The thing you can never get from an algorithm is that it doesn’t have a story behind it
Last month, HBO launched “Recommended by Humans” a handy website for figuring out which of its original shows to watch. But instead of using computer algorithms to sort through its vast catalog, the company hired people to make their case for each series by making video testimonials.
We’ve realized that recommendation algorithms aren’t as infallible as tech companies once made them out to be, and that handing the curation job to people still has value.
You read one article on homicide and very soon it feels like the world is doomed with similar news popping up everywhere in your feeds.
The trouble is algorithms do not understand what a great news article may be but judge it by its clicks and similarity to what you have shown an interest in and such articles get tons of clicks because of their sensational nature
Customers look for products across multiple journeys in our retail ecosystem
My mom would go into store ask a store associate where an item was located
I like to order online and pick up or sometimes get them delivered
My daughter just wants it to arrive right away
Search query volume is increasing year over year along with the nature of search terms
The industry is moving towards voice and visual search
Per Google Trends, 55% increase in queries looking for ideas and inspiration
In 2016, ~58% of Google searches were mobile and >20% of mobile searches were via voice.
Gartner estimates that voice and visual search will be at least 50% by 2020
Voice search has been the greater focus of development than visual; Google voice accuracy reached 95%, similar to human accuracy
Visual search more focused on discovery while voice search focused on immediacy of information
55% of US households by 2022 representing 70M households and 175M installed devices
Examples of Walmart queries:
how to cook roast beef
what walmart has bikes inner tubes
where to buy better homes & garden glider
when does the jewelry dept close
How do customers search on Walmart.com
Understand the intent no matter how it is expressed
Match the items in the catalog that match the intent – depends on catalog quality; goes for recall
Optimizes for engagement
Using Customer mindset/empathy to guide the customer
Precision vs. recall; entity graph
Build trust with the right experience and customer experience…..low assortment or non-negotiable……measure the impact to customer journey and use multi – objective optimization to reduce the gap between expected outcomes from customer and actual outcomes from algo
Design of experience and algo aligned with customer expectation – top 1
Expectation from visual search are taking search to a new level