Growth Memo

Growth Memo

Share this post

Growth Memo
Growth Memo
The baseline fallacy of SERP Features
Copy link
Facebook
Email
Notes
More
User's avatar
Discover more from Growth Memo
Growth Memo distills rigorous, data-driven research on SEO and organic growth into clear, actionable frameworks that empower senior marketers to make strategic decisions in an AI-driven search landscape.
Over 19,000 subscribers
Already have an account? Sign in

The baseline fallacy of SERP Features

We cannot measure the success of SERP Features without measuring baselines. Otherwise, we won’t know if we’re successful because of our actions or luck.

Kevin Indig's avatar
Kevin Indig
Nov 15, 2021

Share this post

Growth Memo
Growth Memo
The baseline fallacy of SERP Features
Copy link
Facebook
Email
Notes
More
Share

Get your content indexed instantly with Real-time IndexNow

ContentKing now supports Real-time IndexNow!

Using real-time monitoring, ContentKing tracks every change on your website and can now automatically notify search engines when content is added, updated or removed on your site — all in real time.

Try Real-time IndexNow


Any modern SEO strategy should cover SERP Features optimization. The modules Google pulls from vertical search and the knowledge graph to augment search results have a significant impact on organic traffic: Featured Snippets draw a lot of clicks, which means the remaining results get a lot less than when a Featured Snippet is not present. Image boxes are even worse!

Every business falls into an SEO strategy archetype and every SEO archetype is best suited for a set of SERP Features. Recipe aggregators dominate recipe packs. Marketplace aggregators occupy Featured Snippets. Publisher integrators cover video carousels.

But when measuring the number of SERP Features for our site(s), we often forget the baseline. Say, your clicks from organic traffic increased by 20% week over week. That's a good result! However, if rank is stable and your impressions also increase by +20% W/W, you're not performing better. The increase is simply the result of more people searching for the keyword the page ranks for.

This is a common fallacy: we judge progress in isolation but forget to measure it against potential, which in return leads to survivorship bias. We think our action led to results while actually other factors were at play. We take the success as a tactic but fail to replicate it and are confused as a result.

SERP Features are especially sensitive to baselines because they a) fluctuate constantly and b) we don't have a lot of data resources to monitor them. Time for some examples for how to do it.

Example 1: Wikipedia and Featured Snippets

According to Semrush, Wikipedia owns the Featured Snippet for 391K keywords on mobile and 1.6M on Desktop (November 2021). In absolute terms, a good performance. In relative terms, not so much.

See, that's where the baseline fallacy is at its finest work. We hear 1.6M Featured Snippets and think "wow!" But measured against the base rate, the example is not as stunning anymore.

Wikipedia ranks for 15.7M keywords that display a Featured Snippet but doesn't occupy the it (Desktop, November 2021, USA). On mobile, it’s 2.1M. In other words, Wikipedia captures only 10% of available Featured Snippets on Desktop (1.6M / 15.7M) and 19.5% on mobile (391K / 2.1M).

Now, Wikipedia doesn’t do intentional SEO. But, If I were doing SEO for Wikipedia, I would seek to match the article content with the passage or list that appears in the Featured Snippet. There might even be an automated way to do that.

Example 2: New York Times

Based on Semrush data, the total opportunity for nytimes.com to rank for People Also Asked features (PAAs) is 14.3M (Desktop, November 2021, USA). However, the Times ranks in only 348K PAAs, which is 2.4% of the total market! Pretty low, and an opportunity for competitors.

For the New York Times, this could pose an opportunity to answer more questions in the content to rank for more PAAs. SERP Features can lead the way here. You could export the PAAs you can find, match it to existing content, and add answers to the questions retrospectively.

Example 3: eBay

I also want to highlight a successful example, and that is eBay. Images play an important part in ecommerce because they literally display the product to customers. Since they can't see and touch it offline, they rely on strong visuals.

eBay performs well for image packs: 22% coverage on desktop (5.7/25M keywords) and 15% on mobile (23.3/132K) in the US (November 2021).

Baseline for competitors

This all goes to say that you should monitor the total available number of SERP Features that are relevant to you in a 3rd party app (or homegrown tool). It's the only way to truly understand if you grow with the market or not.

You can do this for you competitors, too. Knowing which and how many SERP Features they occupy is a powerful way to either get a competitive advantage, defend your business, or attack another one.


Subscribe to Growth Memo

By Kevin Indig · Hundreds of paid subscribers
Growth Memo distills rigorous, data-driven research on SEO and organic growth into clear, actionable frameworks that empower senior marketers to make strategic decisions in an AI-driven search landscape.

Share this post

Growth Memo
Growth Memo
The baseline fallacy of SERP Features
Copy link
Facebook
Email
Notes
More
Share

Discussion about this post

User's avatar
2024 predictions
#234 🔮AI, Organic Growth and Winner / loser predictions
Jan 3, 2024 • 
Kevin Indig
37

Share this post

Growth Memo
Growth Memo
2024 predictions
Copy link
Facebook
Email
Notes
More
2
The first-ever UX Study of Google’s AI Overviews: The Data We've All Been Waiting For
We tracked 70 users across 8 tasks to deeply understand their behavior, thoughts emotions when engaging with Google's AI Overviews and other SERP…
May 12 • 
Kevin Indig
54

Share this post

Growth Memo
Growth Memo
The first-ever UX Study of Google’s AI Overviews: The Data We've All Been Waiting For
Copy link
Facebook
Email
Notes
More
30
What content works well in LLMs?
A new study of brand mentions and citations in Chat GPT & Co
Mar 17 • 
Kevin Indig
51

Share this post

Growth Memo
Growth Memo
What content works well in LLMs?
Copy link
Facebook
Email
Notes
More
6

Ready for more?

© 2025 Kevin Indig
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture

Share

Copy link
Facebook
Email
Notes
More

Create your profile

User's avatar

Only paid subscribers can comment on this post

Already a paid subscriber? Sign in

Check your email

For your security, we need to re-authenticate you.

Click the link we sent to , or click here to sign in.