<img height="1" width="1" style="display:none;" alt="" src="https://dc.ads.linkedin.com/collect/?pid=193356&amp;fmt=gif">
Demo
Login
Menu
Demo
Login

Beyond Attribution: eCommerce Data & Clawback Analysis

by Jordan Ehrlich, on July 23, 2019

Episode 001

REAL GOOD MARKETING

In this week's episode of Real Good Marketing, Brennan and Jordan discuss combining eCommerce data with first-party marketing insights to confidently allocate marketing budget for growth.

(View more episodes here)

Want to build a better marketing strategy with clawback analysis? Reach out to see your customer journey holistically in one platform.

Transcript:

Brennan:
It's not necessarily saying last click attribution's wrong, but only looking at that one number is a myopic way to view things, because there is a bunch of other variables that you have to take into consideration. If I acquire somebody and it has a six ROAS and they don't do anything in the future, that's good, but I would personally rather have a brand advocate that I get from maybe a three ROAS that does more moving forward.

Jordan:
And we're back for another episode of Real Good Marketing.

Jordan:
So this story told, looking back on what you did for this customer, what did you learn?

Brennan:
What I learned through this analysis is the importance that combining your ecomm data with your marketing analytics data is, and to provide a true version of your acquisition tactics. It's important to take consumer behavior, the post-purchase consumer behavior into consideration of your acquisition tactics.

Jordan:
What was the moment helping a customer where you realized you needed a little bit more information to optimize their marketing budgets?

Brennan:
Yeah, so the customer actually brought up this specific idea, but realizing the implications that it has across the other data points we are looking at and the things that's going to impact, was pretty eyeopening.

So with this specific customer, they wanted to grow but weren't seeing good return whenever they expanded out of a certain realm of channels. They were growing year over year which is great, but what we noticed within the data was that what was driving majority of their growth was returning revenue. Right? So people that were, just kept coming back and buying and buying and buying, which is great, right? It's good to have that retention machine running, but if you're not supplementing that you're running the risk of attrition, which I was alluding to earlier.

Jordan:
And what would you guess the percentage of returning versus new customers would be?

Brennan:
Let's say at the time we were looking at it, it was probably 80% returning revenue versus 20% new, give or take. Which is pretty steep, right? And so what we wanted to do was figure out the right places we could place our bets. By places I mean channels, that we could place our budget to increase the mix of new revenue that we were bringing in while still obviously keeping our returning base healthy.

And so we incorporated our first party JavaScript data that we have on the customer's site. Combined that with the specific Shopify data, to understand the true revenue and true acquisition data from Shopify, which is really the central source of e-commerce truth, if you will. Pair that with the unique identifiers that we have in our system, and we were able to run that, the segmentation I was alluding to and so it really helped them feel comfortable in the places that they were spending their dollars, knowing that we would make the dollar back eventually. And that's resulted for them in roughly a 20%,25% year over year revenue growth.

Jordan:
Oh, wow. As it directly related to allocating their budget to those cohorts?

Brennan:
I mean obviously there's a bunch of variables, but us thinking about this differently with them and helping them plan out their demand plan, helping them be comfortable with the dollars where they're spending and the strategy behind that, their business as a whole is up pretty significantly.

Jordan:
Tell me about clawback analysis. Explain that to me.

Brennan:
So the idea around clawback analysis is effectively that you want to segment your users to better understand how they contribute revenue, from when they are initially acquired.

So an example of this would be say you get somebody in January, you spend a dollar and they spend a dollar. So that's a one ROAS right. But you have to understand that, that individual is going to do more moving forward into the future. At least you hope, right? So clawback analysis helps you better understand the revenue that's contributed by an individual group or cohort, if you will, moving forward. So the way that we look at it today is we look at a specific month in which a user was acquired and then segment them by a specific channel to help understand when we acquire somebody via paid search, let's say. They may contribute a dollar in the first month, but maybe in the following three months they contribute five more dollars.

When you look at it that way, it helps you better understand the actual revenue that that group of people is contributing, to then let you be a little bit more loose with your return on ad spend targets moving forward, knowing that you'll make that money back in the future.

Jordan:
This sort of analysis would be kind of like supplemental information to an attribution.

Brennan:
Yeah. Yeah.

You'd look at one attribution model, whatever you're optimizing for and layer in clawback analysis.

Attribution helps you understand what has happened in the way that your revenue could be attributed out to specific channels. Whereas this is more so impacting your media mix modeling. And attribution does feed into media mix modeling, which is understanding where you're placing your bets effectively. So this more so impacts that, because if you know that you can clawback 30% more revenue, you're going to adjust your budget accordingly.

What a lot of marketers do, rightfully so, is focus on a last click attribution model. So they attribute all revenue to when the final marketing touchpoint that drove that conversion. And when you do that, and if you were to only optimize towards a ROAS target, a majority of the time, a majority of revenue is going to stem from marketing tactics or marketing channels in which your consumer base already knows about you, right? So branded search, remarketing, email, and it impacts your ability to acquire net new, right? So get more users to your site.

And so when you're only focusing on the best returners from a last click perspective, you run the risk of hitting attrition, which is effectively causing all your returning users to get tired of you and your brand. Because they're always coming back and buying and you're reliant on your growth. On the folks that already know about you, and relying on that and not back filling that with new consumers. And getting them to become those returning visitors that are buying five, six, seven, eight times. It's a risky game to play.

It's not necessarily saying last click attribution's wrong, but only looking at that one number is a myopic way to view things because there is a bunch of other variables that you have to take into consideration. If I acquire somebody and it has a six ROAS and they don't do anything in the future, that's good. But I would personally rather have a brand advocate that I get from maybe a three ROAS that does more moving forward. So clawback analysis helps us understand the channels in which we can place our bets to understand the percentage of ROAS that we could, or the percentage in which we can lessen our ROAS, is the right way to say it.

Say that when we acquire a customer from paid search in January. They spend that dollar, right? And I think I said in the following three months, they may drive 50 cents in revenue. And what that tells us is that they contribute 50% more revenue than their last click target. Therefore, in theory, you could drop your ROAS target by 50%, because you know you'll make that dollar back moving forward. Thus allowing you to open the different channels in which you're engaging in, and to try and capture those net new consumers.

And I wouldn't necessarily recommend do it to the exact percentage in which the revenues increased, but it helps you understand to the length at which you could, right? So if you're increasing revenue by 30% in the following three months, maybe you're going to drop your ROAS target by 10%, 15%, knowing that you'll make a good chunk of that back into the future.

Jordan:
When you lower your ROAS target, what does that look like as far as budget allocation goes?

Brennan:
Yeah. So it depends on the channel, but a lot of the time it could either be increasing the daily budget on a paid search campaign that could be limited by budget or it could be allocating your budget to a more prospecting focused type channel, like a paid social or non branded paid search or display.

Brennan:
It helps you figure out where you can place your dollars in terms of the funnel, if you will. The awareness side versus consideration versus conversion.

Jordan:
Okay. So then after you came to understand these customers and the value that they would bring after that first acquisition, what did that mean for the customer? What did you tell the customer to do after?

And by customer, I mean demand Gen customer that is using our platform to do this analysis.

Brennan:
So the specific example was, we recommended that we do some spot buys through affiliates with the big Ebates folks and things like that, to acquire new customers in a slow part of the year. Knowing that rates would be lower, knowing that it's pretty slow for everybody. But in the following months, people that we acquire from affiliates start contributing much more revenue. it's a strategic bet to make to say, "Hey, I know it's a slower part of the year, but we want to acquire the net new and then allow our retention machine to convert them a bunch in the next following months."

Jordan:
Were there any other channels besides affiliate that you ended up saying you should invest in this right now.

Brennan:
Yeah, paid search. Paid search was one where just knowing that their current strategy, like I said earlier, is very re-marketing focused, brand focused, with barely any spend going to non-brand. So I was saying, "Hey, knowing that you're doing re-marketing, allocate some more budget to non-branded stuff. So just non-branded text ads. Things that aren't brand focused. And then that will inherently improve your re-marketing efforts because you're doing a nice job with that already. Just increase the pool size," Is effectively all we were saying.

Jordan:
Tell me what inhibited this company from truly understanding these customers beforehand. What analytics were they working with and what was sort of that barrier that they just couldn't quite get past to say, "Oh yeah, we know this audience does this, let's invest in this channel because of the value later that they'll bring to us."

Brennan:
So I would say for what they were doing in the past, they were relying on the last click return on ad spend dollars. Like that's how they were allocating their budget. And so they weren't really ever doing this analysis in particular. It's really hard to do it. Shopify isn't inherently a marketing tool, right? It's an e-commerce platform. So they have some marketing data, but that's not their bread and butter.

Brennan:
What we're able to do and then, like Google analytics is more of a marketing analytics platform. Does it have econ data? Yeah. But it's not an e-commerce platform. Right? Being able to combine the central source of the truth for e-commerce data, i.e. Shopify, and have marketing analytics data from our first party JavaScript. Being able to combine that information gives us a really concise, clean version of our customer.

Jordan:
So, that's kind of what I love to get at is bringing in each of these episodes a new combination of data. And in this case, the combination of data was Shopify e-commerce data and first-party marketing analytics.

Comments

DemandJump Blog

From display & search marketing wisdom to general growth tips, here are a few resources to help your digital marketing teams work smarter, not harder.

Subscribe to Updates

Customer Experience Update