Open Road Integrated Media has been an active client for the past couple of years. I have been intrigued by their claim of having the only really automated ebook marketing system in existence. I can’t say I have the inside knowledge of every other big player’s operations to confirm that is true, but it certainly looks that way from the outside. Near as I can tell, the other players do their digital marketing, for ebooks, audio, and print books, using human beings to pursue opportunities on the titles their digital tools tell them are the most susceptible to improvement across the thousands of titles they control. But I don’t believe any other player has built anything comparable to what Open Road has.
The short summary is this. Open Road is managing digital marketing for more than 40,000 backlist titles, which include those they had signed up for themselves during their early life when grabbing ebook titles not covered by then-current contracts fueled their growth, plus a larger number that have come in the past few years from publishers using their Ignition marketing service. Their system canvasses for marketing opportunities across a range of opportunities we’ll describe further below, and then takes an action on approximately 1800 titles every day. That adds up to nearly 55,000 titles receiving a marketing action each month, or more than 1.3 marketing actions per title available each month.
(Obviously, many titles have more than one action taken for them and some titles don’t get any in a given month.)
That summary alone should help any other publisher that thinks they have a fully automated marketing system decide whether they’re comparable. The internal self-check is: do their monthly marketing actions number 1.3 times the number of backlist titles they have available to sell?
After they accept titles for their system (and most that publishers submit are approved), Open Road does a metadata review and improvement exercise for every title before they activate the actual marketing. That is also an automation-assisted exercise and it results in improved metadata that the publisher is free to use for their own marketing efforts. (Remember, Open Road is marketing ebooks only, although, obviously there is a “lift” effect from their efforts on the print and audio versions of the same books.) That’s a considerable ‘free” benefit for their publishers that, in my opinion, Open Road has made too little of in their self-promotion.
While much of what Open Road does for the titles rests with its own audiences, they have also built a system for “newsjacking” that allows them to pick up signals that are having a direct impact on hits to product pages and conversions at retail. What are the signals Open Road is looking for? This surely resembles what their competitors also seek. Changes in the news cycle that emphasize a specific subject covered in a book. Social media activity around a subject or author. Comments from an influencer. The author passing away. A new book by the author being issued. They’re acutely aware of the intersection of the subject, author, social media platform, and where any particular story broke.
And, when they spot an opportunity through that scan, what are the tools Open Road’s automated systems are directing them to employ?
Like other big publishers, Open Road has a number of “vertical” websites and each issues a companion daily newsletter. These are Early Bird Books (a daily eBook “deals” newsletter and website); The Lineup (true crime, horror, paranormal); The Portalist (science fiction and fantasy); Murder & Mayhem (mysteries and thrillers); The Archive (history and non-fiction); and A Love So True (romance). They also have two newsletters with consumer interests that range across their websites: The Reader (literary fiction) and Open Road Books for Young Readers. Open Road uses these tools themselves and also “sells” placements to other publishers not signed up with them for overall marketing through Open Road’s Ignition offering. (That is, they offer those human marketers at the other publishers the opportunity to decide to buy access to highly segmented niche audiences.)
And, like most other marketing-active publishers, Open Road is doing promotions with Kindle, Apple, Nook, Kobo, Hoopla, Overdrive, and Scribd regularly. They also work with BookBub, Humble Bundle, Epic, Fable, Library Ideas, and Odillo. And they place social media ads with TikTok, Pinterest, Facebook, Twitter, and Instagram.
In other words, the books the scan picks up as opportunity titles are arrayed against dozens, if not hundreds, of potential ways to exploit the moment for them. And 1800 separate instances of putting the book in front of a selected audience are executed. Every day.
Sometimes the action requires “work”, like nominating titles for promotions. If it does, that is also automated. This is another important distinction between what Open Road is doing and what everybody else is doing. Filling out the nominations for many of these opportunities is a non-trivial exercise if it is being done by “hand”. It is self-limiting. Open Road doesn’t suffer from those limitations. If 250 titles today should be submitted to BookBub, the nominations are submitted for all of them. Open Road believes that they deliver a 500% improvement in the number of promotions accepted, no doubt fueled by the massive increase they create in the number of promotions properly applied for.
The leverage this automation provides was illustrated by an exchange I had with an Open Road executive a few months ago. I “sagely” observed, “I bet you can learn a LOT from ten dollars spent on advertising on Facebook.” What I meant was that the Facebook data on respondents to ten bucks worth of ads would provide important clues about which Facebook users to “choose” to expose to a larger campaign.
“Yes, you could,” I was told. “But wouldn’t you rather put down your ten dollars where you knew you would get thirteen dollars back?”
That response rang two big bells for me. One was that Open Road knew, or could at least make a well-educated guess, which titles would deliver a $13 return on a $10 bet. But the bigger one underscored the massive advantage of their automation. If you had to use humans to place those $10 bets on Facebook, you’d lose money getting $13 back. Doing it 100 times or 1000 times with automation is very profitable. Doing it manually would cost more than it would gain you.
Some of the actions taken each day are follow-ups to prior actions. For example, social media ad placements that deliver above threshold ROI metrics are iteratively escalated to ever higher levels of investment until the metric tells the system to stop or scale back.
There are two useful ways to view the benefits the Open Road system consistently delivers. The simplest metric is the change in sales (revenue) across a basket of 1000 or 5000 or 10,000 titles. The short summary of that is that overall monthly sales revenue for the cohort will double in about 60 days and continue to increase less dramatically for a long time thereafter.
But the other is how many titles Open Road can move from “just about dead” into ranking among the top 100,000 titles, which triggers Amazon and Google algorithms to kick in that provide even more of a boost. It is a big deal to a publisher if they can take 5000 virtually dead titles and find after 90 days with Open Road that 500 of them are in that top 100,000 category. That really constitutes a “new lease on life.”
Since all very big publishers know that they have literally tens of thousands of titles their human marketers will never touch, it makes sense for those who really accept the proposition that automation can achieve real results to turn the marketing of those over to Open Road. (And that’s even before the publisher examines the “standard” Open Road “deal”, which takes nothing from the sales any title achieves until it passes its prior year’s revenues. Open Road takes all its compensation from the increase in ebook sales and nothing from the additional sales growth the publishers will achieve on the print and audio editions.)
And that leads to consideration of one more distinguishing characteristic of Open Road’s approach compared to the automation-assisted methods that look like the industry standard. There is a practical limit to the number of titles any other marketing organization can execute a promotion for on any given day. That depends on the extent of their staffing and the degree to which they have streamlined their efforts, including their use of proprietary tools.
But Open Road’s system works better for every title as more and more titles are added to the system. There are more titles with metadata that stamp them as candidates for a push on any given day. Those additional titles, in turn, generate more data that makes the system do a better job of picking titles for the opportunities they see going forward. As titles get added to the universe, the number of moves grow at pretty much the same pace.
It looks from here like Open Road’s automated marketing is a killer app. It is the result of tens of millions of dollars of technology investment and more than a decade of constant daily adjustments and improvements by an integrated team of software engineers, data scientists and marketing analysts. Any publisher who doesn’t have comparable automation is going to find sooner or later that it makes sense to turn over the titles that their humans are unlikely ever to touch to an automated marketing system. If that’s true, we’ll soon find out whether Open Road is unique, and can remain unique, or whether they are just the first to go public with a way of working that nobody else is talking about.
Given that it was built iteratively over a decade, with additional improvements to the algorithms taking place every day, Open Road has confidence that their system would be very hard to replicate. This is reflected in their openness in exposing the details of what they do. They tell their publishers about every move they make in periodic review summaries and they are open to sharing as much detail about how their system works as a publisher wants to know. While I first learned about how intricate the system was through my association with them on other questions, they answered every question I asked when I took on trying to understand it well enough to explain it.
As I said at the top, if anybody else has comparable capabilities, I haven’t seen evidence of it. My hunch is that while a rapidly increasing number of publishers will sign up to work through Open Road’s Ignition service, some of the biggest players may want to work with Open Road to build their own customized version of Open Road’s system. It will become table stakes to be automated to that level, and as you realize how much there is to automate and how many calculations and algorithms there are to develop and monitor, it might seem attractive to pay the rent to start with somebody else’s framework. But that’s something that hasn’t happened yet. And it is also an opportunity for anybody else out there, if there is anybody else, who has comparable capabilities.