Keyword randomly dropping in and out of SERP's

[chaplinashley]

Since around May 2017 ive seen keyword rankings dropping in and out of the SERP's on a weekly basis. What i mean is that a certain keyword will one day rank page one and then literally not rank at all the next, and then regain its position again a few days later to then loose it again. Its happening regularly but only for a select few keywords. Its happening for a few clients in different niches.

Im wondering if anyone else is experiencing the same issues?

Could it be Google or just issues with the SERP tool (SEMRush) which we use to monitor rankings? Ive contacted SEM Rush and they deny any issues with their tracking tool.

[keyplyr]

Could it be Google or just issues with the SERP tool (SEMRush) which we use to monitor rankings?

One way to find out is to manually do these keyword tests at Google Search.

Ive contacted SEM Rush and they deny any issues with their tracking tool.

Nothing negative to SEM Rush, but isn't that the norm?

SERPs are often volatile with results from automated tools that constantly send queries. That's the trade-off using tools compared to manual labor. Even doing it by hand there's a certain percentage of flux, especially over time.

[pvgseoexpert]

There might be possibility of Google Dance... but yes I think you can make a point of that keywords which shows a fluctuation in ranking and track them manually... usually this doesn't happen. Also so check or keep on eye on when google crawls your website by checking cached... Thanks :)

[htmlbasictutor]

Barry Schwartz has been writing about a possible update over the weekend of July 8-9. Maybe you are fluctuating due to this?

See his article Signs Of A Google Search Algorithm Update Over Weekend

[NickMNS]

It might be that Google is intentionally messing with all the SEO tools, gaming the gamers. Mozcast has been reporting value near or above 100 for nearly two month now. Typically in the past 100 was only reached when there were major updates. We are not seeing anything like this today, yesterday or for the past couple of month, and certainly not on ongoing basis as tool would suggest.

[tangor]

What's the time frame? I mean in ANNUAL or better. Weekly or hourly or micro minutes which some think is the g web just don't count.

Chronicle the daily in and out of that keyword for a YEAR then report back on percentage of page one and not page one (without revealing the keyword, of course, we don't discuss keywords at WW)

Keywords, in general, fluctuate. Always have. What works on Monday might not work on Tuesday, but does work January, February, March, April, May, June, July and not so well the rest of the year. Any site dependent on a single keyword will suffer the same fate so it is always suggested to target more than ONE and if one of those targeted gets figgity to not obsess, just keep plugging away.

[Writerly]

This fluctuation could be an outcome of Google's constant index update. As for the SEM Rush, even if they had some issues with their tools, of course they will deny them, but SERPs rarely give you stable results from automated tools.
It can be time-consuming, but if you want to find out if it's Google's fault, I believe that you should manually do keyword tests at Google Search or consult another keyword searching tool.

[NickMNS]

The fundamental problem, be it with doing the test manually or with using a tool, is that one keywords != one serp result. There is personalization, type of device, browsing history, geo-location, time of day, time of year, plus probably many other factors that impact what Google chooses to include in the search, not to mention that any permutation and combination of these factor that will change the result. So yes you can go and search in incognito mode and get one serp but it wont be representative of reality. And it is impossible for an individual or a tool designer to determine what would be a representative sample.

The best representation is from search analytics in GSC, and even that is flawed.