Thursday, September 6, 2018

Marketing Analytics: Why I Obsess Over Rates

I don't mean interest rates on my savings account, which is still paltry BTW. =)

I see marketing analytics reports all the time from clients that focus on absolute numbers like visits, conversions, clicks, opens. These are good to know, but I find these metrics meaningless in a vacuum. Is 10,000 clicks in a month good? It depends...

Instead I like to focus on relative metrics like calculated rates and ratios.

Rates are great to track metrics over time, to offer context, and to compare to industry benchmarks because rates are normalized. This means you can compare rates over different time periods to see if something is wrong or if something is doing surprisingly well all of a sudden. This also allows a smaller company to compare themselves to a larger company, generally speaking.

Let's look at how to use rates in common digital marketing reports to reflect the most insights.

Website Analytics

Example 1: Conversion Rates. Every website needs to have a primary goal or "conversion" event that is clearly and cleanly tagged on the website. That can be an e-commerce transaction or lead gen form completion. Everyone always then tracks the number of "conversions" each week or month in absolute terms.

Most, but not all, clients look at conversion rates. Yes! But there are multiple ways to define "conversion rate." Some define it as conversions / site visits. That's not bad and Google Analytics offers this in their standard reporting. The problem with just looking at this "start to finish" conversion rate is it does not help you identify where the leak in the conversion funnel is. Furthermore, this is often a very low number like 0.05% so it's really hard for senior management to fully understand such a metric.

What every marketer should do is also break down the customer journey into smaller bite-size paths. In the illustration below for an online lead gen site, by simply breaking the customer journey in half to calculate a Consideration Rate and a Completion Rate, you now have more actionable insights.

Purchase Funnel Rates
This helps identify where you have leakage in the funnel and how to address it. For example, if form completion rate is low, maybe your form is too long or confusing. Also, this completion rate should be easier to digest because it usually ranges between 5%-50%. So if you have a completion rate of 10%, you can ask yourself does that feel low when 9 out of 10 people bailed. Much easier to comprehend than the 0.05% conversion rate example above.

This form completion rate is also ideal for industry benchmarking because the "scope" of your metric is just a form and form completions. If you used the 0.05% conversion rate, the scope of that metric is all visits to your site for whatever reason and all the pages on your site. So it's really hard to compare apples to apples on such a broad conversion rate. Take retailers who live and die by shopping cart abandonment rate, which is the reverse of completion rate. Industry-wide cart abandonment rates are relatively easy to find and benchmark to your own cart abandonment rates.

Before moving on from this example, I wanted to mention the Consideration Rate. This is worth monitoring because if it's low, it means users are not showing any interest in your product. So you have to ask yourself questions like "Is our CTA not prominent enough?" or "Are people really not interested in our product offering?" The former can be a user experience issue, while the latter is a product or product marketing issue.

Example 2: % of Totals. This is a very simple ratio and calculation, great for providing a frame of reference. Take the classic Top Pages by page views report.

% of Total Pageviews Example

Most clients will quickly grab this data from their web analytics tool that shows pageviews in rank order, but sometimes discard the % that is often in the dashboard. I like to also provide the third column that is % of Total Pageviews because then you can better appreciate how popular a page is relative to other pages on your site. It's usually harder for senior executives to understand if 72K pageviews for the Product Overview page is low or high, but 25% of all pageviews is easy to grasp and to conclude it's very high for one page to garner.

I calculate this % of Total for a lot of site metrics beyond page views, such as % of Total Video Plays and % of Total Downloads. Again, the goal is to show if activity is concentrated among a handful of content or assets or more fragmented and distributed across all of them. If it's more concentrated, figure out what makes them so popular and do more of that!

Example 3. CTR. Clickthrough rates (CTR) are often associated with paid media ads and email links. But I like to calculate CTR for website analysis when analyzing CTA button clicks or on-page analysis to see what users chose to click on when presented with a host of options. Here's an example from a directory search results page. CTR is calculated for each link based on link click / page views. This is also a more visually appealing way to show these stats than a standard table. I also like to develop a similar slide for CTRs on the global navigation menu. One critical tip: usually you must setup custom click tags to calculate the CTRs I've discussed.

CTR Website Example
In terms of benchmarking, CTRs are great because you can compare pre- and post-launch CTRs if you redesign your website.

Email Performance

Most clients already focus on open rates and CTR. 👍 But I still see reports like this occasionally:

Email Report Without Open Rate Example
For most marketers, the opportunity here is for benchmarking rates. Epsilon publishes a great quarterly report that is free that is full of email benchmarks by industry and by types of emails (e.g., editorial/ newsletter, marketing, service)!

Last comment on email metrics. Don't forget to look at CTOR (Click-to-open rate), which is clicks / opens. CTOR is different than CTR, which is clicks / delivered. Monitoring and benchmarking CTOR can help you address issues with the content or CTA within the email body and usually rules out an issue related to the subject line.

Social Media Campaigns

Engagement Rate Example. For social media campaigns, paid or organic, one of the most popular primary KPIs is engagement rate. Most marketers define engagement rate (a.k.a. interaction rate) as (reactions + comments + shares) / posts. But if I'm a marketer, what constitutes a good engagement rate? A social media analytics tool, such as Quintly, is great for answering this question. It not only allows you to compare yourself to other competitors, but it has also created industry averages. In the example below, let's say I'm BMW USA. I can compare not only my Interaction Rate to direct competitors Audi and Mercedes, but Quintly also shows the average Interaction Rate for the Top 10 US Auto manufacturers. Quintly has other useful social media benchmarking tips on a recent blog post so I won't go into any more details here.

Quintly Interaction Rate Example

Paid Media Campaigns

Example 1: SEM & Display. Almost all media managers report out on CTR, CPC, and CPM for paid search and display campaigns. Yay! The suggestion I have here is be sure you ask your media partner or publisher who you are working with for industry benchmarks on these common media metrics. For example, your Google rep will often put together a quarterly report showing you your SEM spend, CTR, etc., relative to other advertisers in your category or industry (see example below). They will usually slice this by brand vs. non-brand and desktop vs. mobile. One caveat with these Google benchmarking reports. It will almost always show you are under-spending and under-performing because you are being compared to a Category Leader Average not the entire Category Average. Google obviously has an incentive to motivate you to spend more 😉

Google SEM Benchmark Example

Similarly, for display campaigns, find out what's the industry CTR from your vendor. If you're running rich media unit ads, find out what's a good engagement rate for user interaction with your unit in your industry.

Most digital advertisers also look at efficiency using Cost Per Acquisition (a.k.a., Cost per Conversion or Cost Per Lead). I also like to look at Click-to-Conversion rates.

Example 2: Video. For video ads, you need to look beyond video starts or plays. For sure, look at video completion rates. Or even better, look at milestones like quartiles. Below is an example of a Video Completion Rate Funnel with quartile milestones. This helps you identify where people are dropping off. For example, in pharmaceutical videos, users often drop off when the super long Important Safety Information (ISI) begins to play so I often place a milestone marker at the ISI start.

Video Completion Rate Funnel Example

For benchmarking purposes, ask your media partner or publisher for average completion rates for your industry. For example, they should be able to tell you the average completion rate for a :30s video is 30% in the financial services industry.


So there you have it. Now you know why and how I am obsessed about Rates and Ratios for my clients' marketing analytics reporting. All of the above are very simple concepts. You just need to spend a little bit of extra time in developing your measurement and tagging strategy and also on reporting. But in return, you become awesome at storytelling with data!


Disclosure: I am a strategic advisor to Quintly.

No comments:

Post a Comment