Awful title, I know.
After reading James Gurd’s post on Econsultancy, I wanted to jot down my thoughts on how, ideally, I would integrate integrate personalisation technology into an ecommerce business.
This post is probably a little incomprehensible and idealistic – but hopefully, these make sense.
5 steps on the road to success
Cringeworthy heading. But starting small whilst you get comfortable with personalisation technology, taking each step at a time, will allow you, your team and your business to grow more confident when dealing with personalisation. Often, people are scared and aren’t willing to take the plunge from the tallest diving board, but helping them to climb the ladder and pointing out how far they’ve come will help them make the final jump.
The five steps on this road are
I’m going to go through each one – probably briefly, some more indepth.
But first… Continue reading
Originally posted on Quora
There are a number of things that affect the performance of recommendation systems, but the biggest question is what is the right metric to track and evaluate performance by?
The most obvious answer would be conversion rate – the percentage of all “visits” to a site that turn into a “purchase”. However, this metric isn’t really that great for monitoring the performance of a recommendation engine. Continue reading
IntentHQ have put out a slidedeck explaining why personalization based on what people actually do is worse than what people say they do. Continue reading
In our constantly connected world, a device is just a proxy for what really matters — getting to know your customers. Devices provide context, helping us learn what matters to a consumer in a particular location and at a particular time. Coupled with the intent provided by search, this is incredibly powerful. It can help retailers deliver relevant suggestions, essentially recreating those shopkeeper conversations at scale. The right message at the right moment is the next level in customer service — it can quickly and easily turn intent into action.
Context also allows retailers to better than ever anticipate what a customer might need based on when, where and how they arrive at their site and help them decide how to respond to them. People are constantly looking for product information, deals, local availability and local discounts online — and retailers who aren’t there to supply the right information when people raise their virtual hand will lose out.
via Shopping Then and Now: Five Ways Retail Has Changed and How Businesses Can Adapt – Think Insights – Google.
“I never had to look for growth. Tesco grew by 10% a year in an industry growing at 2% because the voice of the customer gave the direction.”
via Internet Retailing » How retailers can, and must, learn from data: interview with Sir Terry Leahy.
Absolutely agree with this, from analyst Susan Aldrich. Continue reading
Reading this article (On Algorithm Wars and Predictive Apps – Datanami, Forrester analyst Mike Gualtieri talks about four principles to distinguish between a regular app and a predictive app – but I think these are really prominent not just to apps, but to experiences across any channel. Continue reading
Commerce Sciences has raised $1.8m from investors, including Eric Schmidt’s Innovation Endevours. But what does it actually do? How does it work? Is it actually any good? Continue reading
Originally posted on Quora.
It depends what you’re trying to do.
If you’re looking to power product recommendations on an eCommerce site, then there are many companies who can help you do this. They range from entry-level (e.g. Barilliance) to mid-level (e.g. Peerius) to enterprise (e.g. PredictiveIntent, RichRelevance). Here are some key points you should be looking for. Continue reading
Amazon leads as paid search advertiser on Google UK | The Drum.
So Amazon is Google’s biggest advertiser. Many people are saying they don’t necessarily need to be buying the traffic.
But, I think what they actually want is the data that this brings. Data in the form of “what visitor X is looking for” – which can be used to power product recommendations on the Amazon site (if you start browsing around), or by email (if you’re an existing customer).
The email example is a powerful way to retarget customers. Amazon know that a customer (who has previously purchased X, Y and Z) is now interested in product B or something from category 3. If the customer arrives from a Google ad and starts clicking around – great. If not, they have still captured this behaviour and can use it in trigger emails to reach out perhaps a day or two later with “people like you also bought” recommendations – potentially recapturing a sale. This isn’t even cart abandonment – it’s visit abandonment.