5x ROAS for eCommerce SaaS (In Depth)

I am going to walk you through an eCommerce SaaS google ads case study with a 5 figure monthly ad budget, from being a company with account maintenance challenges, room for improvement on their returns, and questions on the value of paid search to the organization, to generating a stable 5x-7x return on their ad spend and with confidence in their reporting capabilities.

Before we begin, I think it’s important to understand a little bit of background about the company, the product, and the economic climate at the time. This may not be the most interesting stuff if you just want some hot tips on how to improve your Google Ad account, but I encourage you to read the background, as it will help you get more out of the overall Google Ads case study and help you grow your business with paid search. If you would just like to see the results, jump to the end of the case study where all the results have been summarized.

In order to protect their competitive position, we’re going to keep details about the company a little vague, but the details we can share are as follows.

ECommerce SaaS provider

In business for over 15 years with an established brand

Spending approximately 20k a month on Google Ads

When I came on board, this was a company who’s owner has multiple businesses and was focused on their next venture and was looking to delegate aspects of maintaining their revenue stream. Multiple agencies had been involved previously that weren’t quite a fit, and as such ads were being managed in-house by a head of marketing.

The product is a software as a service product with online signup and use. It helps eCommerce store owners connect with a specific type of supplier, get access to products quickly, and allows them to integrate the products with their store via API connections to most major ecommerce platforms such as Shopify, Amazon, eBay, BigCommerce, etc.

Once the store and supplier integrations are set up, the product provides a number of syncing, inventory, and order management capabilities.

I knew from our conversations before getting started that this eCommerce SaaS company had a lifetime customer value for their subscribers of ~$800 . This was a blend of ambitious first time store owners, and veteran store owners with long time stable and profitable businesses that would subscribe to the product and use it to scale their eCommerce empires.

I reviewed historical data within the Google Ad account and found a number of questions about the value of their investment that were difficult to answer with their current reporting infrastructure. One of the first things that we noticed when doing our initial due diligence was that Google Ads tags were not present on their website and that conversions and audiences were built off Google Analytics and not leveraging Google Ads tracking.

Most ad campaigns had the default settings and were leveraging auto applied suggestions. Most ad groups did not have responsive search ads and many expanded text ads were lacking optional headlines and descriptions that presented an opportunity to increase CTRs. Ad extensions were not being leveraged to their full capability and many had outdated copy.

There were many issues to tackle and since this was an older account with a long history, there were hundreds of ad groups, and thousands of keywords. I discussed with the client, educated them on the opportunities for improvement, and discussed that it would be a process to move to a better end state.

Eventually I agreed upon a goal of bringing the ad account to a 3x return on ad spend with an estimated 3-6 month timeline for doing so and expectations of demonstrating incremental improvement over that time period. For simplicity’s sake, I decided I would measure progress towards that goal by multiplying the number of paid subscriber conversions against the average lifetime customer value to get an estimated revenue. We would use this estimated revenue and divide it by the ad spend for the month to calculate ROAS (return on ad spend).

150 Paid Subscribers Acquired $800 ACLTV (Average Customer Lifetime Value) 150 * $800 = $120,000 Estimated Revenue $20,000 Direct Ad Spend $120,000 / $20,000 = 5.0 ROAS (Return On Ad Spend)

The first issue I tackled after we began was to clean up the reporting infrastructure for the account. As I mentioned earlier, Google remarketing tags were not present on the account and conversions had been imported from Google Analytics rather than leveraging the native Google Ads conversion tracking.

This is often times a sub-optimal setup and that was the case here. It’s an easy mistake to make because the impacts of it aren’t well understood and Google themselves regularly present it as an option for conversion tracking. To understand why this is a sub-optimal setup requires an understanding of the differences in attribution models between Google Analytics and Google Ads.

Google Analytics uses what’s known as last non-direct click attribution modeling. What this means is that conversions will be fully attributed to the last non-direct channel.

A direct visit is when someone explicitly types the URL of your website into their browser, visits via a bookmark, or otherwise comes directly to your site.

What this means is that if there is an intercepting channel, (such as an organic visit from a search engine, or a social media referral) that comes after your paid search click, then paid search will effectively be erased as a channel and you won’t get ANY credit for a conversion, even if that visitor was originally sourced from paid search.

This can result in the loss of some pretty valuable data and really under count what the impacts of your paid search program are. Imagine a customer found the product via paid search on Google Ads, subscribed to the email newsletter on their first visit, and then returned shortly after via a click from an email to subscribe.

With conversions imported from Google Analytics this would not be counted as a conversion and you wouldn’t get any visibility into what the keyword or search term was that was responsible for acquiring that customer. You would however get that information when using native Google Ads conversion tracking which will persist and count conversions, regardless of whether there was an intercepting non direct channel (such as organic, or email) between the paid search visit, and the end conversion.

I installed their conversion tracking which included a conversion for a free trial signup, and a conversion for a paid subscription within Google Tag Manager. I also installed Google Ads remarketing tags for remarketing audience building.

After I moved to native Google Ads conversion tracking, our attribution modeling journey wasn’t done yet! I still identified another beneficial change to the reporting infrastructure in the form of attribution modeling for multi-click conversion paths.

Consider a scenario where there are multiple participating clicks in a conversion journey via paid search. Let’s say a user first encounters the website with a search term like “eCommerce software”, later in their research, they search for “best ecommerce software for shopify” and lastly after they are done comparison shopping and have decided to go with your brand they search for “company brand name” and click.

Position 1 – “ecommerce software”

Position 2 – “best ecommerce software for shopify”

Position 3 – “company brand name” (Converts here)

How do you distribute the credit for each click? This client was using last click attribution models, which is the default, for all their conversions. In the scenario described above, all credit would be assigned to the last click, “company brand name”.

With up to 60 day conversion cycles, and a known set of multi-click conversion paths, this did not make sense. We were missing visibility into valuable information on which search terms, such as “ecommerce software” were driving awareness and actual customer acquisitions. We knew we had to assign at least some credit to top of funnel keywords. We reviewed the options for switching the attribution model, with the options illustrated in the graphic below.

After some discussion and analysis of the options based on their analytics data, I decided that position based was the best fit for this particular client. If you work for Google and you’re reading this though, we’d really love a reverse time decay model!

Putting some thought into these settings is an important step because they directly affect the number of conversions you can attribute to Google Ads, and where the credit is distributed. This is important because much of Google ads bidding these these days is done via AI and machine learning.

Machine learning and AI is very powerful because it can automate a lot of tasks and drive real time decision making with minimal human interaction. In general though, machine learning and AI is very effective at making the same decision that has been made in the past based on a set of available information, not necessarily at making quality decisions based on a set of new criteria. What this translates to is that the better you tune your conversion tracking and attribution models, the more informed the conversion optimizer becomes and you are able to get a better ROI. When Google Ads has better information about which searches led to conversions in the past, then it can make better decisions about which bids will lead to conversions in the future.

It’s also critical to have the correct conversion tracking and attribution models in place so that you can make quality decisions about how to make strategic changes to your ad account. Installing these tags and configuring these settings lays the foundation for accurate reporting and performance measurement.

With the Google Ads conversion tracking in place, appropriate attribution modeling settings, and conversion tracking windows set for what made sense based on the nature of their business’ purchasing cycles, we were ready to sally forth and begin bidding on Google Ads again. On to the next part of our Google Ads case study: updating our keywords and the search terms they generate.

The next part of our Google Ads case study will be to talk about how I updated keywords and search terms. With a 5 figure monthly spend and a long account history, there was rich data available to analyze. Legacy conversion tracking described above made some of it inaccurate, so we had to lean on the side of caution, but there were several things that jumped out at us as opportunities for improvement.

Disabling Auto Applied Suggestions

Inquisitive search terms

Lots of exact match keywords

Broad Ad Groups

Just a quick reminder, search terms are the exact phrases that a searcher typed into Google and keywords are the target search phrases that you as the advertiser enter to target.

Google makes it very easy to just set an amount of money you want to spend every month, and forget about your account. One of the ways they do this is with auto applied suggestions. Google will scan your account for opportunities and automatically apply them to your account after a 14 day notification period. For accounts that aren’t actively managed it’s not a bad feature, but for active account managers it’s clearly not a fit.

One of the biggest criticisms of the feature is that it largely just finds ways to expand your reach, and spend more money, relying heavily on machine learning to generate leads. Yes it can help you generate more leads, but often times with low efficiency in spend. For advertisers without PPC expertise, it really is a great feature, but for 5 figure spend accounts with active management, it is not, and as such, I disabled it. Over the years, the auto applied ad suggestions had instituted a number of changes in the account and which were a contributing factor to the lower than possible return on ad spend that I uncovered in my due diligence. Many of the issues described below were implemented because of this feature.

One thing that jumped out at me was that was a pretty substantial volume of search terms were inquisitive in nature. What we mean by that is that the search term contained phrases like “how do i”, “what is”, “where can i find”, etc. These types of phrases were converting very poorly, both for free trials, and ultimately paid subscribers.

We compiled a list of low converting inquisitive phrases, put them into a negative keyword list, and applied the negative keyword list to all the campaigns in the account. Negative keywords can always be removed later to expand reach, but our initial strategy for generating 3x ROAS hinged upon betting on obvious winners.

Another thing I noticed was that over time, the account had grown by adding more and more exact match keywords.

Exact match keywords can be a good fit for low spend accounts that need to have a very tight focus and where every dollar is critical. This however was an established business with a little bit more risk tolerance and a healthy ad budget.

As you scale an ad account, maintenance becomes a real concern and maintaining hundreds, if not thousands of exact match keywords can become a burden and also significantly limit your reach.

One of the things I did to improve the health of the account, and reduce the maintenance required is identify areas where exact match keywords could be consolidated into phrase match and modified broad match keywords.

Since I was going to be actively maintaining and monitoring the account for negative keyword candidates moving forward, this made a lot more sense because it would simplify the account structure but also increase the base of keywords that the account would target. I would control erroneous or low quality spend with negative keywords.

One of the other things I observed in the ad account was that there were a number of overly broad ad groups. What we mean by this is there were ad groups with a mix of keywords targeting fundamentally different searches. For instance there were ad groups that had keywords like “shopify ecommerce software” and “sell more on amazon” mixed in the same ad group.

This can present a problem because the ads within the ad group rotate regardless of search term. If you have an ad written for people looking to sell on Shopify then there is a chance that you can show them an ad for selling on Amazon instead.

In Google Ads you want to have your ad groups and the supporting ad copy to be focused on a tight theme where the ad copy for the ad group will resonate with all or most of the possible searches that will appear in the ad group.

Doing this will increase your click through rates (CTR) and your quality scores. Higher quality scores mean you pay less when you win an auction, so you can save a lot of money by putting some extra effort into your ad groupings and overall account structure.

I identified a need to split ad groups or reorganize keywords into different ad groups throughout the account in order to make sure the ad copy resonated with the searchers. Speaking of ads, let’s move on the the next section of our Google Ads case study and talk a little bit about bringing ad copy up to par.

Structural changes and conversion settings can go a long way to improving the quality and capability of your Google Ads account. These largely help show the right people the right ad, but at the end of the day people are being shown an ad and have 2 choices, click or don’t click. This is where ad copy comes in. Is it compelling? Does it resonate with what the searcher is seeking? The next section of our Google Ads case study for this ecommerce saas provider is to talk about updates to ad copy.

Ad copy is one thing our client excelled in. As you can imagine, being in business for 15 years and serving thousands of clients along the way, they had a pretty clear value proposition, and understanding of what it is the market wanted. There was very little need to experiment to find what sort of copy was generating clicks and ultimately conversions.

There was however a number of ways we were able to add value by increasing the quality of ads via subtle changes. These can be summarized as:

Moving To Title Case

Refreshing Copy To Match Current Offerings

Maximizing Ad Real Estate

Increasing Ad Volume

Many of the ads in the account were written in sentence case. I knew that converting ads to title case is a quick win that can generate lift in the near term. It’s a best practice in Google Ads to write your ads in title case. What’s more engaging?

sign up for a free trial today -or- Sign Up For A Free Trial Today

If you thought the second version was more engaging, you’re not alone. Google themselves recommend putting your ads in title case and says so in their certification courses. If that’s not convincing enough I looked at a study that compared ads in sentence case to ads in title case and found these results:

CTR improved by almost 1 full percentage point.

Conversion rate (CVR) was twice as high as the CVR for the ad version in sentence case.

There were 3 times as many conversions in the title case ad.

More revenue per conversion was generated via the title case ad.

While I was updating the ads in the account to title case, I also took the opportunity to update ad copy to match the business’s current offers. Remember back earlier in this eCommerce SaaS Google Ads case study I told you that the company helped store owners connect with a special type of supplier? Well the volume of those suppliers had grown substantially over the years, but the ad copy had not. Phrases like “over 100 suppliers” or “connect to more than 10 eCommerce platforms” were updated to match the business’ current offers and changed to phrases like “Over 250 Suppliers” and “Connect To 25+ ECommerce Platforms”. I applied changes like this to both ads and ad extensions.

The next change I made was to make sure that I was maximizing ad real estate. I noticed that were a number of ads in the account that didn’t have all the available fields filled out within the ad. What this translates to is a whole lot of available real estate on the search engine results page (SERP) that was being left on the table.

Let’s look at a few examples.

Consider the ad below, the previews, and the real estate that it occupies. Notice the empty fields in the Headline 3, display paths, and description 2.

These ads don’t take up very much real estate on the SERP. They don’t grab as much attention as they could. They are also going to have a lower ad rank which translates to being shown less often and costing more per click. These types of ads can still generate clicks and conversions, but maximizing the use of available headlines, descriptions, ad extensions, and other ad settings greatly increases the real estate on the SERP that the ads occupy. This will increase click through rates and conversions, and at no additional cost. In fact, the higher click through rates can improve quality scores and save you money.

Compare the images above to the fully filled out expanded text ad below that is leveraging available ad extensions to see how much more real estate on the SERP you can occupy. (This has a side benefit of pushing your competitor’s ads out of view!) More engaging, right?

The last change we made as part of our efforts to update ad copy was to make sure that each ad group had at least 2 expanded text ads and 1 responsive search ad. This increases the ad volume in the account and gives each ad group 3 chances to enter the google ads auction and score the highest possible ad rank.

Many ad groups in the account did not have responsive search ads and in our experience responsive search ads are often the top performing ad in the ad group in terms of both click through rates and conversions. So, make sure you are leveraging responsive search ads!

Ad copy should be periodically refreshed and tested to see what works and what doesn’t. Having at least 2 expanded text ads and 1 responsive search ad in the ad groups is a great way to make sure the next version of your ads are of a higher quality than the last version. You can infer the types of copy that are resonating with your audience by reviewing the ad groups and seeing what is performing well. When your ad groups only have 1 ad, this isn’t an option as there is no basis for comparison.

Pro Tip: Create duplicates of your ads in the ad group and change only the landing page to get a better feel for which landing pages are performing better!

This was an in depth project that took a lot of effort but the the end results from working with this client have been fantastic. Some of the results were as follows.

Multi year highs in transaction counts and new customers acquired

Adding thousands in new monthly recurring revenue

A stable 5x-7x return on ad spend in their paid search account

Reduced burden on the internal marketing department