The impact of a Google Update
Expertise SEO
By Anique Schoot
Google rolls out one update after another. Some bigger than others, and the various updates also vary considerably in noticeability. In some cases, Google itself gives a description of the changes that are coming, but at least as often it remains a mystery which buttons are being turned. For SEO, it is therefore actually always reactive work after the roll-out of such a new update.
Always up-to-date and smarter
‘Google is getting smarter’, you sometimes hear in the world of search engine optimisation (SEO). And that getting smarter happens through its frequent algorithm updates. Using a complex algorithm based on hundreds of variables, the search engine calculates which web pages are the most relevant results for an entered search query.
The outcome of the displayed search results is tested again and again against incoming user data. Were the results indeed helpful, or did visitors immediately click away again? Did the top results indeed get the most clicks, or did users scroll through looking for a better answer to their search query?
With algorithm updates – and especially with the more substantial Google Core Updates – algorithm changes are taking place, aiming to understand the searching user and the content of web pages even better.
The September 2022 Core Update
Some time ago, the Google Core Update of September 2022 took place, which started rolling out on 12 September and was completed on 26 September. About the content of the update, Google remained secretive: it only referred to information about Core updates in general. Now that this update is over a year old, it is time to draw our own conclusions.
To determine what impact a Google algorithm update has had, we can look at the shifts it has brought about. Measurement points: the day before the new update started and several days to weeks after the update was fully rolled out (this can sometimes take weeks).
For our own analysis, we collected 47,187 keywords from a wide range of themes and industries and retrieved the top 25 search results for each term. We compare this data at 2 points in time: 11 September and 20 October 2022, after the impact of the update has had some time to become visible.
With this analysis, we take a look at the changes before and after the Google Core Update from different perspectives. Thus, we look at:
- The differences in the composition of the top 10 for each keyword. Are the URLs that rank in the top 10 before and after the update the same, and if so, did the URLs maintain exactly the same position? We can look closely at this composition of URLs for each of the keywords;
- The Share of Voices (SOV; a plot showing the market distribution within the top 20) for each theme/industry from before and after the update, and the percentage differences between these SOVs;
- At the domain level, the differences before and after the update: which URLs within a specific party’s domain had the most keywords with a lot of change in their top 10? And in which keywords per URL was that difference mainly?
Shifts in the top-10 after Google September update
A comparison of all 47,187 keywords between time point 1 and 2 yields the following insights:
- 55.43% of the search terms have the same URLs in the top 10 in both the first and second periods. This means that for about 45% of the keyword URLs disappeared from the top 10 and were thus new entrants. Overall, there is 0.08% where the entire top 10 is new.
- 0.53% of search terms have 10 URLs that have the same position in the top 10 in both the first and second periods. 2.33% of search terms have 0 URLs that have the same position. The most common occurrence is 18.82% that 4 URLs still have the same position.
The further distribution of matching URLs and positions is as follows:
Processing the data in boxplots, we can clearly see where the centre of gravity of the changes and similarities lie.
Unsurprisingly, in most cases the composition of the top 10 remained the same, with the percentage of changing URLs decreasing rapidly as the number of URLs increased. Had it been the other way around, this would have meant Google would have reconsidered the search intent or best results for many keywords. In other words, that the search engine was not doing its job too well before the update.
In the second boxplot, we see that the most frequent occurrence was that 3 to 4 URLs changed position within the top 10. Again, a result in line with expectations: more unlikely is that no changes were needed at all, or that all URLs were at a non-optimal position.
The above ratio of shifts is average, but the proportions vary by theme. For instance, the composition and position of URLs for themes such as cooking, financial, casino and personal injury remained fairly similar, while the boat and wallpaper sectors were subject to the most changes overall.
Shifts in different sectors
We have divided the 47,187 search terms into 29 themes, so that we can also roughly measure the effect of the latest Google Update per industry. This way, we can zoom in on the successes and losses of specific parties that have gained or lost relatively much. To illustrate, let’s take the clothing sector.
In a Share of Voice that looks at the percentage differences within the top 20 between the two points in time for a selection of the larger fashion web shops, we see that most of the parties stick around the middle. They have gained or lost visibility slightly (change in indexing percentage) and/or gained or lost a few tenths of positions on average (position change within the top 20), but the changes since the update are not drastic. In general, the clothing sector has improved: the average position has decreased slightly about as often as it has increased, and the indexation percentage has increased significantly more often than it has decreased.
However, we do see 2 big outliers to both sides. At one end we have Only.com, which has risen almost 15% in average position. At the other end of the spectrum, we have Zalando.co.uk, which has dropped as much as 30% in average position, but in return is shown in the top 20 with a whopping 210% of additional keywords.
Shifts by keyword
We can also measure the impact of the Google algorithm updates on a keyword basis. As we saw earlier, in most cases there were 3-4 URL changes within the top 10. So we can consider this as ‘normal’ for the September update. It is more interesting to zoom in on keywords where there were large changes in the top 10 – for example, the 0.08% where all 10 URLs are different from before the update – to determine how Google’s understanding improved or changed after the update.
One of these keywords was ‘industrial dining chair’. The position distribution in the top 10 looked like this on 11 September (left) and 20 October (right):
URLs first measurement | position | URL present second measurement | URLs second measurement | position |
maxwonen.nl/tafels/eettafels/industriele-eettafel | 1 | ❌ | dimehouse.nl/stoelen/industriele-eetkamerstoelen | 1 |
dimehouse.nl/tafels/eettafel-industrieel | 2 | ❌ | bol.com/nl/nl/l/industriele-eetkamerstoelen/25160/34301 | 2 |
industrieelwonen.nl/tafels/eettafels | 3 | ❌ | debommelmeubelen.nl/meubelen/stoelen/eetkamerstoelen/industriele-eetkamerstoelen | 3 |
robuustetafels.nl/tafels/industriele-tafels | 4 | ❌ | maxwonen.nl/stoelen/eetkamerstoelen/industriele-eetkamerstoelen | 4 |
bol.com/nl/nl/l/industriele-eettafels/25165/34301 | 5 | ❌ | robuustetafels.nl/leren-eetkamerstoelen/industrieel | 5 |
profijtmeubel.nl/eettafel-industrieel | 6 | ❌ | gigameubel.nl/industriele-meubels/industriele-stoelen | 6 |
gigameubel.nl/industriele-meubels/industriele-tafels | 7 | ❌ | profijtmeubel.nl/eetkamerstoelen-industrieel | 7 |
houtmijn.nl/industriele-meubels/eettafels | 8 | ❌ | beslist.nl/products/meubilair/meubilair_389370_389409/c/stijl_test~8064049 | 8 |
eijerkamp.nl/eetkamer/eettafels/industrieel | 9 | ❌ | sohome.nl/zitmeubelen/eetkamerstoelen/industrieel.html | 9 |
tafels99.nl/eiken-tafels/industriele-tafel | 10 | ❌ | countrylifestyle.nl/zitmeubelen/eetkamerstoelen/eetstoel-industrieel | 10 |
Interestingly, the entire top 10 before the update was rolled out consisted of irrelevant search results for the logical search intent behind this keyword, namely: industrial-style dining chairs. Instead, the search engine only displayed pages with industrial-style dining tables. We can conclude that Google did not yet understand that a ‘dining chair’ is a piece of seating furniture.
After the September update, the picture is completely different. The entire top 10 now consists of URLs related to industrial dining (room) chairs. The algorithm has learned that dining tables and dining chairs are closely related (and thus most likely also often mentioned together in the same breath on the same page), but not the same.
Always up-to-date in your own sector?
The impact of a Google Update not only applies to September’s Core Update, but we can and will make it measurable for every major algorithm update. This will give you the most important changes at a glance and, if necessary, you can respond immediately with your SEO strategy. Stay up-to-date and keep an eye on our website after the next Google Core Update! Is your own sector not covered in our blog? We can run our Google Update Effect Measurements and market-wide analyses for any online market, so feel free to contact us for a customised analysis.