Learn how one startup doubled CPL and saved money

In this case study, a focus on keywords and cleaning up data to better understand CRO led to a 55 percent decrease in purchase CPA and larger deal size leads.

Chat with SearchBot

Stop me if you have heard this one before: We want the cheapest lead possible.

One consistent theme that I hear from startups and international brands is people asking for a cheaper lead each month and year over year.

However, that is the faster way to waste money and, in some unfortunate events, go out of business. There are a couple of reasons why you don’t want the cheapest lead:

  1. Paying Customers: A cheap lead looks great early on but as you spend money each month and see fewer leads convert into a customer, it becomes increasingly hard to justify your spend on paid advertising.
  2. Wasted Time: If your leads are not converting then this means you are wasting a lot of time on business development that is not helping your business grow. You want to spend time on the leads most likely to convert.

Let me give you a real-world example. We spent the last eight months working with a SaaS startup that competes with Shopify, WooCommerce and Magento in the e-commerce space. We took on the SaaS startup when their cost per lead (CPL) was $49 and their cost per acquisition (CPA) for a paying customer was $1,734.

The startup wasn’t happy as the number of leads that turned into customers was inconsistent from month to month, and they were wasting thousands of dollars on keywords that were not converting. Worst of all, the data they had was a mess to sift through due to a poor strategy.

This meant we had to get the advertising spend under control, make sure we had proper data to work with and truly understand the type of customer who was going to convert. When we did all of the above, we achieved the following, even though we doubled their CPL in the process:

  • Lowered their purchase CPA by 55 percent
  • Larger deal size leads that the sales team could sell to
  • Had consistent sales each month compared to the last three in-house people
  • Brought in customers that have a similar or higher LTV

CPA chart

How did we do this? Let me walk through our plan of attack as we launched this brand in eight countries within the first three months of working together.

Working with data

Data is at the heart of what do we as PPCers. However, the client didn’t have a data framework to understand what was working on Google and Bing. The first thing we did was an audit on their advertising and analytics accounts to make sure everything was set up correctly.

Once we fixed a few issues and made sure we had a UTM tracking bible in place that everyone in the company understood, we got on with the heavy lifting.

Having a consistent and standard naming convention saves time and money. A good example is how your company names paid search traffic; is it CPC or PPC?

Having a single source of truth makes it easy to compare results as the company grows and accumulates more data. This startup was going to have a large amount of data as we planned to launch in eight countries within the first three months of working together.

Getting spend under control

One of the things we saw early in our account audit was a heavy focus on display ads. We turned those off as they were not converting into paying customers. We wanted to focus our spend on keywords to help us make sure our ad copy and landing page matched the searcher’s intent.

For startups, we look at building campaigns based on the following keywords: brand, competitors and then non-brand keywords. If we have enough traffic, then we look at doing remarketing. Lastly, we always test a dynamic search ads (DSA) campaign to see if it works for the brand. Frederick Vallaeys has a great article covering DSA campaigns and automation.

Brand campaigns are self-explanatory. Competitor campaigns are a little more tricky. We had a budget that allowed us to focus on the top five competitors in this space including Shopify, WooCommerce and Magento. However, if the budget your working with is an issue, then we recommend only focusing on the top two or three competitors.

We built a campaign for each competitor and used the following seven keywords below. The reason we gave each competitor their campaign was to help control spend and easily see which was converting and profitable.

keyword list

Over the years we have found that these are the seven ways people do their research when looking for software and comparing options. People searching for a competitor are lower in the funnel (or closer to deciding to buy).

The phrase or BMM keyword may capture traffic from current customers who want to call up your competitor’s support team. It’s a good idea to use your negative keywords to block traffic related to customer support request.

Once we had set up the brand, competitor and DSA campaigns, we waited until we had enough data to launch a remarketing campaign. The remarketing campaign was going to target those who had become a lead but not a paying customer.

If you have already paid to get someone to your site and have them become a lead, you should try to convince them to become a customer (when it makes sense). However, don’t just show ads to leads non-stop, that is the fastest way to annoy someone.

For non-brand campaigns, we looked at keywords including +Online +Shopping +Platform or +Create +Online +Store. We would only bid on non-brand keywords when we had room in our budget. It’s always best to start with your lowest hanging fruit, which is usually your brand, competitor and remarketing campaigns.

Customer insights with CRO

Understanding your customers and why they buy is essential. This helps you write better ad and landing page copy that speaks to a customer during their research. Gathering this customer insight falls under conversions rate optimization (CRO).

CRO is not about a/b testing or merely increasing your conversion rate, which is a misnomer in the name. If you do CRO right, you can convince customers to buy higher tier packages from your pricing pages or acquire customers who have a better lifetime value (LTV).

Both of these don’t require a better conversion rate but do require talking with customers to understand them better. When we do CRO work for clients, we run surveys, online polls and even meet customers face to face. We want to understand the reason they bought and the value they are getting out of the software. Never be afraid to talk with customers at a regular cadence.

We even do online research on Reddit, TrustPilot and TrustRadius to see what some of the common problems are in each industry. There is a lot you can learn about a product on the sites, which can also work out to help developer better ad and landing page copy.


Working on a SaaS brand takes a strong strategy, with the right data, to make choosing a path a lot easier. We couldn’t have succeeded if we only did one thing above. We had to fix their data, get spending under control and understand why customers were buying in the first place.

These strategies led to a 55 percent decrease in the purchase CPA for our Saas brand in 2018. As we plan for 2019 with them, we are looking at new opportunities in new countries as well as new ad platforms like Facebook and Instagram advertising. The future is bright as long as we don’t get hung up on the cheapest lead possible.

Contributing authors are invited to create content for Search Engine Land and are chosen for their expertise and contribution to the search community. Our contributors work under the oversight of the editorial staff and contributions are checked for quality and relevance to our readers. The opinions they express are their own.

About the author

Duane Brown
Duane has been called an international man of mystery and digital nomad by friends. He has lived in 6 cities across 3 continents and visited 40 countries around the world. He uses his curiosity for people and love for people watching to run better marketing campaigns for clients. After leaving Toronto in 2011 to gain an international view of the world. He has worked for Telstra in Australia and brands including ASOS, Jack Wills and Mopp (bought Sept. 2014) while in London, UK. He now lives in Montreal, Canada helping brands grow through data, CRO and marketing at Take Some Risk Inc.

Get the newsletter search marketers rely on.