Posted on 8th November 2023
The Search Engine: A product design lifecycle
There it is - Google's homepage circa 1998. A still recognisable colourful logo accompanied by the all important search box that was to become the de-facto gateway to a myriad of information. For over a quarter of a century the company have dominated the general purpose web search market. But over time the design of the search engine has morphed to become a much more controversial beast following the modern paradigm of 'free' online services: To generate huge advertising revenues, information about the user became the product, and the functionality became a secondary by-product of this vast operation specialising in the mining of personal data.
Here's an early search results page from Google. Key features are:
- The search box
- Advanced search
- Categories
- Search context bar (showing the query, number of results, pagination info)
- The web page results
It was a simple and functional design. Not particularly aesthetically pleasing but many will remember it for Google's killer feature (its ranking algorithm and sheer amount of pages indexed). It was the fastest way to find information online. It worked and people appreciated the simplicity of it.
Things started to evolve and Google's home page offered useful categories of information that would allow the user to filter their results. Several iterations of design changes made it easier to use and many people regardless of their experience with computers found it made the internet much more accessible to them. But there was always the temptation to build more features and implement tricks to cement their position as the dominant search engine.
In the mid 2010s the results page shifted and changed repeatedly with new elements creeping in and crowding out the original feature that was king in the early days: A simple list of search results. In this example you have to go below the fold just to see the second actual web result. An annoying pop-up asks the user if they want to switch their default search engine to Google. And behind the scenes they are most certainly harvesting user data at this point without their consent.
Many would argue from this point forward it just kept getting worse:
Design analysis
Let's ignore the commercial reasons for now and the fact that by this time they seem to be breaking their 'don't be evil' company motto, and instead look at the design changes that have taken place.
Too many things
It's clear increasing numbers of features have been added to the results page. What's interesting is how isolated those seem to have become both visually and functionally. For example it really isn't clear that the category buttons at the top filter the results. Their proximity to the search results needs increasing. There is also a huge amount of wasted space on-screen. Adding all these new features has negatively impacted the layout of the search results, leading to a confusing experience where a user must hunt around the screen to find what they are really looking for.
Feature duplication
Do these drop down options look familiar? That's because they are duplicating the filter buttons across the top. I've no idea why, but this isn't the first time I've spotted feature duplication like this in modern websites and apps.
Too many icons
Icons definitely have their place, where a key feature can be quickly and concisely illustrated without the need for text. To me the important icons to keep are the X to clear the search and the magnifying glass for the search button. The big mistake here in my view is the confusion caused by the addition of a microphone and something resembling a camera inside the search box. I know exactly what those icons do, but I doubt my 70 year old mother who has been effectively using Google for decades does, nor does she probably care about those features. The problems here are ones of poor visual hierarchy and relevance to the primary usage. I know 'search by voice' has a big accessibility benefit, but I'm not sure cramming that and 'search by image' in the search box before the magnifying glass is the best approach.
Conclusions about the design
A very long essay could be written about the pros and cons of this design. But to try and summarise my impression of the current Google search results page: It isn't working well and is a regression in terms of Google's capability to fulfil the original purpose of being a search engine. Going further I would pose the question what does it say about Google as a company? To my mind this is a business with a significant identity crisis. I feel this is a perception that could be solved by a back-to-basics UX approach.
My modern search engine design
I'm not going to attempt to redesign the Google results page in its entirety. As I've already alluded to I think some significant user research is required before that task is embarked upon. But I thought it would be a useful exercise to try and create a new design for a generic search engine that is fit for modern times. Here's a couple of quick sketches of how it could look:
Some ideas that I had which would need a few iterations:
- Move the login / account related features near to the logo of the company: proximity to its purpose and also no cog or 'hamburger' icons which might get missed.
- Putting the text based information that augments the query with useful information on the left hand side which is a wasted space on Google
- Simplify the categories displayed: limit it to about 5 maximum. This is an extreme example where we've relegated everything apart from web, image and video. A lower number of categories shown at once also enables you to have icons representing a result from each category so when you are on the 'all' section you can see what type of result each one is.
- A drop down for the text/audio/visual search options.
- Going back to old-school basic links for 'similar queries'. These could be powered by algorithms like the 'people also searched for' and 'relevant queries' features in Google's search, but the user doesn't really care where they come from and it's confusing to have too many different things that pop up depending on the search. Aggregating things like this in simple sections of the page where the user will learn to expect them lessens the cognitive load of the page structure changing regularly.
The end, or is it?
Well certainly as far as Google's days as a genuine search engine it could be argued it has reached the end of its lifecycle. Though it could also be argued if the concept of a search engine is relevant today. Without proper investment into examining the user experience it's hard to draw any conclusions. As far as my design goes, clearly it needs a lot more variations before converging on anything useful. Ultimately tweaking the design of a search engine without proper user research won't yield the best results. But I might try and take it further and create a prototype at some point just for fun.
To be honest I still can't get over just how simple and effective the original Google search results design was.
If you enjoyed this, you can read the next post in this series here.