What is programmatic SEO and why don’t I care?
Howdy there, I’m Nick! On a typical day I lead Dataherald’s ops and CX teams, but today I am the author of this blog post and your personal tour guide to one of the world’s most (kind of) exciting topics - programmatic SEO. I understand the title of this blog post suggests that I should be telling you what programmatic SEO is, but I’m going to throw you a curve ball and tell you what it is not, because I think that may be more important. Programmatic SEO is NOT a programmatic way of monitoring or improving your site’s SEO. For that matter, it has very very very little to do with SEO at all.
Essentially, programmatic SEO is a misleading name for a strategy that enables you to create thousands, or even millions, of unique (yet templatized) web pages at scale with minimal time and effort. In most cases, people don’t simply want to create these pages for funzies, rather they want their newly created pages to be discoverable on the web. Thus, if you’d like to please the SEO gods at Google, traditional SEO principles apply. Your content needs to be unique, valuable, and crawlable.
So, perhaps a better term for this technology would be Programmatic Content Creation or Data-powered Content. Names aside, here’s the nuts and bolts of it: let’s say you run a recruitment agency in the US and you specialize in placing plumbers, electricians, and carpenters into great career opportunities. Since you are an industry leading marketer, you know that your target audience is typically not looking just for new career opportunities. They are looking for career opportunities in those three fields and in specific locations. Thus, you turn to a programmatic SEO service, upload your database with 50 states and 3 job titles and generate 150 templatized landing pages titled something like: “Career opportunities for {Job_title} in {State}.” Congratulations, you just completed programmatic SEO 101!
Unfortunately, the above example does not check the boxes of “unique and valuable content,” so Google is likely not going to rank those new pages you worked so hard on. To ensure these 150 pages have unique and valuable content, you are going to need to include information specific to each combination of job title and state. There are two ways to achieve this, manually or programmatically. For this simple example, creating unique content for 150 pages may be a viable option. However, if instead of targeting these searches at the state-level, you choose to target the top 2,000 cities, you’ll soon have more content to create than you can shake a stick at, and trust me, you have better things to do than shaking sticks.
That’s where option two comes into play, creating unique and valuable content programmatically. To solve this challenge, we at Dataherald play jump rope with the worlds of AI and big data. Merging these two technologies means that companies looking to deploy marketing content at scale can do so in an efficient and data-centric manner. How would this apply to the 150 pages we previously discussed? Glad you asked! Data and AI can be leveraged in a multitude of ways to enrich these pages. For example, you could include internal data like open job listings or company reviews as well as external data like industry trends, salary ranges, demographic shifts and more. Next, take all of those data points and mix in just a touch of AI magic and you are officially the proud owner of 150 hella unique and valuable landing pages targeted to help your ideal prospects find your solution. The coolest part is, when your content is powered by data, you have a new story to tell every time the data refreshes, ergo your pages never go stale and Google never forgets about you.
To quote every high school essay ever written “in conclusion…” programmatic SEO is a growth channel utilized by many companies, like Zillow, Zapier, Stack Overflow, etc. to create and publish marketing content at scale. Those companies, however, weren’t as lucky as you, they each had to develop this technology in-house. You, on the other hand, just have to talk to Dataherald :)
