Skip to content
3a_-_logo_ecom_ads2-_1_-300x134
kisspng-google-adwords-google-partners-advertising-pay-per-google-partner-5b0ef14e9d4ce7.5072306115277059346443
PPCSamuraiPlatinumBadge_440x200p–2

CASE STUDY: BABY APPAREL BRAND

Quadrupled Non-brand revenue + increased profitability

Baby Apparel Brand

This is an 8-figure, US-Based, Baby Apparel brand.

Nick, the owner, contacted us as he was missing leadership on their Google Ads. Their agency was doing the bare minimum and even though he wasn't an expert, he knew he was leaving money on the table.

There were some quick wins but the long term wins is what we're going to focus on in this case study:

  • Quadrupled non-brand revenue and sales YoY while improving ROAS to 225% (was 146%)
  • Tripled overall Google Ads Revenue to $3.2M (was $1.03M)
baby apparel

INTRODUCTION:

To illustrate the timeline and progress we split this case study into our 4 major processes:
  • In-depth Account Review
  • Performance baseline
  • Trimming the fat
  • Scale

All figures in the image below are shown in USD. And July-March 2020 are compared to same period last year. (Total Revenue)

Posh_-_Compare_YoY_-_All_-_July_-_March_2020__WATERMARKED_FINAL STAMP

| PHASE 1 - IN-DEPTH ACCOUNT REVIEW |

This is where we establish the status quo. Before we go in and optimize, we first need to know where the account is at.

Where are the opportunities, what are the weaknesses? Any low hanging fruit?

Here are the big takeaways:

BRANDED TRAFFIC WAS SPREAD OUT OVER ALL SEARCH CAMPAIGNS

Due to bad negative keyword management, branded search terms were sneaking in through a lot of non-brand keywords.

(A lot of agencies are using tactics like this to make their “non branded” traffic appear better than it actually is. )

GOOGLE SHOPPING PERFORMANCE WAS BASIC

Your standard “one campaign, one adgroup and thousands of SKUs” was applicable here. No control, and therefore wasting a TON of BUDGET.

PRODUCT FEED WAS NOT OPTIMIZED TO MANAGE 3000+ SKUS

If you want to manage thousands of SKUs with shopping it’s essential your key attributes are filled out properly and you leverage custom labels to create an effective architecture in your Google Ads account.

Write this down: Google shopping success = 70% product feed, 20% campaign architecture and 10% bids.

ADS WERE BASIC AND MISSING HOOKS AND ANGLES

Due to account architecture, the ads were not dialed in into the intent, nor was it highlighting the keywords in the headline.

TOO MANY INTENTS COVERED IN ONE ADGROUP

As there were intents mixed in one adgroup it’s very hard to write an ad for the different intents. You don’t want to do this as it will lower your CTR. On top of that, you want to create a relevant user experience, and linking to one URL for different intents isn’t a good one as you have no control over which ad will show up per adgroup.

FEW AUDIENCES AVAILABLE

Smart bidding and retargeting campaigns can flourish when you have sophisticated audiences available. As machine learning leverages your audiences and is trying to connect the dots it’s essential you have powerful, sophisticated audiences in place covering different time periods and events.

DUPLICATED TRACKING

Due to the installation of a shopify app by the client a second pixel was installed which literally destroyed the performance overnight. All campaigns had to go back into learning mode (Took us 3 weeks to get it back up and running - stability is everything).

| PHASE 2 & 3 - PERFORMANCE BASELINE & TRIMMING THE FAT |

So we know now what the weaknesses are and what needs to be changed to get the account to perform at a higher level.

SIDENOTE: Setting up new campaigns is resetting Google’s algorithm so you want to make sure you do this gradually and not all at once.

Here’s what changed per point.

BRANDED TRAFFIC WAS SPREAD OUT OVER ALL SEARCH CAMPAIGNS

In this case, it looked like NB did fairly well. However, after dissecting the traffic and excluded ALL brand-related terms the non-brand traffic was actually in DECLINE.

You always want to exclude your brand terms from ALL your non-brand search campaigns to make sure you’re isolating brand traffic in its own campaign.

Why?

Four reasons:

  • Get a great overview of how your brand is developing via volume impressions
  • Control a 100% impression share on your branded terms as you should
  • Control your messaging for brand-related queries
  • And show the truth about how your non branded traffic is performing so you can adjust and craft a dedicated strategy

GOOGLE SHOPPING PERFORMANCE WAS BASIC

Optimize the product feed to segment products based on colors, margins, categories and material. Then leverage these attributes to create multiple campaigns giving you the full control over:

  • Budgets
  • Bids
  • Nature of the traffic (TOF MOF BOF)
  • Different SKUs

ADS WERE BASIC AND MISSING HOOKS AND ANGLES

Revamping the search campaign structure making sure that similar intents are grouped on the SAME adgroup. This gives us the opportunity to write highly relevant ads towards the intent and link to the most relevant landing page. Overall we’re creating the best possible user experience from search to check-out.

On top of that we’re adding multiple ads per adgroup that hit different hooks or angles so the smart bidding algorithm can learn which angles work best with what target group.

NOTE: As A.I. and machine learning will only get smarter, bidding becomes less important.

However, they aren’t creative enough to create appealing angles and hooks for ads and therefore you need to map out who your potential buyers are and what angles might appeal to them. Then let the machine do the work for ya. Times are changing!

FEW AUDIENCES AVAILABLE

Creating more audiences in Google Analytics for us to leverage. This is something you want to do from day 1 as it can’t really retract data. So if you want to create an audience of let’s say ATC last 540 days, it can max retract that of last 30 days. So you want to make sure you have created all your sophisticated audiences at the start. Otherwise you’re losing out on a lot of opportunity.

DUPLICATED TRACKING

This one was tough. I can’t stress enough the importance of stability when working with machine learning. Your whole performance comes crashing down and burning. In this case the client had no idea a pixel was created through an app and it took us a few days to see it. It’s like teaching a kid how to ride a bicycle. It gets better and better everyday. Then one day it falls off the bike and it forgot how to ride it. Starting all over again. Luckily Google just introduced seasonality bid adjustments which can take into account fluctuations. However, there isn’t a solution for duplicate tracking LOL.

TRIMMING THE FAT

After acquiring new data we benchmarked the performance. From that point forward it's a matter of reducing ad spend where it doesn't convert. That phase is what we call "trimming the fat".

We FIRST want to trim the fat before we start scaling, otherwise it's like rowing a boat with holes in it.

We want to make sure we have the leanest set up so we can instruct the system with high quality data + volume.

All figures in the image below are shown in USD. And July-March 2020 are compared to same period last year. (Non-Brand)

Posh_-_Compare_YoY_-_Non_Brand_-_July_-_March_2020_NEW

| PHASE 4 - SCALING |

SCALING SEARCH CAMPAIGNS

We get asked about this phase A LOT. So I’m going to show you how we did it for this client.

First of all, you want to identify what traffic is TOF, MOF and BOF. For example, keywords like “sneaker” is TOF as it’s very generic, we don’t really know what they’re looking for outside something with sneakers. As it’s very generic a lot of traffic will be fetched when you book this keyword.

“Nike Sneaker” is what I would identify as MOF as it’s more specific and shows a clear intent to a brand.

“Nike Air Force One Green” is your typical BOF intent traffic as it’s specific enough to know exactly what they’re looking for. Makes sense?

So there are various ways to go from here depending on the keyword match types you want to use.

I don’t want to get into the details of that too much so here’s what we’ve done:

# TOF CAMPAIGN #

Contains (modified) broad keywords. These keywords fetch search terms related to the keywords we booked. Singular, plural, synonyms, but also queries that contain the words of our keyword. Example, “+sneakers” can trigger the following search terms:

  • Sneakers
  • Nike sneakers black
  • buy Adidas sneakers footlocker
  • Trendy sneakers celebrity 2019

Different intents can be triggered and it can generate a lot of “white noise”. This is your campaign that is going to take your most time to manage.

# MOF CAMPAIGN #

This campaign was completely covered by our dynamic search campaign who was bidding on keywords based on the schema data on the product pages.

Google crawls the product page and based on the content, it starts bidding on keywords to trigger your ads with dynamic headlines and URLs to the right products. A great way to find new keywords you haven’t thought off before.

# BOF CAMPAIGN #

This campaign only contains exact keyword match types of the search terms that have converted in the past.

So [nike sneakers] will only trigger auctions when somebody is searching for:

  • Nike sneakers
  • Sneakers Nike
  • Sneakers from Nike

The auctions it triggers stays true to the intent of the keyword without adding extra words.

We exclude all the exact keywords from our TOF and MOF campaign to make sure we’re funneling the right traffic to the right campaigns + messaging.

Here’s the important part. As we’ve identified and split the different funnel traffic we can now:

  • Allocate different budgets to the different campaigns for optimal control
  • Set different KPIs per campaign
  • Have a better overview of the performance

TOF traffic has a lower conversion rate and thus ROAS than the BOF campaign. So by splitting them in separate campaigns we can set different tROAS targets and budgets.

The TOF campaign is highly scalable, is a great source for finding people who haven’t interacted with your brand before but is at the same time a source of expensive traffic due to its generic traffic.

What most advertisers do
What you should do instead

SCALING SHOPPING CAMPAIGNS

The same approach we implemented for shopping campaigns.

However, before you can do so you need to optimize your feed in a feed manager. In this case we leveraged attributes to segment products on:

  • Category
  • Color
  • Material
  • Margins

Then we created shopping campaigns with high and low priority to filter out TOF and BOF traffic so we can again:

  • Allocate budgets accordingly
  • Set different bids and targets per shopping campaign and nature of the traffic (TOF or BOF)

NOTE: This approach only makes sense when you have high volume traffic. We have to take into account that just a few click per month on a campaigns is useless and then rather combine multiple elements to send valuable campaign traffic to the algorithm.

CONCLUSION

Not touching things after implementation is optimization on itself. Let the machine do the work while you “pilot” the campaign and be “the teacher” to the algorithm.

Be patient - Google Ads is a different beast than Facebook ads. Learning phases are different and longer but if you crave more consistency in your business it’s the way to go.

NOTE: Take into account the sale delay for your website. For this client, the time lag to a sale was about 11-13 days. So don’t draw conclusions too quick. Let the machine do its work and remember: Patience is profitable.

---------------

If you want to grow your brand outside of Facebook ads and are looking for a partner that takes leadership of your account, then book a FREE strategy call with us and we’ll have a look if we can help your Google Ads account to the next level.

Don't take just our word for it

What others say about us

So you do your due diligence. Why would you spend your precious time going on a call with these guys? Exactly. You need some stories that resonate with your situation. So go check some of them out, if it helps.

"In the First 6 Months We Saw an Increase of 200% in Spend and Return."

"Five months ago we were averaging at a Return on Adspend of 150%. Right now we're averaging at a 450% ROAS while continuing to expand the campaigns...Which is amazing!"

"We Now Have Multiple Google Ads Clients Coming In And It Brings In a Lot of Extra Revenue."

We deliver onlne marketing done with you services that are fully crafted to your needs. There's no "one size fits all" solution, nor is there one magic framework that rules them all. We are flexible and as long as you're committed, we're 100% confident we'll take you to the next level(s). Want to find out if we're a good fit? Hit that button and apply for a strategy call.

3a_-_logo_ecom_ads2-_1_-300x134

Digital Marketing Consultancy

eCom Ads BV

Bos en Lommerplein 270-300

1055RW Amsterdam

the Netherlands

Copyrights © All Rights Reserved By eCom Ads

Your content goes here