transactions, but it also gives us real insight into how a new platform might be impacting marketing efficiency, regardless of the single channel efficiency.
New people are coming onto new platforms all the time, whether that's people aging into using something like Twitter, or people aging out of using something like Facebook. I think you need to be constantly testing and reviewing impacts, whether it's different creative, different channels, or different devices. And, I think constantly iterating and listening to what your customers have to say and what data has to say is incredibly important.
If you do those things really well and you focus on the metrics that are tied back to the objective of the business, then anyone can really do performance marketing.
