Thoughts on offering programmatic SEO as a service

Last week, I wrapped up a unique project with a client. They hired me to implement programmatic SEO on their website. This was my first time ever offering this service, and I learned a lot in the process, so I thought it would be interesting to share!

First, if you don't know what programmatic SEO is, I did a write up on my blog about it last year when I set up a little experiment around it. Basically, it's when you target long tail keywords that follow a pattern. Things like "best coffee shop in Dallas" and "best coffee shop in Denver" are good examples.

Using a large data set (or most recently, AI), you can automate creating content that answers the user's query, allowing you to target tons of user searches, and get lots of traffic for minimal work. My experiment targets keywords in the gardening niche, and just surpassed 2k users per month at 6 months old.

(@thefedoration also recently did a write up on his approach to progrmmatic SEO that you can find here)

For my client, I generated a little over 2000 landing pages from two spreadsheets of data. Each page targets a long tail keyword in their niche that gets about 100 - 10 user searches per month. Since the volume for these keywords are so low, there's very little competition when it comes to targeting them.

When these pages mature, they should bring in over 20,000 hits per month!

I spent 57 hours total on this project. If you break it down I think that's about 30 seconds per page.

These hours included keyword research, creating templates for the content, normalizing the spreadsheet data a little, importing the pages into the user's site, and incorporating the client's feedback when they had it. For this project, I did not use AI or GPT-3.

Looking back, importing the pages took up more time than I thought it would. The client uses a Wordpress website and I was unfamiliar with the best way to import a large amount of pages at once with that platform. Now that I've done it once, importing pages for any future clients with Wordpress should be faster.

Still, if future clients are using a platform I'm not familiar with, I'll likely budget extra time to explore the API for whatever site builder they're using. No one wants to manually go in and create 2000 pages on a site!

Another area where I wasn't expecting to spend so much time was normalizing the spreadsheet data. In my opinion, this creates the most opportunity for increased scope, since you'll probably have to aggregate or manipulate the data a little to match the kind of content you want to create.

The client is in an older industry with huge incumbents. They couldn't compete with traditional SEO, but there was a surprising amount of opportunity with programmatic SEO. I see a lot of potential for people who have difficulty implementing SEO to get traffic much faster using a programmatic strategy.

Overall, I really enjoyed this work. I feel like I made a bigger impact on the client's business. More so than simply pushing out a new feature at least.

Curious if anyone else has offered programmatic SEO as a service, and what their thoughts are on it.

For those who are interested, I wrote more about this topic and my original SEO experiment in my April retrospective.