First though, let’s take a look at the UK’s top 12 toys for Christmas 2014, according to the Toys Retailers Association

For some of you this may be just a random collection of barely recognisable words, to others these are products haunting your every waking moment…

  • Bop-It! Beats
  • Frozen Snow Glow Elsa
  • Frozen Sparkle Doll
  • Kidizoom Smart Watch
  • Little Live Pets Bird Cage
  • Minecraft Figures
  • My Friend Cayla
  • Nerf Demolisher
  • The Lego Movie: Benny’s Spaceship, Spaceship, SPACESHIP!
  • Toot-Toot Animals Safari Park
  • Transformers Stomp & Chomp Grimlock
  • Zoomer Dino

I can crack as many jokes as I like, but if I don’t wake up to Benny’s Spaceship on Christmas morning I’ll be throwing the hissy fit to end all hissy fits. 

On with the search…


As you’d expect Amazon has a comprehensive and fairly intuitive search. I only have to begin typing ‘bop it’ and the product ‘bop it beats’ appears straight away… 

However if I just search for ‘bop it’, the supposed most popular model doesn’t appear on the search results page till halfway down.

There seems to be a lack of continuity in relevance here.

Similarly if I type ‘elsa’, the ‘elsa snow glow’ doll appears at the top…

… however the ‘elsa’ search results page shows the snow glow doll in second place, under the other popular Frozen toy this year.


These are very minor inconsistencies that won’t really matter too much when it comes to each product’s discoverability.

If I type in ‘benny’s’ the LEGO movie spaceship toy appears right at the top (as opposed to…uh… the disturbing Michael Haneke film Benny’s Video).

However ‘lego’ is too broad a search term to bring up one of its most popular products.

Similarly the brand name ‘nerf’ doesn’t reveal ‘nerf demollisher’ till quite some way down the suggested items. 

I do now however have a Nerf Zombie Strike in my Amazon wishlist… FYI.

Toys R Us

When searching for ‘my friend cayla’ the doll doesn’t appear till the fourth result….

However I am taken directly to its product page, which should help aid conversion.

For other items, Toys R Us’s search is strangely illogical. Here is my search for the massively popular ‘minecraft’. It appears under ‘mickey mouse’ and ‘microphone’.

I assume the automatic suggestions are arranged by popularity. Historically people probably have searched for the other terms more than the one I’m typing, however I imagine more people have searched for ‘minecraft’ in the last couple of years than ‘mickey mouse’ or ‘microphone’. Maybe I’m wrong. However it’s important to remember when building search algorithms to identify recency as well as popularity.

Here’s another inconsistency: ‘mine craft’ versus ‘minecraft’. Minecraft is the proper spelling, but misspellings should always be considered in all good search tools. However the problem here is that Toys R Us has split its products between the two spellings. 

When searching for ‘Transformers Stomp and Chomp Grimlock’ the item doesn’t appear in the automatic suggestions.

It’s also not at the top of the list when searching for ‘grimlock’.

Then on the search results page, one of the most popular toys this Christmas is featured as the second listing.


Argos doesn’t have an automatic suggestion tool in its search bar. Which is very disappointing.

When searching for ‘toot’ (as in Toot-Toot Animals Safari Park) I’m taken directly to the manufacturer’s landing page, where the product I searched for is hidden amongst some other product page links.

If however I search for ‘Toot-Toot Animals Safari Park’ I’m taken directly to the product page.

I imagine most people wouldn’t search for the whole term though, so let’s break it down and try ‘animals safari park’.

Here you can see that the item I’ve been looking for is the very last product on the search results page.

Tesco Direct

As soon as I’ve typed ‘f’ Tesco Direct has suggested ‘frozen’ to me. It knows exactly what’s popular this year.

Then when I continue typing it brings up the exact toy I’m after.

Searching ‘kidizoom smart watch’ leads to a similar success.

Then by searching for just ‘kidizoom’ I’m presented with this clearly laid out product listing page with the two smart watch variations presented on the top row.


Before I’ve even typed the third letter I’m presented with ‘zoomed’ as the top query and the actual product I want in the popular products section beneath.

The same goes for ‘little live pets bird cage’…

Then just to mix it up, let’s try a variation…

Success! Then the same goes for ‘bop’… I’m presented with the most popular model in an obvious and clear way.

Even the very generic term ‘animals’ brings up the Toot Toot Safari Park I was looking for.

And just to test the product results pages, here you can see the most popular version of Grimlock this Christmas is the first listing.

I’ll talk more about how brilliant this whole experience is in the conclusion.

John Lewis

John Lewis’s search bar gives automatic suggestions to a point. It will recognise ‘nerf’ as I type it, but it won’t offer any products within that keyword.

Then after searching for ‘nerf’, its most popular product is nowhere to be seen on its search results page.

Similarly I’m only given three suggestions for ‘lego spaceship’, neither of which I want.

However when searching for ‘lego spaceship’ I’m presented with two items that didn’t come up in the automatic suggestions.

‘My friend cayla’ doesn’t seem to exist here… 

And horror of horrors, neither does ‘snow glow elsa’.


Just for variation, let’s see how the UK’s most venerable and famous toy store handles site search.

There is no automatic suggestion tool unfortunately and then when I search for ‘snow glow’ the Frozen doll doesn’t even appear on page one of its product results page.

When searching for ‘elsa’, one of this Christmas’s most popular items Sparkle Doll’ doesn’t appear till the very bottom.

In conclusion…

Many of these retailers have a specific algorithm that suggests products mainly based on ‘popularity’, then arranges products on a results page under the vague notion of ‘relevance’.

When it comes to what’s actually been popular over the last few weeks, or what is forecast to be popular, that’s where many site search tools fall down. It’s important to take into account recency and serve customers with results that reflect current trends accurately in a more intuitive and agile way.

Just look at Smyths as an example. Although historically popular search terms appear as automated suggestions in the ‘top queries’ section of results, products that are popular right now are also included. It also helps that these are in a separately headed ‘popular products’ section and include thumbnail images to really draw the attention.