We Turned an Idea Into a SaaS Product for Content Marketers—Here’s the Whole Story
Three weeks ago we launched our first software product, Revive. For the uninitiated, this is a free tool that finds content decay on your site and recommends articles to refresh. You can try it out here.
This is a significant shift for Animalz. We’ve been laser-focused on becoming the high-quality content marketing agency for SaaS. But as we’ve grown, we’ve learned a lot about content from working with the best SaaS businesses—everything from workflow challenges to writing prompts to a litany of data hurdles. We want to share what we’ve learned so that anyone working in content marketing can benefit, not just the customers we work with directly. Software is the most scalable way to do this, hence our recent evolution from ‘services’ to ‘software + services.’
One Content Marketer’s Journey from Neuroscience to Writing to Software
For me personally, this is a shift back towards my roots. I’ve worked as a content marketing writer for over five years. Before that, I was a neuroscientist for fifteen years. How does one go from neuroscientist → content marketer → software developer/product manager?
I went to university to study neuroscience, I went to grad school to study neuroscience, and I went to Switzerland and then came to the US to practice neuroscience. Along the way, I worked on such eclectic subjects as how sheep recognize other sheep and Elon Musk’s new favorite subject, brain-machine interfaces.
I was okay at scientific research, but never excelled. What I did do well was write: grant applications, manuscript editing, proposals, press releases, presentations. If something needed writing in the lab, it landed on my desk.
Along the way, I also met my wife. When my research visa was up, we decided to stay in the US. Making a living through writing seemed an ideal option, but like most nascent writing careers, mine didn’t start ideally. I did some straight-down-the-line, terribly-paid freelancing on Upwork and the like. I wrote SEO-fodder for a pittance and got ghosted and ripped off.
But Upwork led to writing for Pocketbook (an Aussie version of Mint), and the first time I got to combine the analytical skills I built up as a scientist with content marketing. Later, the Head of Growth at Pocketbook left to join Canva and took me along as a freelancer. I wrote one of Canva’s first big hits. This article was part of the resume I sent to a sketchy Craigslist ad for writers in New York. That ad turned out to be from our founder Walter Chen, who was starting a new content agency. I started at Animalz, alongside Walter and Paige Picard, on Day One.
I’ve written hundreds of articles for dozens of customers in these four years. I still enjoy writing, but as I’ve gotten to understand content better, I’ve also wanted to bring a more analytical perspective to a field that badly needs it. Can we scale what we’ve learned working with customers and reuse my scientific skills to dive deeper into how and why content works? We want to try, and our first step in that direction is Revive.
What We’ve Learned Launching a Product for Content Marketers
It is still very early in the product life of Revive. We want to continue to make it more and more useful for content marketers. But even in these first few weeks, we’ve started learning more about the very specific challenges content marketers face in their day-to-day work.
1. Traffic Decay Is a Problem
Traffic decay happens to all our customers, even the ones with millions of views on their blogs. It happened to most of the people who signed up with Revive.
So far, we’ve found 40,514 potential refreshes from 923 sites. That averages 43-ish refreshes per site. Of course, this is skewed towards the bigger sites that signed up. A few sites had tens of thousands of articles on their sites with thousands of refreshes found. But the tail is long. Hundreds of sites had between 1 and 10 refreshes needed.
The scale of the problem is shown if we look at the total number of pageviews lost to decay: 52,689,428. That is a lot of views, ~1,300 per article that people aren’t getting. For media companies, that represents a lot of missed ad revenue. For B2B companies, it represents a lot of lost conversions. For other businesses, it is lost sales. For all, it is lost revenue.
It is wrong to say this is an epidemic—it is a natural part of investing in content for organic search. The dynamic nature of content ensures that your articles will lose traffic over time if the world moves on but they stay static. But this is why we chose to start with a refresh tool—we know from our customers that refreshing these articles is a low-lift way to capture those views back. Those 52 million views are much easier to get back through refreshing than through completely new content.
Traffic Decay Isn’t Just a B2B Problem
We built Revive for the kind of companies we work with—B2B SaaS. Though they make up the largest individual cohort of signups, they didn’t even make up a quarter of total signups:
Who were the other users and how could traffic decay impact them?:
- Other businesses. Mostly e-commerce, these are people who wanted to look at the traffic to product pages instead of content the way we think about it. If organic traffic to a product is dropping, that product isn’t going to sell as much.
- Other agencies and content writers. Good to know, this is useful! A number of other content agencies signed up to run their customers’ analytics through Revive. Like us, they want to make sure their customers are growing traffic and see refreshing existing content as an excellent option for doing so.
- Travel sites. A lot of travel sites used Revive. There were always refreshes to find but, these sites, run by individuals, had consistently good traffic. Likely supported by advertising or sponsorship, they want to know which articles need more attention to increase revenue.
- Media companies. With ad revenue and sponsored posts being vital to success, knowing which articles and topics are performing well—and which aren’t and could be updated to do better—is critical.
- Individuals. If you have a site to build your personal brand, knowing which content isn’t performing is as helpful as knowing which is, in terms of understanding your audience and what they want to know about.
We built Revive around B2B SaaS, but the concept is far-reaching. Like any product, it can’t be all things to all people, but seeing different use-cases is helping us further understand the how and why of content.
Highlighting Traffic Decay Is Good, Knowing Why It Happens Is Better
If there is one feature requested by users of Revive so far, it is this: why is this happening? Here are two examples of feedback we received:
“Would be useful to have additional information like how the article dropped in rankings for a particular keyword or whether search traffic was down for that related keywords. Data to help make sense of the drop in rankings/traffic.”
“The results should show whether the traffic increase is due to a loss in rankings or due to overall search volume decrease. The next steps for optimization are very different, depending on which of these is the main cause.”
Is it because the article is too old? Is it because something better has been published? Is it a seasonal aberration? Is it because the search volume for the keyword has declined? Revive doesn’t answer these questions, but it is the natural question to ask when you see an article is declining in views: why?
And, just as importantly, what do I do next?
Future Analytics Need to Be Opinionated
Google Analytics is an amazing tool. It gives you an incredible amount of detailed information about what is going on with your site, for free (or for $150,000, if you don’t want sampled data).
But it lacks a crucial concept—an opinion. It gives you the data and allows you to slice and dice as you want, but once you get even a modicum of traffic, knowing how to look at your data or what to look at to understand your content becomes unclear. With a surfeit of data available to content marketers, just presenting that data is no longer really good enough; they need to know what that data means.
Revive is a step in this direction. We want to highlight critical parts of your data easily, instead of having you trawl through thousands of URLs in Google Analytics. It has an opinion: “Refreshing content is important. It is an easy way to win. These are the articles that need it. This is how you do it.”
This opinion comes from the lessons we’ve learned in our agency. As we know, and some of our feedback has stated, the next step is to go further. We need to tell you why it’s happening and exactly what you should do on an article-by-article basis.
This is the cycle we want to continue at Animalz. Learn something interesting from working closely with our customers. Help them on an individual level. Scale the idea (using software or content) to benefit others. Learn more.
Product Management Is a Lot Like Content—Here Are a Few Takeaways
I’ve learned a ton about product while building Revive, but also learned lessons that will help me write better content as well. Here are a few I’d like to share:
- Good products are only built with feedback. The same is true of content. You might only want eyes on your article once it is polished, but good editing is crucial to a good article. The same goes for product. Only through feedback from the team here at Animalz, and then our customers, did we start to understand what worked and what didn’t about the product. The two examples above are testament to this. Putting something out there in the world gives people the chance to respond. You can then iterate with your audience/users/customers to build something better and never thought of in the beginning.
- In both cases, you have to think audience-first. When you are coding an entire product, it is very easy to get caught up line by line. Should this be a separate function? Will this function execute faster if I remove this line? Getting lost in the minutiae is kinda fun, but you have to remember that you are building this for you; you are building this for others. Users of Revive don’t care about Google Analytics API calls; they care about the results, how easy they are to understand, and whether they know what they are going to do next. The readers of your content don’t care about your clever sentence structure. They care about the results of what you are telling them, how easy your idea is to understand, and what they are going to do next.
- There is more to product than code; there is more to content than writing. Building Revive required not just learning how to develop a product from a software point of view, but also learning product management, product marketing, UX design, customer success, and customer support — all the things I’ve spent four years writing about! Going over the content of our customers was vital to developing the product and getting it to launch. The same is true of content. In 2019, it isn’t just writing. It is understanding customers, managing internal expectations, knowing the right channels for promotion, and being able to see the big picture of why you are writing.
- Don’t only concentrate on the happy path. In product, the happy path is the way you think a user will interact with your product. They will click here and then there and then on that, and, voilà, a great experience. You know the path as you designed it, so this is obvious. Not so to your users. They will click that, then there, then here, and, voilà, a 404 page. The sad path. The Animalz team were experts at finding the sad path through Revive before we launched. But you need to think about the sad path through your content as well. It might be clear to you what you mean, but you already understand your argument. Are you directing the reader down the happy path properly, or leaving them to find the sad path themselves?
How do you go from being a neuroscientist to being a content marketer to being a software developer/product manager? By being eager to learn new things. They can be about brains or about businesses or about bits. One of the reasons I am proud of working at Animalz is that we’re always open to learning. And when we learn new things, we want to share them, whether that is through writing content or through writing code.
The Science Behind 100,000-View Blog Posts
All successful content marketing is successful in the same way.
At least, that was our instinct and we wanted to prove it. We’ve seen many blog posts take off, delivering a steady flow of organic traffic for months and years. We’ve also seen plenty of posts flop, hitting a flatline within a day or two of publishing.
Blog posts tend to follow a few familiar “shapes”—that is, the data plotted in Google Analytics tells the story of the post. You can read these shapes without knowing much about keyword volume or traffic sources. You can probably even tell if the post is going to be a hit or a dud within a few weeks of publishing. But this is all anecdotal. We wanted to find out if the patterns we see are indicative—and, more importantly, predictive—of success. So we looked very closely at a few blog posts that had been viewed more than 100,000 times to see if our observations were backed by data.
The results offer a template for success, but also a new and more scientific way to think about measuring your content.
The 5 Phases of the Content Lifecycle
Thinking about content as either viral or evergreen is limiting. To succeed, a post ideally needs both short-term appeal and long-term engagement. The shape of its performance in Google Analytics tells the story. You’ve likely seen many posts with a spike early on and then….nothing.
We used 100,000 lifetime pageviews on a single post as a benchmark for success, but that certainly isn’t the only way to define it. It’s clear, however, that posts in the 100k club have delivered good value.
Most shared a very similar shape. Here’s an example:
Example #1: An Archetypal Success
This post, like others that are this successful, goes through five distinct phases:
- A spike phase when the post is first published.
- A trough phase where it looks like growth is stagnant.
- A growth phase as pageviews increase over a few months.
- A plateau phase as growth levels out.
- A decay phase as traffic to the post starts to drop off.
The initial spike skews the long-tail view. This post delivered the huge majority of its pageviews well after the spike phase died off. (AdEspresso has published an impressive 13 posts that have earned 100,000 pageviews or more.)
Here is another 100,000+ view example with the same growth shape from Appcues. The spike phase was less dramatic, which makes the growth phase more obvious:
Example #2: A Long-Tail Success
Appcues, The 5 Best User Onboarding Examples
Let’s take a closer look at example #2. When we break down the source of the traffic to this post, we can see what is driving each phase:
The initial spike is driven by community links (shown in orange) like GrowthHackers, Inbound, Designer News, Reddit, The Hacker News, etc. There is also some traffic coming through email (green) and direct traffic (red). A few weeks later, there is a second bump, this time with more traffic coming from email and from social (teal) as a result of a second round of promotion on those channels.
The overall traffic then flatlines for weeks. The time frame between the second email-driven spike and the start of the growth phase is over three months. During this time, overall views bumble along at ~100 per week. It looks like the post is going nowhere.
But then the next phase kicks in—the growth phase. Growth over time comes from organic search traffic. For this particular post, organic traffic really comes into its own after the Week 25 mark. It is at that point that organic search traffic becomes a significant component of all traffic to this post.
In approximately the first half-year this post was live, most traffic comes from direct views, email, and referrals. But in that second half, organic takes over, growing as a proportion of all traffic to the post and growing the absolute traffic to the post.
We can also see that in the trough phase, from approximately Week 12 to Week 25, organic traffic was growing as a proportion of all traffic. The headline numbers were flat, but organic was already starting to pick up the slack from smaller direct, email, community and social traffic.
At its peak, organic represents 87.4% of all traffic to this post. For growth, it is not just the absolute numbers that determine success, it is also the ratio of organic/other traffic.
At the one-year mark, the post transitions into the fourth phase—the plateau. At this point, the post has reached #1 for the keywords related to the post, maxing out the volume of those search terms. It happily stays at this point for over a year before traffic starts to wane.
If we look at the proportion of traffic in each phase for this post, we get:
About 97% of all traffic comes after that initial spike. An incredible 90% comes after the post has already been live for six months.
The fifth phase—decline—is clearly seen in this post. This happens when newer content supersedes this post and starts to cut into the organic traffic to the post, causing a decline in overall traffic, or when search volume for the target keywords declines.
Let’s look at two more examples of the same shape, both from 100,000+ view posts. The first was published in 2015:
Example #3: A Slow Decline
The initial spike, growth, and plateau are all visible. In this case, the article has declined slowly over the past two years, though it still pulls in a respectable ~200 views a week to the blog. Given the post’s enduring value, it’s ripe for a content refresh. With a few changes to the title and some of the content, the decline could be stopped or reversed.
Here’s another example, this one published in 2014:
Example #4: Continued Growth
This post from I Done This doesn’t seem to suffer from decline, even after four years. In fact, incremental growth continues.
Why does the Wistia post trend downward while this post trends up? The difference between these two posts is the subject matter. The first (long decay phase) deals with a piece of hardware that is time-boxed. The second (long growth phase) covers something truly timeless.
What are the takeaways from this?
- Distribution matters. Nearly all successful posts we examined had a strong spike phase. This initial traffic allows you to sow the organic seed. People have to see your post to start linking to it. Additionally, a search algorithm can collect data on how people interact with your new post (dwell time, bounce rate, etc., factor into rankings). Getting it out into the world through Reddit, Twitter and especially your email list is important for the awareness that can lead to links that later help drive organic traffic.
- Growth comes from search. But if you want to make a splash on The Hacker News and haven’t thought about anything past that first day, your post will die and your blog won’t grow. Additionally, past a point, the size of that initial spike doesn’t seem to increase the size or velocity of continuing traffic—sustained success has more to do with on-page optimization, quality content and evergreen topics.
- Volume decides the size of your blog. The traffic on the organic ceiling is fairly easy to measure. When you’re ranking at the top for every target keyword, you’ve topped out on traffic. Choose a mix of keywords that drive traffic (top of funnel) and conversions (middle and bottom of the funnel). Each provides their own value.
- Decline happens. Almost all the time. Evergreen content can continue to grow for years, but other sites will want to knock you off the top of search results. If you have that position today, you still have to work to have it next year.
The last takeaway—this takes time. The success of the posts highlighted above are years in the making. You won’t see 100,000 views from that initial spike, but you also won’t see 100,000 views from organic overnight. It takes time for those views to accrue, but you can see early on whether you are heading in the right direction.
Will Your Post Take Off? Here’s What to Look for
It’s easy to analyze the success or failure of a post a year or two after publishing, but can you tell early on if a post will get traction in search?
This comes down to whether or not you get to phase three, the growth phase. If you have the slope of organic growth within about six months of publication, the post is successful. If the trough has no upside, you will have no growth.
Here’s an example post from Amplitude. Published in 2016, it hasn’t hit 100,000 views but is obviously on the way:
Example #5: Success Is Inevitable
This has reached phase three much earlier than examples #1 (AdEspresso) and #2 (Appcues) above. The initial spike came through email, but within three months, the post is growing through organic search traffic. Phase two barely exists.
When we look at relative channel contributions to traffic, organic is already contributing >90% of all traffic within half a year. This is key to success—the sooner organic traffic takes hold, the sooner you can expect significant (and passive) traffic. You can shorten the trough phase by investing time in promotion and link building after the initial spike. You can also update content and optimize on-page SEO after a post is live to improve its chances for success.
The post continues to grow. It has yet to reach the plateau phase. Considering it is currently on ~75,000 views, it will definitely surpass the 100k mark.
Not every post has to be heading towards the 100k mark to be considered successful. The amount of views you can achieve for any given post is relative to the search volume for its target keywords. Niche posts can be successful within their own restrictions. Here’s an example of a post from Clearbit that is not going to hit 100,000 views but can still be considered a success:
Example #6: A Smaller Organic Win
Clearbit, The Modern Guide to Lead Qualification
It has the initial spike driven primarily by direct traffic, which then drops and growth is taken over by organic search. The slope of the growth phase is shallower but still obvious and starts within the first six months. After a year, growth looks to have plateaued.
The same trend is seen in with this ProfitWell post, though its growth phase lasted longer:
Example #7: Search Volume Creates a Ceiling
ProfitWell, How Revenue Recognition Works in SaaS
Traffic to posts like these is bounded by its fairly niche topics. But—and this is an important “but”—this post is targeted directly at the buyer personas for one of their products. The people finding it through search and reading it are exactly the people this company wants to reach.
Of course, it doesn’t always work out. Often, it doesn’t. Here is an example of a post from a successful company with a successful blog that never reached the growth phase. The trough phase became a flatline:
In this case, 48% of all traffic to this post came on day one. Only 7.28% of traffic so far is from search. This is extremely common. The 100k club posts we’ve used as examples above are outliers. Here are some things to look out for in the days and weeks after launching a post that you hope will bring in lots of traffic:
- A small initial spike. If you don’t see good traffic to your post in week one, it doesn’t mean all is lost but does mean backlinks aren’t going to be forthcoming. Getting that spike means people are aware of your content. This is an important first step.
- Week-over-week organic growth. Especially in the first six months, you should be seeing steady growth each week. This means the post has gained a foothold and this traction can lead to a flywheel effect—the more people are exposed to the content, the more opportunities there are to link to it, which improves the rankings and exposes it to more people.
- Outsized returns from long-term traffic. Tom Tunguz says that “Only about 1⁄3 of the views on a typical post on tomtunguz.com are generated on the first day.” We have found the same. For the posts highlighted above, we’ve found that <10% comes in the first week. ~90% of all traffic to successful blog posts comes later.
These are heuristics. Your mileage may vary. But by looking at your Google Analytics through the prism of the five phases highlighted above, you can understand “success” in traffic terms for your blog and learn what works for your readers.
The Compound Effect of Content Marketing
Successful posts grow through good initial distribution and steady organic growth. Successful blogs grow by layering successful posts on top of each other.
The first post highlighted in this article is just one of many on AdEspresso’s blog that has passed 100,000 views. When we layer the top performing posts on top of each other, we can see how organic growth over time leads to overall blog growth:
These are just the top 10 performing posts and already weekly views are at 20,000. AdEspresso has posted thousands of posts. Their overall blog traffic looks like this:
They hit almost 150,000 weekly views. This is purely through the compounding effects of content marketing. One post layered on top of another, layered on top of another, etc. As long as you can reach the growth phase on a regular basis, you will be able to grow a large blog.
Blogs even have emergent properties—they are greater than the sum of their parts. The better individual posts perform, the more search engines will see your blog as an alpha resource and rank you higher.
But this success is predicated on individual posts. They have to perform. When you know the signs of success, you can drill down on what works for your readers and what doesn’t both in terms of initial and long-term distribution. By doing that, you can replicate success over and over, building layer upon layer of great posts, and build a great blog.