Rozee Digital Case Studies

Troy Lee Designs


10x Facebook ROAS

6x Google Shopping ROAS

25% Revenue Increase YoY


The company is a relatively well known brand to those within the motocross and mountain biking space. As the space was getting more competitive & organic reach was in decline, the brand wanted to expand it’s direct website sales while also boosting brand recognition to increase retail sales.


Competition: Not only was there competition with other brands, some much larger, but we were competing with their own retail partners for the customer relationship.

Budget: We were given a modest budget for 2019 to prove out a paid media investment with large goals to hit.


With Rozee Digital’s experience working with active apparel / lifestyle brands they hired us to plan and execute their first paid media plan to run through 2019.

They are known for their images and videos, we had an incredible creative department at our disposal to create eye catching ads, engaging videos & build on the brands particular voice and brand messaging.

We put a plan together with the following tactics:

  • Branded search through Google & Bing
  • Google & Bing shopping to expose new customers to the brand that were actively looking for helmets, but were still researching brands
  • Facebook & Instagram ads with a funnel from cold traffic to retargeting driving sales directly through TroyLeeDesigns.com


Because they had an internal creative department, Troy Lee provided all social ad assets.

Rozee Digital built a custom shopping feed to use for Facebook catalog, Google Shopping & Bing Shopping.

Executed media buying across the 3 platforms including audience development, product, keyword testing & full reporting.

Remarketing Funnel

Usually, the first step when a new client comes to us is revamping their remarketing funnel.

Often, it’s the most overlooked, low-hanging fruit within Facebook Ads and Google Ads.

What a lot of people do is they run big audiences, like 60 or 90-day website visitors, set-up a few ads to and that’s it.

But they forget that there is a real human behind that screen. They often get so pissed at their brand from seeing the same two ads over-and-over again for 60 or 90 days, that they will not buy from that brand even if they would need the product!

Most likely, you know how it feels. We all have been victims of this.

Therefore, we employ a granulated and micro-segmented approach in remarketing. It depends on case-to-case, but often after optimisations, we’re seeing about a 20-30% boost in remarketing ROAS compared to the “90-day approach”.

It’s the same structure that we’ve applied for this client.

We split people by intent:

  • Warm Upper Audiences – Content Engagers
  • Warm Bottom Audiences – Website Visitors
  • Hot Audiences – Add to Cart / Initiate Checkout Based on Event’s Triggers
  • Past Customers

And then segment them further by days:

  • 1-3 days
  • 4-7days
  • 8-14 days
  • 15-30 days

This allowed us to swap ads depending on the phase a customer is at and day range.

The best part is that it allowed us to create a specific experience website visitors will be taken through to the minuscule details.

What we usually do in our remarketing funnel is:

1-3 days – We remind customers about the product, introduce different USPs, introduce the brand, it’s mission, drop in a few review ads.

4-7 days – We start building up the “craze.” We run a lot of review content, influencer videos, 3rd party articles about the brand to make it seem like it’s “everywhere.”

We also introduce low stock alerts through Dynamic Product Ads carousels so they would seem genuine.

8-14 days – We start stacking small discounts, offering some other ways to buy the product, for example, a mega-bundle.

15-30 days – We increase the discounts as a last try to get the customer to convert.

New Prospecting Campaigns Structure Capitalised on Facebook Automation to Decrease CPA’s

Another thing that was a significant deciding factor in this client’s success was the utilisation of the right account structures for prospecting audiences.

We can’t stress enough importance to this.

In 2018, and 2019 it was usual to run Facebook Ads in a very controlled and micromanaged manner.

But things have changed now.

Artificial Intelligence is making huge advancement leaps, and Facebooks A.I. engine is quickly becoming one of the smartest A.I. engines out there.

It knows even such things as where and when you’re most likely to buy so that it could show you the right ad at the right minute. It knows whatever you’re buying a product at 7 am while eating breakfast, or at 10 pm when you finally put the kids to sleep and can finally rest…

And the deeper you dig into the Facebook automation the creeper it gets.

It knows your hobbies, your browsing behaviour changes. How quick you’re changing websites, what type of posts are you engaging with, how long do you stop the scroll for on particular posts, etc.

There are millions of signals you’re sending it.

It’s downright creepy.

But as eCom growth experts, it’s our job to utilise this the best way we can for our clients.

Thing is older account management structures that do not allow you to leverage its power to the maximum.

What is working for us right now is what is called C.B.O. (Campaign Budget Optimisation) and our Ad Buying Framework we’re currently using – D.A.S.M. – Dynamic Ads Stacks Management.

What it entails is creating 3-4 “Stacks” of 1 C.B.O. Campaign with 5 audiences and 4 creatives in each.

What we do is set-up optimisation goals as Website Conversions – Purchase, to let FB AI. know what the desired outcome is. 7-days click and 1-day view optimisation window.

Now for audiences is where the magic happens. We use open or minimal broad targeting, letting AI pick the highest probability converters.

Usually, we have one audience with completely No Targeting set-up; maximum restrictions can be in the age gap and gender.

We have 2 5-10% L.A.L. best-performing audiences in a range of 15-30MM.

And 2 Interests that are at least 50-100MM in size. Sometimes even using “expanded interests” feature on Facebook.

After setting it up, we give it a few days to run.

Clean up 1-3 worst performing Ad Sets, and we’re left with a working “Stack.”

In case these “Stacks” does not work out, what we do is combine a few different “Stacks” winners into one “Best Audiences Stack.”

Now the whole D.A.S.M. framework coverts much more than that. The beauty of it is in how you’re able to dynamically test audiences and creatives at the same time while having a smooth account evolution approach that doesn’t put you in a “locked way of doing things,” but allows you to evolve the structure as things change.


Our budget throughout 2019 was in the low 5 figure range (confidential). Our split was 2/3rds Facebook Ads & 1/3rd split between Google & Bing

At least 50% of budget was consistently spent on prospecting

Facebook proved to have a much higher ROAS on prospecting so we kept Google & Bing supporting brand & retargeting


Through extensive testing and optimization, we were able to keep an average of 4X ROAS prospecting and 16X ROAS retargeting over five months as we worked to scale their advertising campaigns to over six figures in monthly revenue.

  • 25% Increase YoY website sales.
  • Prospecting ROAS – avg 4
  • Retargeting ROAS – avg 16

About The Client

Apparel / Outdoor brand




Consumer Goods and Services

Services Provided

  • Google Ads
  • Bing Ads
  • Dynamic Remarketing
  • Google Shopping
  • Facebook Advertising