ARS®/Rescue Rooter® is one of the U.S.’ largest independent home services companies, offering heating, air conditioning, indoor air quality, plumbing, drain cleaning and sewer line services from company-owned locations across the country. A Zion & Zion client since March of 2009, we serve as the digital marketing agency for 35 ARS®/Rescue Rooter® location-based lines of business throughout the country.
When it comes to paid search, it’s fairly standard for digital agencies to declare their job as complete once the click has been delivered to the client.
In some cases, the agency will take responsibility for the actual conversion, be it an eCommerce transaction, a phone call, or completion of a contact form. However, the process that leads from click via paid search to final sale goes through a variety of steps and systems that are not easily linked or able to be viewed holistically. Our goals are not simply to generate clicks and calls for ARS®/Rescue Rooter®, but to understand and involve ourselves in the broader process of generating ROI for the client.
In the case of our paid search marketing for ARS®/Rescue Rooter®, there is a funnel that flows as follows:
- The customer clicks through to the landing page
- The customer decides whether to call the client
- The client’s customer service rep (CSR) either does or does not book an appointment
- The appointment yields revenue ≥
We performed detailed analyses on not just costs and conversions (i.e. calls), but also efficiency of CSRs with respect to appointment booking and average transaction sizes (based on post hoc revenue reporting).
The analyses we present in this case study highlight the extreme differences in performance by day and time of day for a variety of metrics and how those metrics compound to affect ROI. In particular, we present two different ARS®/Rescue Rooter® markets, but conceal the names of those markets to preserve confidentiality. You will note that the patterns found in one market differ substantially from the patterns found in the other, further highlighting the need to perform individual deep market analyses.
Our team amassed a considerable number of actionable insights based on our analyses. Many of the insights facilitated our agency’s management of the client’s paid search marketing, and many allowed us to provide guidance to the client on potential ways to optimize their management of budget and staffing. We provide a sample of those insights and their implications here, however there were dozens of insights with corresponding implications. We refer to the two ARS®/Rescue Rooter® markets in this case study as Market A and Market B.
NOTES: All metrics shown are reported as relative to the average.
For example, a Cost Per Click of -3.3% for Market A on a Thursday means that the cost of a click in Market A is 3.3% lower on a Thursday than the average cost of a click in Market A across all days.
Another example would be that Booking Rate of +2.7% for Market B on a Wednesday means that the booking rate in Market B is 2.7% better than the average booking rate for Market B across all days.
The matrices for each metric show Day of Week in the column labels and Time of Day in the row labels.
The reason that the Time of Day row labels only show data from 7am to 9pm is because the limited amount of data outside of that time range was so limited (e.g. very few clicks and very few calls) that outliers skewed the visual conditional formatting of the analysis. For example, if you look at the Average Transactions Size matrix for Market A, imagine, hypothetically, that there had indeed been only three calls at 3am in total across all Wednesdays. Now imagine that one of those calls was booked and turned into a transaction worth $50K. That would cause the conditional formatting to label that cell as bright green because it would be the best day and time across all days and time for the size of an average transaction. And all other cells in the Average Transaction Size matrix would shift their formatting to the white to red end of the spectrum making visual inspection of differentiation difficult. That fact, combined with the fact that we want to act on data that is likely statistically significant drove us to decide to focus on a Time of Day range that contained adequate volumes across all metrics in the analyses.
All references to Days of Week and Times of Day are based upon the date/time of the click. That is to say, for example, when you look at the ROI metrics for both Market A for Wednesday at 9am, that cell is calculated based off of: the cost incurred by all clicks that occurred in the 9am to 9:59am window on Wednesdays, AND the revenue ultimately collected for calls that are associated with those clicks. Note that a small amount of error is associated between the alignment of click and call times, however, our research shows that amount of error is minimal.
Insight 1 – Market A: ROI on click cost incurred on weekends is higher. This is due to the compounding effects on weekends of: lower weekend Cost Per Click, more Calls Per Click, and better CSR Booking Rate. And, while weekend Average Transaction Size is actually lower, this factor is outweighed by the compounding positive effects of other factors.
Insight 2 – Market A: While cost per click is generally (i.e. on Mon-Weds and Thurs-Sat) lower in the late afternoon and evenings than during other times of the day, ROI in the 7pm to 9pm time frame is generally the highest. However, the reason for this higher ROI differs depending upon the day. For example:
- The higher ROI between 8pm and 9pm on Mondays and between 7pm and 9pm on Tuesdays is primarily attributable to a combination of lower Cost Per Click and higher Average Transaction Size
- The higher ROI between 7pm and 9pm on Wednesdays is primarily attributable to a combination of lower Cost Per Click and better Booking Rate
- The higher ROI between 7pm and 9pm on Friday and Saturday is primarily attributable to a combination of lower Cost Per Click and more Calls Per Click
Insight 3 – Market B: Insight 1 from Market A (see above) is not true for Market B. While Sunday is the highest ROI day for Market A and Saturday is the second highest ROI day for Market A, this is not the case for Market B. Market B’s highest ROI day is Saturday, with Sunday being Market B’s second to lowest ROI day.
All of the above insights, and the many other insights we garnered, had not only bidding, ad scheduling, and ad copy implications for our paid search campaigns, but also potential call center scheduling and technician scheduling implications. For example:
- The client originally had concerns over lifestyle implications for technicians working on weekends, however, the high weekend ROI for Market A and higher Saturday ROI for Market B led us to recommend compensation changes as a solution based on the increased profitability.
- The client can revisit call center staffing based upon examination of whether Booking Rate variation was inherently due to Day of Week and Time of Day or whether Day of Week and Time of Day were simply a proxy for certain CSRs being on duty. That is to say, was Market A’s high Booking Rate at on Wednesdays in the 8pm to 9pm window due to the date and time, or was it due to the fact that someone exceptionally good at booking appointments was taking calls in the call center at that time. If certain day/time windows are inherently better for appointment booking, the client can shift the CSRs that are better at booking appointments to other day/time windows.
- Greater transaction sizes during the day on Saturday in Market A can be explored further with respect to whether the transaction size is tied to particular technicians that work during that day/time window or whether the greater transaction size is inherent to the day/time window itself. A method of exploring this would be to analyze the overall performance of technicians that are on duty in that day/time window relative to other technicians.
Over the years of working with ARS®/Rescue Rooter®, we have repeatedly demonstrated an ability to leverage our business and analytics capabilities to continuously improve performance and provide guidance to the client. And again in this case, our analysis drove insight and opportunity on two fronts. First, we provided operations insight and optimization opportunities to the client in the management of their call center and allocation of sales and technician resources. And second, we were able to strategically redistribute campaign funds to drive ROI up to 32.9% higher.