Bad Data Journalism (aka Black Hat Digital PR) – It’s Everywhere

Note I've broken / removed some hyperlinks in this article.

  What is the biggest "mistake" people make r.e. interpreting Google Trends data?

    Not knowing that the, "... numbers are ... scaled on a range of 0 to 100 based on a topic’s proportion to ALL searches on ALL topics" (https://support.google.com/trends/answer/4365533?hl=en). 

    Sometimes it appears that this data is knowingly (?) misrepresented (stating data on RELATIVE popularity represents data on ABSOLUTE popularity) to grab headlines via "fake news" in "content marketing" efforts by Search Engine Optimisation (SEO) / Public Relations consultants . For example, circulating a press release with a 'topical' / current news headlines like "The highest level of internet searches in recorded history were made in the UK for the phrase ‘nose job’ last month on Google.co.uk . Searches went up by 200% in June 2022, data commissioned by cosmetic surgery specialist http://BorisEdwards .co .uk found." . This is for the purposes of trying to acquire coverage and / or backlinks for SEO purposes . It even gets printed by papers with SEO departments like the Daily Telegraph , who are able to fact check this fake news. 

Exhibit A

  It was reported on https://www.telegraph .co.uk/money/consumer-affairs/workers-fleeing-britain-escape-skyrocketing-prices/ on 6.5.2022· that “The highest level of searches in internet history were made in Britain for the term ‘move abroad’ last month on search engine giant Google. Searches rose by 1,000pc in April, data commissioned by emigration specialist Reiss Edwards found.” 

   When I checked, the data for April 2022 referred to in the article was not yet available in Google Keyword Planner on 7.5.22. From my check this data was made available around 11.5.22 . Were they making it up? 

    Here is data for the year up to the end of April 2022 for Google in the UK (the three monthly and Year on Year search volume trend is +22% for both):

Exhibit B

     I saw on Linkedin a Content Marketing case study where a Digital Marketing Company had circulated a press release with some research on Google searches for the phrase ‘delete Twitter’ which had helped their (gambling) client acquire backlinks and media coverage.

   The case study stated that “Google Trends has shown a 910% increase in searches for this term, in the past day”. I got into a discussion in the comments on the post , stating that  Google Trends shows the relative popularity of a search term, not absolute.  Thus you cannot infer "a 910% increase in searches for this term in the past day" from this data, only a 910% increase in the relative popularity of searches for this term in the past day. See https://support.google.com/trends/answer/4365533?hl=en .

Exhibit C

     Even the likes of an influential Silicon Valley venture capitalist & analyst can get their analysis wrong it seems.

Exhibit D

  Out of interest I recently looked up the price of the Ether cryptocurrency on coindesk .com/price/ethereum/ where there is a "search trends" graph -  see below. The explainer pop up misrepresents what Google Trends data is as does the linked page from the "Learn More" text: https://resources.intotheblock .com/indicators/social/search-trends (I've broken the links).

Exhibit E

  I saw Mark Rofe write on Twitter that it was "Interesting to see [Digital PR agency] https://search-intelligence .co.uk make use of the website http://journoresearch .org to presumably do their outreach so execs have a journoresearch .org email address." with the inference that this will give more credibility to their link building / SEO efforts .

   Using a different email address doesn't mean they are conducting valid statistical "research" . For example, journoresearch .org's data "analysis" for their client as published on Hull Daily Mail that "analysis of Google Trends by Casina Grounds revealed that search volume for [the phrase] 'learn sign language' rocketed by 1,011 per cent in the space of three hours on Saturday, October 23, when Strictly was on air." is not correct .

   This (absolute search volume) data is not available in Google Trends - it only measure's relative popularity of search terms and topics: https://lnkd.in/dTyZUnay . When will journalists realise that a lot of digital PR is "fake news"?