Before we chart new featured snippet territory, we’re doing a quick update on some past data.


While we’ve been off chasing searcher intent and peeking into “People also ask” boxes, featured snippets haven’t been far from our minds. Lately, we’ve been wondering about snippet stability — when you win one, how long do you get to keep it?

So, we did what any good data wonks would do: we got researching. And as we number-crunched snippet volatility, we also updated some findings from our original study (and its refresh!) to see just how much everyone’s favourite SERP feature has changed. Here’s what’s new.

The number of featured snippets are still on the rise

In our initial 2016 study, featured snippets appeared on 9.28 percent of our one million SERPs. In 2017, they’re showing up on 23.25 percent of them. This makes a 54.99 percent increase in snippets over our July 2016 refresh, and a whopping 150.40 percent increase over our January total.

It’s apparent that Google isn’t close to being done with featured snippets, and we wouldn’t be surprised if they keep multiplying.

Swipe for more →

January 2016 July 2016 April 2017
Number of snippets 92,832 (9.28%) 149,980 (15.00%) 232,451 (23.25%)

Skyrocketing snippets

Featured snippets show no sign of slowing how often they appear on the SERP.

List snippets continue to grow in popularity

You may remember that paragraph snippets dominated the scene back in January 2016. This is likely because snippets were still a relatively new SERP feature, and it was easy for Google to pull these answers, as-is, from their source material.

When we checked in on snippets six months later in July, we saw a sharp decline in paragraph formatting, with its loss going almost equally to both list and table snippets. It was clear Google was experimenting with these fancier formats — ones that, if not already formatted as such, take a little extra work to lay out.

Nine months after that, paragraph snippets have pretty much held steady as the preferred format of choice, but we’re now seeing a drop in tables, and another big jump for lists.

We’re going to chalk this change up to the fact that lists are easy for searchers to scan (and for Google to compile), while the complexity of table snippets typically results in confusing info and a less-than-great user experience.

Swipe for more →

January 2016 July 2016 April 2017
Paragraph 82% 63% 60%
List 11% 19% 29%
Table 7% 16% 11%

Paragraph is the favoured format

Paragraph snippets continue to rule the roost.

Featured Snippets and PAAs are BFFs

When we first calculated how often snippets shared the SERP sandbox with other features, we saw that they were most friendly with videos, “People also ask” boxes, images, news, and shopping boxes.

Fast forward a year and change, and the relationships between snippets and all but one of those SERP features has started to crumble, with images taking the largest hit of the lot.

Only the bond between PAAs and featured snippets has grown stronger, now appearing together 70.68 percent of the time. Given how similar these two SERP features are, this may be a complete surprise or totally expected — they either compete with or complement one another. Are we headed for PAA takeover? Could we see a hybrid of the two one day — a PAA box with a pre-expanded, snippet-worthy first question?

Swipe for more →

SERP feature January 2016 April 2017
Images 13.9% 9.4%
News 5.4% 3.0%
“People also ask” 22.0% 70.7%
Shopping 4.0% 3.6%
Video 24.5% 22.8%

How well do snippets play with others?

The relationship between snippets and most other SERP features has weakened over time.

Rogue snippets are breaking (first position) rank

In the early days, featured snippets always took the top of the SERP — so much so that we often said they occupied “position 0.” Today, snippets are starting to fight with other SERP features for that number one rank. A hair over four percent of our featured snippets appeared in position two:

  • Rank 1 = 95.88%
  • Rank 2 = 4.07%

(And 127 snippets fell among ranks three, four, and five.)

So, who’s managed to snatch the victory spot away from featured snippets? After some poking around, we saw shopping boxes appear in the first position most often, followed by home services and news results.

In other words, paid placements and trending stories can make a more established answer take a backseat.

Featured snippet source position has dropped

In January 2016, we found that the majority of our featured snippets were being sourced from position two (29.60 percent). Now, we’re seeing almost that same percentage (28.91 percent) come from the third place rank.

And, in general, the sweet spot where featured snippets are sourced from has shifted from the two-to-four range (66.30 percent of all snippet URLs), to the three-to-five range (63.91 percent).

While we can’t quite put our finger on exactly why this is, a small contributing factor may be the four percent of featured snippets appearing at a lower rank.

We also noticed that a month after we ran our data, Google tested pushing the organic URL completely off the SERP. Recent checks in STAT show this wasn’t a permanent decision, but needless to say, we’ll be keeping an eye on snippet source position.

Swipe for more →

Rank January 2016 April 2017
1 0% 0%
2 29.60% 4.76%
3 20.46% 28.91%
4 16.24% 20.04%
5 12.88% 14.96%
6 10.42% 11.14%
7 4.02% 7.56%
8 2.18% 2.55%
9 1.53% 1.39%
10 1.03% 0.94%
11-12 1.65% 1.12%

Featured snippet source position

Google’s starting to pull featured snippets from further down the SERP.

Up next: Featured snippet volatility

Now that you’ve been refreshed on the basics, it’s time to look at featured snippets in a whole new way: their stability on the SERP.

We tracked a mountain of snippets over a 20-day period to see just how shifty they might be — like how quickly they come and go, or how often they change hands. Want to see how we pulled off this crazy experiment? Give the methodology a read. Want to jump straight into the insights? Feast your eyes on the full research findings.

If you’re curious about STAT’s ability to track snippets — and other SERP features — for your own analysis, we’d be happy to show you how. Just say hello and request a demo!