We recently met the team from Klevu, a sophisticated search technology company.
Klevu is an NLP-based (natural language processing), self-learning e-commerce site search solution, designed for enterprise-level and mid-level online retailers. Founded in 2013, Klevu is used by over 3,000 merchants, from all over the world. Klevu’s primary objective is to enable online stores of all types and sizes to have the most advanced and powerful, yet affordable search solution.
The majority of Klevu’s clients are using the Magento platform and they have an existing integration with both Magento 1 and Magento 2. Some of Klevu’s clients include Agent Provocateur, Oneills.com, Jack Daniels, Yamaha and Zimmermann.
Some of Klevu’s core features include:
Natural language processing – Klevu’s natural language processing capabilities represents one of their core benefits over their competitors, as they go deeper into understanding the query and the intent behind it. An example of this could be a user searching for “floral drapes” when a product is listed as “curtains with flowers”, Klevu would understand the similarity in the query and serve the correct products.
Automatic catalog enrichment – Another key feature of Klevu is around how they enrich the product catalog to add further context. Klevu take as much data around the product as possible and then add synonyms and additional adjectives and similar terms in order to match far more queries. This is one of the key features for stores with larger product catalogs and is a big part of how they handle synonyms and errors / mistakes.
Content search – Klevu also allows merchants to serve and promote content as part of the search function, allowing for queries like “delivery information” or “terms and conditions”, as well as enabling users to serve things like buying guides for more generic queries.
Self-learning capabilities – Another very important feature for users is Klevu’s ability to self-learn – meaning results are automatically optimised based on how users are engaging with results and purchasing behaviour.
Advanced merchandising and boosting – Klevu gives merchants far more control over which products, attributes and categories are being promoted for different queries, by providing boosting rules. Users can simply add different weighting scores against each product, category or attribute in order to boost their visibility.
Instant search layout – Klevu’s instant search box helps to improve the user journey and speed up the product discovery process, as can be seen below. When a user starts inputting a query, results are served via the instant search box far quicker than if they were to go to through to the results pages.
This example also shows how faceted search can be used to improve the search experience, which can also be used on the results page.
Fully customisable layout – The layout of the instant search box and the results page is completely flexible and the CSS can also be edited from the Klevu admin interface.
Advanced reporting suite – Klevu provide comprehensive reporting around your search function – including detail on the keywords being used, the location of searchers, the keywords driving sales and common errors that could be optimised.
Millions of customers can now login and pay with the payment and shipping information stored in their Amazon account. Join thousands of merchants and start building a strong, loyal customer base. Amazon Login and Pay with Amazon makes it easy for customers to buy what they want, and get on with their lives. Forget countless usernames and passwords — your Amazon account details are all you need.
Amazon Payments is available as a Magento extension/module, reducing time-to-market & development costs.
Here’s a video which positions the offering.
Here are some example Magento sites that have implemented Amazon Payments:
Since the dawn of the Internet, people have tried to abuse websites for both sport and profit – especially e-commerce websites which by definition possess a number of customer data entry points.
As the abuse became profitable, the scale of abuse grew using automated software (sometimes referred to as”bots”). To prevent bots from overrunning sites with spam, fraudulent registrations, fake sweepstakes entries, and other nefarious things, publishers responded by testing users to see if they were human or not. These approaches are typically referred to as CAPTCHA, which per Wikipedia, is an acronym for “Completely Automated Public Turing test to tell Computers and Humans Apart”) is a type of challenge-response test used in computing to determine whether or not the user is human.
Generally, e-commerce websites would use CAPTCHA to strengthen the security around the most sensitive account access points. Typical access points on an e-commerce website would include:
Sign up for a new account or register
Sign up for email newsletters
Request a catalogue
Login to make a purchase
Get a quote
What happens – if you are unprotected – is that your customer and email databases end up being mass-populated with spam email addresses: it is costly and time-consuming to weed these out.
Google’s definition of the issue reads as follows: “CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) is a type of security measure known as challenge-response authentication. CAPTCHA helps protect you from spam and password decryption by asking you to complete a simple test that proves you are human and not a computer trying to break into a password protected account. A CAPTCHA test is made up of two simple parts: a randomly generated sequence of letters and/or numbers that appear as a distorted image, and a text box. To pass a the test and prove your human identity, simply type the characters you see in the image into the text box.”
Rather than use the conventional “old school” approach which produces notoriously difficult-to-read squiggly text, Screen Pages has recently deployed a “choose the correct image” approach which effectively rules out bots (well, until the AI in bots means they can interpret visual content. On Donald Russell for example, you can need to identify the steak (for example, on its email newsletter subscription page: https://www.donaldrussell.com/emailsignup)
Google has also released a new method called “reCAPTCHA”. “reCAPTCHA is more adaptive and better-equipped to distinguish legitimate users from automated software: this updated system uses advanced risk analysis techniques, actively considering the user’s entire engagement with the CAPTCHA—before, during and after they interact with it. That means that today the distorted letters serve less as a test of humanity and more as a medium of engagement to elicit a broad range of cues that characterize humans and bots.”
Here’s Google’s helpful intro to the subject and it’s new approach in a video:
We’ve already implemented this on one website – on Watco’s B2B “get a quote” facility.
For information, implementing some form of CAPTCHA facility on this important data capture forms is generally less than a day’s work: a small investment compared to the pain of purging all those spam email addresses from your customer databases.
We recently met with the folk from Nanorep, an impressive online customer service hosted software application with a well-crafted Magento extension.
Nanorep is a “self-aware support system” that is capable of learning the answers to repetitive questions and updating its knowledge base software in real time. This innovative & automated software approach to self-service can dramatically improve the friendliness of e-commerce sites with resultant cost savings, increased conversion and improved customer satisfaction.
Nanorep widgets on any page allow to you to ask questions in self-service mode across multiple channels. There is no need to wait for responses from customer support agents – via email, telephone or forums. The responses provided by nanorep are instantly available and contextually correct – queried by “natural language” free-form questions/searches with responses driven by its clever search technology. Its knowledge base is “self-learning”: it instantly updates with questions and answers and learns in real time what is statistically relevant. If answers are missing or don’t hit the spot, you can enter answers directly through Nanorep’s web-based console.
As a website customer – on the front-end – you can easily use the widget to find answers to all product-related queries on site. Because each customer that visits your e-commerce website is theoretically unique, individual responses are based on user type, product,location, technical specs, stock, care & maintenance etc. So, the answer to a question depends entirely on the context in which the question was asked.
The Nanorep widget
Shoppers can ask any kind of question on any product page and get instant answers which help them to make up their mind. nanoRep addresses three specific types of question:
This does not refer to a specific product, nor to any action that the user may have taken. It is simply a general question — maybe about your company, or the process of ordering something or about who the management are. The answers are automatically retrieved from the nanorep Q&A knowledge-base.
These refer to a transaction (in general), either a product that has been bought or that is being considered to be for purchase. The answers are automatically retrieved from your Magento product catalogue data base.
The personalized question refers to a specific transaction, such as an order placed, or a service which was disrupted or anything else that happens to the individual customer. The answers are automatically retrieved from your Magento database.
Consider the following questions: How much does shipping cost? Is this available in my area? What payment methods do you accept? Each of these questions requires a specific personal response. The cost of shipping is not uniform; it depends on the product, on availability, on geo-location, and even the user.
Question topics automatically retrieved from the Magento database include:
Orders statuses (status, purchase and shipping dates)
Further, all nanorep self-service functions are available in multiple languages and is fully optimized for all mobile devices and smartphones.
The functionality of the Magento extension includes:
Automatically pulls personalized ordering information from the Magento database.
Automatically pulls products attributes information from the Magento product catalog.
Automatically installs the nanoRep widgets per product, as well as the API and back office.
Automatically installs an SEO-based knowledge-base for every product.
Shoppers can ask personalized questions about the status of their shipping, billing and orders, and get instant and accurate responses.
Shoppers can ask their questions in any of 64 different languages, and nanoRep will automatically provide translated answers.
The Nanorep module usefully includes analytics and reporting that show real time insights to the mind of the shopper: making it possible to know precisely how NanoRep has delivered value:
Merchants can see every question ever asked by their shoppers, filtered by product or category, and use this data to optimize the knowledge-base with more instant answers for essential information.
See which products get the highest number of attention by the shoppers with the highest level of shopper interest.
See the list of the last three questions that lead to every purchase.
See which products are being asked about most frequently.
Actionable BI that shows which products receive the highest number of unanswered questions, allowing you to optimize the knowledge-base, based on the most needed information and especially on the most relevant products / categories.
There’s a great example on Zaggora, a Magento website.
Magento has released the newest version of Magento Enterprise Edition 1.14.1 and announced Community Edition 1.9.1.
Here’s a summary of the new capabilities:
Visual merchandising tool for product categories
Magento Enterprise Edition 1.14.1 enables retailers to organize products on category pages dramatically faster and easier. The new drag-and-drop visual merchandising tool lets merchants preview what a category page will look like to shoppers and adjust product positions with the click of a mouse. This lets them test and iterate to create more visually appealing pages and more easily make real-time updates in response to changes in consumer demand and inventory levels.
The visual merchandising tool also saves merchants time when establishing or updating categories. It gives them the ability to bulk-add or remove products, set up rules to automatically assign products to categories, and create dynamic categories (such as “new products” and “on sale items”) that pull in products from other categories based on product attributes.
Other new merchandising features, including the ability to consistently fix “hero” products at the top of a category and automatically sort products according to special rules, help to further optimize category sales.
New “swatch” capabilities empower merchants to show off a broad range of product attributes to boost conversion rates. Richer at-a-glance product information, like available colors, fabrics, and sizes, improves consumer confidence to proceed with their purchase.
Better collaborative support in the Admin Panel
This makes it easier for multiple administrators to make product updates at the same time, without diminishing performance.
Long list of other enhancements
Magento Enterprise Edition 1.14.1 includes over 70 updates to related products, payments, security, and other features as part of our commitment to continually improve product quality.
Updated Google Analytics
Magento Enterprise Edition 1.14.1 has been updated to support Universal Analytics, the new standard for Google Analytics. With this update, merchants can define more custom dimensions and metrics for tracking and gain access to ongoing feature updates that will only be available on Universal Analytics.
Here’s a quick summary of what Apple Pay can mean for payments (courtesy of Retail Dive):
1. The fact that leading banks and credit card companies are partnering with Apple is a sign that the shift to mobile payment systems is getting underway.
2. Card numbers are never sent to Apple servers or retailers. They are stored instead in the new iPhones’ Secure Element, a built-in security method.
3. The payment system will only be available on iPhone 6, iPhone 6 Plus, and the Watch.
4. Apple Pay will be available at 220,000 retail locations starting in October. But when compared to the over nine million U.S. retail locations that currently accept credit and debit cards, Apple Pay has some catching up to do.
5. The payment system will only be available on iPhone 6, iPhone 6 Plus, and the Watch (when the Watch comes out next year). It also relies on near field communication (NFC) technology, something only 10% of retailers currently have.
6. If customer preferences really do change from plastic to mobile, the days of credit cards could soon be over for retailers.