January 3, 2021

PPC Case Study: 94% Reduction in Conversion Costs and 764% increase in Conversion Rate

The Problem: $284 Phone Calls

November 2019 - Feb 2020

Who would pay $284 for a phone call? Can you imagine spending $5,000 for 17 phone calls? Before my role at Discount Dumpster, this was the reality. Agencies came and went without having a positive effect on the bottom line. Unfortunately, this is the case time and time again. In another role, it was the same story.

After the first few months on the job. Here is what happened. Cost per Conversion dropped, by a lot.

Feb - May 2020

3 months later, the average CPC is down by 77% from $13.88 to $3.38.

Ten months and $24,000 later we are still going strong.

  • Conversions - $284(16) to $16.99(1323) = 94.1% Cost Reduction
  • CPC Average - $13.88 to $4.43 = 68.08% (3.13x)
  • CRO Average - 3.41% to 26.06% = 764% (7.64x)
Feb - December 2020

The Solution: Key Problems Addressed

Reorganized the structure of the groups into SKAG's, also known as single keyword ad groups, and optimized bids.

Why Negative Question Words?

Companies with a larger budget with the resources to convert users at an earlier stage of the funnel might find question words an excellent place to start advertising along the consumer journey.

For example, blogs starting with "How to" or "Best", about your product or services could be used as top of funnel touchpoints. Additionally, your sales team could follow up with leads on a later date. However, most small business owners do not have the resources to spend on top of funnel search queries.

Additional Funnel Opportunities

  • Sales team follow-ups
  • Blog posts addressing question words
  • Retargeting unique offers based on the query
Removed Search Queries "How" "Price" "Prices" "Cost" "Job"

End to end Attribution

Because of the use of Callrail and Salesforce, I was able to track conversions end to end and determine which calls resulted in a sale.

What I found after doing data analysis through Tableau is that top of funnel searchers turned into customers, only after additional engagements on later dates. However, not at the same rate as the first time, bottom of funnel callers. This is because people at the bottom of the funnel do not need any more convincing, they're people ready to order.

"Prices for Dumpster Rental" - Phrase Match - "Dumpster Rental

Explanation: Someone looking for "Prices", is still in the exploratory phase, they will likely call around asking for prices and may or may not end up ordering with us.

"Cheap Dumpster Rentals Phoenix AZ" - Phrase Match - "Dumpster Rentals"

Explanation: Someone looking for "cheap" has a subjective opinion about what cheap means. You and I have different opinions on what cheap means.

Additional Funnel Opportunities

  • Unique landing pages based on queries such as "Prices" showing users our unique selling proposition and what makes our product stand out.
  • Mapping the consumer journey around question words to create content around those queries, such as "Budget Dumpster VS Discount Dumpster", blogs written by influencers.

Why Manual Bidding?

If your quality scores are bad, and your landing page is not optimized, costs can get out of hand fast.

Automated campaigns need hundreds of conversions over a short period for automated strategies to work effectively, Google can get by with as little as 20, but the more you have the better. One of the advantages of automated bidding strategies is that it can use data not publicly available to advertisers, such as the previous search queries, and search intent, this can theoretically optimize a campaign.

However, as an advertiser you lack control over what queries or information the automated strategy is optimizing for, often you end up bidding on queries that would result in your desired action(phone call), but may or may not result in a sale.

For example, an automated strategy that's focused on phone calls will try to get phone calls no matter what the user's intent is(In this case we want users to order a dumpster). It could optimize for people trying to call your competitor. Or looking for jobs. Or maybe they're curious about the dimensions of a similar product. It is important to note that semi-relevant queries are not inherently a waste of money; they could be opportunities to increase revenue. But you also have to ask yourself if you want to pay Google money to act as a Q&A phone service.

What I've found is that when you look at the search query report, as seen in one of the sections above, you'd find that the automated bidding strategy was bidding for search queries from people asking questions, or searching for a competitor. Once again, this strategy can work if you have the resources to answer questions or to offer signup rewards to customers willing to switch. But you're also paying twice as much to answer those questions. This feels like a double whammy to me; paying twice as much for queries that have a lower probability of becoming a customer.

Another effective way to reduce costs is by optimizing a campaign for the Google Adwords Quality Score. Which is covered in another case study.

Additional Funnel Opportunities

  • Using keywords and conversions to drive the SEO strategy, which keywords result in the most sales

In the example below, there are queries for the same market, both at different prices. The automated campaign used factors that are not publicly available to bid up the same query, based on the premise that the person had a higher probability of making a phone call. However, the conversion rate did not increase in the automated campaigns. What ended up happening is the costs went up and that's about it.

Same Keyword, Same Market High CPC - Automated Campaign (non-SKAG)
Same Keyword, Same Market Reduced CPC

New CRO optimized Landing Pages with an Average 26% Conversion Rate

If you reduce cost per clicks from $12 to $4(3x), you are on the right track, but if you also increase conversions rates from 4% to 8%(2x), then you've effectively done the following to your conversions

3x2 = 6x more conversions for the same cost.

While I would have been happy with an 8% conversion rate, the results were much better. The previous agency was directing users to a standard location page, which was a mistake in my opinion. This standard page had a horrible PPC conversion rate. Though it's organic search conversion rate was somewhere between 7-12%. It's speculative to wonder if the irrelevant phrases from the automated campaign took the conversion rates down, instead of up.

You can see the original Phoenix Metro landing page here. It has a conversion rate of 1.48%, with a Cost Per Phone call at $400. It did not change very much between October and November 2019.

$400 Cost/Conv and 1.48% Conversion Rate
Sitewide Conversion Rate Sep 2018 to Feb 2020 - 3.4% - $14,000 for 66 Phone Calls.

Optimized Landing Pages

From the screenshot above, the conversion rate between September 2018 to February 2020, was roughly $14,000 in Ad spend and averaged a conversion rate of 3.41% over its lifetime. Resulting in 66 phone calls at a lifetime average cost of $205.78 per call. If one were to double the conversion rate, then one could double the number of leads received with the same ad spend.

7% would have been fantastic, but I was able to reach an average 26% conversion rate through trial and error, while A/B testing pages on Unbounce. Call quality actually went up during the same timeframe, as we were negativing keywords that were irrelevant to our intended goal, phone calls, and sales.

Experimenting with colors and structure, as I did with UX experiments in Google Optimize also helped the conversion rate.

On a fundamental level, the things that matter the most on a highly converting PPC landing page are above the fold, the very first things a user will see. The bullets below are the design essentials for above the fold content.

Conversion Rate Optimization Essentials

  • Lack of Menu
  • Reviews of Business
  • Company Logo
  • Image of Product

Old Page vs New Page

Conclusion

Roughly one year later and $24,000 in Ad spend, the Cost per Acquisition is still at a healthy $16.99 for every unique phone call.

Results

  • Conversions - $284(16) to $16.99(1323) = 94.1% Cost Reduction
  • CPC - $13.88 to $4.43 = 68.08% (3.13x)
  • CRO - 3.41% to 26.06% = 764% (7.64x)

7.64 X 3.13 = 23.9 Times More Leads with the Same Budget

Leave a Reply

Your email address will not be published. Required fields are marked *

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram