Is Google search 'good' search?

I have written about Google recently (here), but this time I want to consider search in particular.

Apparently around 80% of search is Google search, with up to 94% of secondary school students using Google for their search needs (Choice Media Channel, 2024; Connolly & Branham, 2024). However, the Alphabet business model, advertising, and ranking algorithms - that's the parent company of Google search - raise some flags with the transparency, quality, and functionality of the search product (Choice Media Channel, 2024; Connolly & Branham, 2024).

Like the "Black Box" of Van Bertalanffy (1968) where transformation happens behind the curtain, Google's search algorithms and ranking too is opaque. We enter our search terms, and whatever arrives we must accept (Choice Media Channel, 2024; Connolly & Branham, 2024). We have no ability to interrogate our results any further than the search results we are delivered. Or we have to run another search, refining our terms. 

Further, Alphabet will prioritise the products of its advertisers - aka "self-preferencing" - those of us who don't pay (Gleason et al., 2023), and - even worse imho - advertisers may pay a premium to artificially inflate their search ranking (Choice Media Channel, 2024; Connolly & Branham, 2024). 

With each successive search, Google's search engine itself is gaining a clearer picture of our user behaviours, interests, and demographics which they can parcel up and sell to any who want it (Choice Media Channel, 2024; Connolly & Branham, 2024). Well: we accepted the EULA (end user licencing agreement) in order to set up our Google products initially, which transfers our data into Google's ownership. That old adage, that if we are not paying, then we are the product (Johnson, 2001) holds true. But it can lead to "filter bubbles", where results we see may be biased due to what we have previously sought, rather than new results (Choice Media Channel, 2024; Connolly & Branham, 2024). 

Another curl with this idea is geographical bias, where search results may depend on where we live, meaning our searches are less replicable (Briscoe & Rogers, 2024). We may also get results ending at paywalls or in rotted links, making search more frustrating than it once was (Choice Media Channel, 2024; Connolly & Branham, 2024). You may - like me - have stopped even noticing that Google proclaims "estimated" results are in the millions... while the viewable results are probably on a tiny fraction of that; perhaps fewer than 0.0001% (Briscoe & Rogers, 2024). 

Speaking of volume, the sheer number of under-curated records makes it hard for us to find high quality search results - finding our needle in that vast global knowledge haystack -  the standard Google search tools are focused on finding commercial and travel solutions (Choice Media Channel, 2024; Connolly & Branham, 2024). While we can use Boolean search in Google searches, they do not always work, and many users are unfamiliar with them (Choice Media Channel, 2024; Connolly & Branham, 2024). 

Then we have the frankly dishonest behaviours. Google has been accused of misappropriating content from smaller third-party sites for use in its own knowledge components (Gleason et al., 2023), that scraping of summary answers at the top of Google's results list may "directly answer their questions [...], leading to good abandonment [...] of search activity" (p. 245) resulting in reduced web-traffic and revenue for the content creators themselves.

And lastly we get those AI "Hallucinations", where AI provides odd, wrong, or dangerous results (Lingard, 2023). Google Gemini has returned AI overview content suggesting using glue to hold cheese onto pizza, that Barack Obama was the first Muslim US president, and that children would gain health benefits from drinking wine (Choice Media Channel, 2024; Connolly & Branham, 2024). While they definitely don't pass the adult sniff test, we need to remember that those who are searching may not be adults.

So is Google search still 'good' search? I think it is still fairly good, but to me it is no longer the gold standard it once was.

I am now - mostly; firmly - a DuckDuckGo user. 

Sam

References:

Briscoe, S., & Rogers, M. (2024). An alternative screening approach for Google Search identifies an accurate and manageable number of results for a systematic review (case study). Health Information & Libraries Journal, 41(2), 149-155. https://doi.org/10.1111/hir.12409

Choice Media Channel. (2024, August 8). Why Can’t I Just Google It? [video]. YouTube. https://youtu.be/C6xnB7sfWj4

Connolly, M., & Branham, R. (2024, August 8). Why Can’t I Just Google It? [slide deck]. Choice 360. https://www.choice360.org/wp-content/uploads/2024/06/Why-Cant-I-Just-Google-It_Choice-ACRL-Webinar_080624.pdf

Gleason, J., Hu, D., Robertson, R. E., & Wilson, C. (2023, June). Google the gatekeeper: How search components affect clicks and attention. In Proceedings of the International AAAI Conference on Web and Social Media (Vol. 17, pp. 245-256). https://ojs.aaai.org/index.php/ICWSM/article/download/22142/21921

Johnson, T. (2001, June 23). You're not the customer, you're the product. Google Groups. https://groups.google.com/g/rec.arts.tv.interactive/c/EY-nOSMlAPE/m/GtKHp6qQBiUJ?pli=1

Lingard, L. (2023). Writing with ChatGPT: An illustration of its capacity, limitations & implications for academic writers. Perspectives on Medical Education, 12(1), 261-270. https://doi.org/10.5334/pme.1072

Von Bertalanffy, L. (1968). General System Theory: Foundations, Development, Applications. George Braziller.