The Concept of Input Metrics for SEOBy Martijn Scheijbeler Published June 7, 2022 Last Modified June 1, 2023
Sessions, transactions, and revenue are not metrics that an SEO shift overnight. It’s a matter of often waiting for Google. They’re considered output metrics, the work that you’ve put in results in those.
I just put down the book: Working Backwards by Bill Carr and Colin Bryar. It’s about their many learnings working for years at Amazon. It documents the process and policies that were put in place to make it the true giant it is today (including the theory behind pizza teams, the bar-raising process, etc.). One of the chapters talks about the concept of Input Metrics. Something that we often forget about in the context of SEO (or driving traffic via other channels).
Input metrics are considered metrics that can help define progress towards an output metric.
- How many products do you have (today versus a week/month ago)? Metric: # of SKUs.
- How many pages do have a custom/unique title or META description? Metric: # of optimized pages.
Analysis Decision Tree
They can be influenced by the work that you do daily. Let’s say that your (output metric) revenue is down. You would likely follow a decision tree like this:
- Did the number of transactions decrease or did the average order value decrease last week?
- If transactions decreased, did our conversation rate change?
- If the conversion rate didn’t change, did sessions change?
- If sessions were down, what channels caused this? Let’s say for a minute that all channels were flat except for Social Media.
- If Social Media was down on the metric of sessions. What caused it?
A simple explanation often follows: We just posted less on social media channels on Tuesday because of X. Aka, you defined the input metric that is in your hands to change: the # number of posts. My take is that any metric you report on should be able to either trigger an action, report an outcome, or show industry-level trends/benchmarking. Knowing certain metrics is useless without context via either benchmarking or historical data.
Example: # of Posts > More Social Posts > More Traffic > Larger Community
While I led Marketing at The Next Web, we looked at our variability in traffic to figure out how to grow our audience more sustainably over time (instead of relying on content going viral). Besides the apparent focus on SEO, we realized the # of published posts was a big input. It was not too surprising, but it was an input metric that could significantly impact sessions. For example, it led to us republishing or reposting old content on Social Media and having more writer support on weekends to create a steady stream of content. Without knowing what impacts traffic on a channel, you’ll have difficulty changing your approach.
Reporting Input Metrics versus Output Metrics
What metrics should you be reporting to your boss or upper level? The ones that show the impact, which is most often output metrics in my opinion. You want to show business results as that shows your contribution to the business’s bottom line (at the end of the day we’re all getting paid based on that).
→ Tom Critchlow has written a bit more about this subject in his blog post: “Some Notes on Executive Dashboards
However, it doesn’t mean that you shouldn’t have to report for input metrics. If you know the % of pages that aren’t indexed, you have an input metric that needs to change to grow the output metric: sessions over time. If you run an e-commerce store, the number of products has an impact on their availability. In RVshare’s case, we can have thousands of RVs, but if they’re not available at the highest peak of the year, it’s still not going to help us (or the owner) grow.
Example Input Metrics for SEO
Input to Output can often be visualized as a funnel, as in SEO, there are many steps that eventually lead to an outcome:
Crawl to Indexation: Each combination of steps will provide a way to input the next outcome. Increasing the number of pages should increase how many are crawled.
- Number of pages (input)
- Number of pages that are crawled
- Number of pages that are submitted (via XML sitemaps)
- Number of pages that are indexed
- Number of pages receiving traffic
- Number of pages driving revenue (output)
What do I need to get started?
Getting this insight goes back to having access to the right data. Most of what I just talked about is available through free tools: Google Search Console and Google Analytics is your best friend. The next best source is your internal data or CMS, which can provide insights into the quality of content/products/etc.