Publish Like a Gramp

Your Grandpa is probably good at many things – mine was good at carpentry and gardening – but I’ll bet you 0.0001 Bitcoin that they both aren’t very good at web publishing.  Here are some terms you’ll still hear today from web publishing companies that originated in your Grandpa’s era, and what you should be using instead.

Page Load Time

Sometime after the end of the Cold War, analytics companies, including Google Analytics, started recording and reporting on Page Load Time.  Today, March 23, 2021, it’s still the default view when you click on Site Speed in Google Analytics – even though it utterly useless.

Page Load Time is the total time it takes to load the entire page and its contents, including all resources on the page such as scripts, images, CSS, etc. While this might sound like a great way to measure site speed, I actually can’t think of a worse way.

Below I describe two scenarios on the same page. Let’s imagine that this is a simple HTML page with a widget lower on the page, off the screen. It takes 1 second to load the page and 1 second to load the widget. 

Scenario A: I load the HTML page normally, so it needs to load everything, including the widget before it shows the content.  The content renders to the user after 2 seconds and the page load time is also 2 seconds. Easy.

Scenario B: Same page, but this time since I know the widget isn’t visible, I’ll load it later. The page loads and displays the content to the user after 1 second, twice as fast as Scenario A –  Great Result!  The page then waits for the render to finish and loads up the widget after a second pause, so it’s ready if the user scrolls down.  The Page Load Time is now 3 seconds.  Recap – the user sees the content in half the time, but the Page Load Time is 50% longer.

You might be thinking that the above is some strange edge case, but it isn’t.  With the rise of lazy-loading, this is currently happening on nearly every modern website.  If it isn’t happening on your site, you’re probably behind.

To this day, I see publishers, and worse, companies that are supposed to be helping publishers focus on this archaic metric. If you’re talking to an “expert” and they mention page load time, run, don’t walk, to the Zoom end call button. (or should I say hang up the phone since we are speaking in outdated terms)

What should you use instead of Page Load Time?  The Core Web Vitals.  The Core Web Vitals consist of 3 measurements:

Largest Contentful Paint (LCP) – How long does it take for the user to see the primary content on a page.

Cumulative Layout Shift (CLS) – Does the content on a page stay still, or does it shift around like a recipe site?

First Input Delay (FID) – Does the page respond to user input quickly.

The Core Web Vitals are fantastic measurements of speed-related user experience.  Drop the pagespeed scores, drop the page load times and focus on these and your users, and Google, will thank you.

CPM / RPM

1995 brought us two things that continue to stick around even though they probably shouldn’t – Boyz II Men and CPM.

CPM (Cost Per M (Roman Numeral for 1,000)) –- is a pricing unit used for the purchase of online advertising.  More specifically, it is the price paid for the display of 1,000 ads.  A CPM of $5.00 means you’ll pay $5.00 for every 1,000 times your ad is shown.  RPM is the publisher version of CPM and is just a measurement for how much a publisher gets paid per 1,000 ads shown on their site.

Earnings Per 1,000 Visits (EPMV) is a measurement of revenue earned from every 1,000 visits to a website. It’s also sometimes referred to as session revenue.

While CPM continues to be a valuable pricing unit for the purchase of advertising, it is a terrible metric for web publishers because it only measures one part of a much more significant interaction with a user.  To explain what I mean, let’s think of a coffee shop or any retail business.

When customers visit a coffee shop, they will likely purchase a coffee; however, they may also buy a donut or perhaps one of those delicious round cakes on a stick.  RPM, the measurement of the revenue from a single ad, is the equivalent to the price of coffee.  EPMV, the measurement of the income from a visit, is equivalent to the total amount of money spent by a customer during their visit to the coffee shop (coffee, donut, cake stick).

Let’s imagine that coffee was $2.00, but you increase the price to $3.00 – your CPM has gone up!  However, with the increased cost of coffee, more customers decline to purchase additional items. Your EPMV has fallen, along with your revenue.  Ask any store owner, and I’ll guarantee you that they would like to maximize profit per customer, not yield per item.  Publishers should be no different.

Ok, that’s for a coffee shop, but what about a website?  Well, just like the number of items a customer purchases, the number of ads a visitor sees on your website is not fixed.  You can add more ads or remove ads.  You can change the size. You can move them around.

	Scenario A (one ad on the page)	Scenario B (two ads on the page)
Top Ad CPM	$10.00	$10.00
Bottom Ad CPM	N/A	$2.00
AVG CPM	$10.00	$6.00
Page Revenue	$10.00	$12.00

Things get even more complicated when thinking about multiple page views in a visit.  Putting more ads on a page might cause visitors to leave your site, reducing the number of ads they see on that visit.  Decreased ads viewed can again result in an increase in CPM but lower revenue. 

When a metric fails to measure the result you’re looking for, it is a lousy metric.  If you’re talking to an expert and they are telling you how great their CPMs are, this is a sure-fire sign that they are stuck in 1995.

So why do these metrics continue to be used?  Honestly, I’m not sure, but I have a couple of theories.  The first is that change is hard.  Ask your Grandpa – sometimes the world passes you by and you fail to adapt.  I certainly think that’s the case for some of these companies. In other instances, I believe that they continue to use them out of convenience.  It is hard to optimize for an entire user visit or maximize the purchase value in a store. It’s much easier to focus on RPMs or the price of a cup of coffee.  These companies take 20% of publishers’ revenue no matter what, so they try to concentrate you on RPM, which is much easier to game.

P.S. Grandpas are awesome. I hope to be one, one day.