My (tentative) plan for building readership

This is the sort of introductory post that should really have been at the beginning, but when I began this I was in the middle of figuring out how to utilize Google’s Accelerated Mobile Pages.

I’m Larry Felton Johnson. I use my middle name to differentiate myself from the many thousands of other Larry Johnsons, including the former NBA star, and a county commissioner here in metro Atlanta.

I’m the editor and publisher of the Cobb County Courier, and a member of LION Publishers.

I started the Courier, an online-only independent news site, to cover local news in Cobb County, Georgia, adjacent to the City of Atlanta to the north and west.

For the first three years I was in a building phase, and while I had dedicated readership, it was very small. For the first year 200 readers was a good day, in years two and three readership would fluctuate between 200 and 1,200 depending on whether any specific stories took off.

In January of this year I changed hosting platforms to help increase the speed of my page loads, started fanatically bearing down on SEO practices, and boosted our daily page view expectation to between 1,500 and 3,000, with spikes up to around 10,000.

Since most of that increase was due to a rise in Google page rank, and most of it was being sent to my AMP pages, it created an interesting dilemma I described in an earlier post.

So what’s the plan?

I’m still the only full-time staffer at the Courier. While I have a good team of freelance reporters, they are just that: reporters.

So I have to edit, report as needed, sell ads, and take care of all the social media and website development.

Doing it has given me a good idea of the things I need to do.

First let’s examine where my traffic comes from.

Facebook

For the first three years the overwhelming majority of my traffic came from shared content on Facebook. Of all social media platforms I deal with, Facebook is the most hassle for the least benefit at this point, though.

Facebook will change its algorithms and yank the rug out from under small publishers in a hearbeat. They’ll set up programs that require publishers to jump through multiple hoops, and dismantle them and set up something new within a few months.

At this point I have no idea what the best practices for working with Facebook are, and since Google accounts for over 70 percent of our traffic now, I have to apply Pareto’s 80/20 Principle, and put my immediate effort into making use of the Google traffic.

Google

Google, particularly via the AMP traffic, is the main source of traffic to my site, and I have to take it the most seriously.

But I can’t afford to rely on Google to be a consistent and stable partner any more than I can Facebook. Google commands a huge share of the ad revenue, and the amount they pass along to publishers through their Adsense program is pathetic.

So what does that mean for me?

I have no control over Facebook, I have no control over Google. The thing I do have control over is the Courier’s periodic newsletter.

At the moment the newsletter is sent weekly, and has fewer than a thousand subscribers, even though its free.

So in keeping with Pareto’s rule, I need to improve the newsletter to make it valuable to the subscribers, and increase the subscriber base by at least a factor of 10. In a county of 760,000 people that’s not a far-fetched goal.

So I’ve been working on capturing newsletter sign-ups via the AMP articles as a first step. It took me a few weeks to figure out how to do it, and now I need to refine the pitch I’m presenting on the AMP in-article pitch.

Even with my Facebook presence I need to put the remaining 20 percent of effort guided by the Pareto Principle into funnelling readers into the newsletter.

So my plan is to build the newsletter using mostly Google, but also Facebook,

So I need to:

  • Make sure that first-time visitors to my site always see the newsletter signup pitch
  • Make sure there’s compelling material in the newsletter to entice readers to visit the Courier site
  • Spend most of the effort getting the AMP signup form right
  • But also not totally drop my efforts to encourage Facebook users to sign up for the newsletter

Of course there’s a lot more to the strategy than that, much of it geared toward creating a regular loyal readership. But that is for future posts.

The need for a balance of traffic sources

As the Editor and Publisher of a small local news sites I wear a lot of hats.

The two that create the most conflicts are my hat as editor and my hat as ad salesperson.

Until I can hire a salesperson this is going to continue, and I’d like to focus on one of the stress points of these two roles.

As an editor and reporter, I was trained to evaluate the newsworthiness of stories by the following rote list of criteria: Impact, conflict, novelty, prominence, proximity, and timeliness.

Not all the criteria are met with every story. The “man bites dog” stories have novelty, but usually don’t have a great deal of impact (except for the dog, who’s unlikely to become a regular reader).

But as an ad salesperson, another thing I need to focus on is the channel/source/medium with which the reader arrives.

I explained a little about this here and here.

Some ads are highly visible on desktop computers (sidebar ads) but do poorly on mobile devices and are invisible on AMP pages.

Other ads will do well if a lot of readers visit the home page, but not if most readers go straight to articles via shared links.

What I did at 4 a.m. today

I know doing ad zone analysis at 4 a.m. on a Sunday seems pathetic, but hey!, that’s my life.

So this morning I downloaded a list of my available ad zones from the ad server company I use, Broadstreet.

My goal was to get a clear idea of what having my ad inventory filled would look like by plugging estimated monetary values into each available zone on a spreadsheet, and from there calculating how much I need to charge to arrive where I want to be in terms of profitability.

One thing became clear.

To credibly sell every ad zone for enough money to achieve the sort of growth I aim for, I have to diversify my sources of readership.

The enormous jump in readership I’ve gotten from the increase in Google AMP traffic is fantastic, and looks great on my media kit. But I need to be able to deliver results for advertisers who purchase on the home page, or category pages, or sidebars, or footers …. well, you get the idea.

So how do I accomplish this?

So far I’m just in the brainstorming phase of a solution, but a few things jump out at me:

  • I need to increase my newsletter readership, and I probably need to increase publication frequency from weekly to at least three times per week. I only have 650 newsletter subscribers, which for a primary site that gets over 100,000 page views per month is very low. What newsletters do is allow me to control the traffic to some extent. Readers are arriving, not to an AMP page, but to the traditional pages via direct link.
  • I need to refocus on traffic from social media, for the same reason as newsletter traffic. Also, If I increase social media traffic without decreasing the search engine traffic, my media kit looks even better.
  • I need to provide some content that older readership, who are more likely to arrive via desktop computer, will read. Local history is an example of that.
  • I need to directly appeal to readers to visit the home page often.

I’ll write more on this as I try things out over the next couple of weeks. I think the first step is doing a real campaign on newsletter signups, followed by trying to make my social media strategy more coherent.

AMP ad sales explained another way

To some extent I’ve struggled to explain, even to other publishers, why it’s imperative for me to master Accelerated Mobile Pages and AMP ads.

The short answer is that I have no choice unless I want to slow down my page load speed and trigger a drop in traffic back to the modest levels I had prior to my sharp increase after April.

As far as I’m concerned, AMP is a big part of the Cobb County Courier‘s future, whether I want it to be or not.

So let me walk you through an exchange with a typical advertiser of the moment, a candidate for political office.

It’s a good example, because candidates do not tend to know much about how ad placement works, the difference between a rotating ad and an exclusive ad, and the difference between viewing the page on a desktop computer (the typical way a candidate tries to look at their ad) and a mobile device (the way 75 percent of readers come to our site).

A candidate for office (whom I’ll call “Candidate 1”) grabbed my prime spot (leaderboard) early with an exclusive buy. The leaderboard, and its larger sibling the billboard, is seen everywhere, on different devices and on every page (home page, category page, utility pages, etc.) Every place that is, except AMP pages, which have their own separate environment.

Since that spot was not available, the very next best slot is the top position on “inside story” ads. The reason it’s the next best is that it’s seen on both mobile devices and desk top computers, whereas a sidebar ad gets shoved so far down the page that the reader never really scrolls to it on mobile devices.

So Candidate 2 bought the top inside story position.

After that, Candidate 3 comes along and can’t get either the leaderboard or the top inside story slot, so she buys the second inside story zone, plus the top sidebar ad.

Candidate 2 visits one of my pages from a desk top computer, and notices that the sidebar for Candidate 3, who is her opponent in the election, is further up the page.

She phones me and says “I thought you told me the ad I bought was the next best position. I went to a page and saw (Candidate 3)’s ad first.”

I tried to explain that while Candidate 3’s ad is further up the page on a desktop computer, the ad doesn’t even show up at all during a session for most users, who show up via mobile device.

We worked out a deal whereby Candidate 2 bought the second sidebar slot, and she could see both ads on her desktop computer.

But let’s make it even more complicated by introducing AMP into the picture.

Whether the advertiser is a candidate or the owner or officer in a business, they want to be able to see their ad.

Most advertisers attempt to see their ad using a desktop or laptop computer. This makes it easy during a sales call, but as I explained above, if they visit the site on their mobile device, I’m likely to get a call or email stating “MY AD ISN’T RUNNING!”

AMP ads create one more layer of possible failure to see the ad, because nearly all my AMP traffic arrives via organic search (usually Google, but also Bing, Yahoo and Duck a Go Go).

So to see the ad, the advertiser has to arrive at the site the way most users do now, which is via the results list of a search.

At the moment my solution has been to send the prospective advertiser a direct link to one of my AMP pages and tell them to open it with their mobile phone, along with an explanation of why that’s the way to see the ad.

But I still anticipate complaints that the ad isn’t running when they spot check the non-AMP version of the page.

The big picture here is still that AMP pages dominate my traffic, so the advertiser gets the maximum number of eyeballs on their ad if they buy one.

But as I stated at the top, this is even hard to explain to other publishers, much less advertisers, who have a wide range of both knowledge of how ad placement works, and general technological knowledge.

AMP and the ad sales dilemma

This post is a modified and expanded version of one I posted to a social media forum dedicated to small publishing ad sales.

For a small publisher, the complexity that comes with a sudden increase in traffic is a good problem to have in some ways, but it’s a problem nonetheless.

In the case of Google’s Accelerated Mobile Pages (AMP) the problem comes from two directions.

The first is suddenly having two completely different advertising environments to manage, with different formats and best practices.

The second is explaining this to advertisers in a clear way.

As I mentioned in a previous post, I’ve had an order-of-magnitude jump in traffic due to changes I made that increased page load speed in late April and early May.

Google loves page speed, and rewards it with higher ranking.

Before then, 20,000 page views was a good month for me.

Every month since May, inclusive, I’ve exceeded 112,000.

But since Google prioritizes AMP pages in its ranking system when if finds them, a very large portion of that jump goes to my Accelerated Mobile Pages, which have their own ad slots, with heavy restrictions on design due to page speed optimizations (for example, the size has to be known in advance so that AMP can set the framework before content is loaded, so no dynamically-sized ads are allowed).

Here is a 30-day period I examined with Google Analytics a few days ago.

Total Pageviews: 112,524

AMP pageviews: 64,376

non-AMP pageviews: 48,148

In a media kit that total number looks great compared to where I was before April, and even traditional access to my site has more than doubled.

But note that 64,376 of those page views generate no impressions for traditional advertisers (whose ads are on the non-AMP pages) whether the ads are well situated for mobile (inside story, leaderboard) or not (sidebar ads).

I made my first attempt to directly sell AMP ads to a customer yesterday, and got an email from him this morning in which it’s clear he thought I was pitching him Google ads, rather than AMP ads targeting a local audience.

The current state of my AMP ad zones isn’t all downside.

I had Broadstreet, the ad service company I use, set up two AMP zones and for the time being rotate Google Adsense ads into those zones.

My revenue from the Adsense ads in the AMP zones caused my total revenue from Adsense ads to increase from about $30 per month (paid every quarter or so), to $350 during the last 1-month period, and a likely $250 for this month (paid every month since both those figures exceed the $100 payment threshold).

That isn’t enough to float the site but is enough to supplement my budget to hire freelance reporters.

This is a good problem to have in some ways, but it’s a problem nonetheless.

My goal is to figure out best practices for AMP ads that are effective for customers, and to learn how to clearly explain the benefits of the ads to customers.

I feel like the canary in the coal mine on this one in some ways. I haven’t received any indication that other publishers are experiencing the lopsided traffic to AMP ads I’m experiencing, judging by the feedback I’ve gotten so far.

Why I’m so focused on AMP

When I first added AMP pages to the Cobb County Courier it was just because someone advised me to do it. At the time I knew Accelerated Mobile Pages had something to do with site access via smartphones and other small-screen devices, but I was still oriented to desktop and laptop computers.

But I installed an AMP plugin and let it construct AMP pages from the content I posted, and didn’t think much more about it until early May of this year.

As I continue my education on AMP (and at the moment I’m reading everything I can get my hands on) I’ll post a series of “What is AMP and how does it work?” articles.

But for now I want to demostrate to you why it is critical to me at this point.

After May, when my project to improve the Courier’s Search Engine Optimization began showing results, and my web traffic began skyrocketing, I noticed that a lot of the new traffic was to my AMP pages.

I monitored it informally, but yesterday I set up a dashboard in Google Analytics to track traffic to AMP pages and non-AMP pages over different time ranges.

I included in the dashboard a table listing all the AMP pages by URL, along with the number of pageviews for each page, and another table with the non-AMP pages with pageviews.

To get summary counts I set up scorecards with total AMP, total non-AMP, and total pageviews of any type counted.

The tables by article urls will allow me to figure out what type of articles are likelier to get traffic to the AMP pages and which do better with traditional pages.

Here are the totals as of yesterday when I set the range to the past 30 days:

Total Pageviews: 112,524

AMP pageviews: 64,376

non-AMP pageviews: 48,148

(The first time I accessed the dashboard the total count of pageviews was off by one, but that seemed to correct on the second access.

This gives a clear picture of why I need to figure out how to make the best use of the AMP traffic.

The proportions are staggering to me, and make several things imperative.

First I need to grasp AMP advertising so that I can sell advertisers on local impressions, rather than intense feature-rich ads.

Second I have to figure out a way to convert readers who come to the site via Google search to regular readers who frequently access the site.

I plan on adding a sticky signup form to my AMP pages as soon as I figure out how to do it. That will be a first step in trying to build reader loyalty.

Finally, even though Google Page Speed Insights gives me higher marks for the AMP pages, I’m seldom out of the mid-range orange-coded numbers. I need to figure out how to be consistently in the green.

Introducing Publishers Notebook

My name is Larry Johnson, and I’m the editor and publisher of the Cobb County Courier, a local independent news site in Cobb County Georgia.

I’m starting this blog (and I have no illusions that it’s anything other than a blog) to flesh out techniques and policies in my role as publisher.

You won’t find a lot of graphics here, except for statistical charts, graphs and tables.

Another reason I started it is that I tend to get obsessive about projects. I’m a member of LION Publishers, and have posted frequently about this on their Facebook group. I began worrying that I risked seeming to spam the group, so I moved the discussion to this blog, where people can visit as they choose.

The specific event that led to this blog was that I monitor Google’s news tab carefully, and the Courier has been pretty successful in ranking pretty high when we keep our content fresh.

But in about March I began noticing that the Patch chain of sites, which has made Search Engine Optimization its business model, often managed to beat not only the Courier, but the much larger Atlanta Journal-Constitution and Marietta Daily Journal in the page rankings, often with absurdly weak content.

So I started a project to, as I jokingly called it, “out-Patch the Patch.”

I began really bearing down on Search Engine Optimization.

I made heavy use of Yoast to improve my rankings, started checking my site against Google Page Speed Insights frequently (I’m still not where I want to be in that respect) installed Hummingbird and Smush, and finally, when I realized my hosting platform was just innately weak, switched platforms.

The results exceeded my highest expectations.

I went from 20,000 pageviews in a good month, to regularly exceeding 100,000, and this improvement was almost immediate and has lasted for three solid months so far.

Here’s a screenshot of Google Analytics comparison of the past 30 days (the blue lines) versus the same period last year (the orange lines).

So what does that have to do with this blog?

I find it best to diagnose things by writing them out, and writing them out in public gives me the opportunity to get feedback from readers.

And I really need to systematize the approach I’ve been using, really understand which changes led to the results, and build on it.

In particular, and I’ll deal with this more in the next post, I need to understand why Google’s Accelerated Mobile Pages (AMP) is now more than 50 percent of my total traffic, and how to take advantage of that enormous increase.

The AMP increase has a lot of implications for how I sell advertisements, how much attention I pay to social media promotion (since Google is funnelling a large majority of my traffic to me) and how to transform readers who came to me through a web search into regular readers.

So my next post is going to think out loud about AMP, a subject I’m just beginning to wrap my head around.