The Price is Right

Pricing is a critical component of a B2B eCommerce site. There’s nothing more disastrous than a customer seeing the wrong price for a product. The emphasis should be on accuracy and processing speed.

It is common for B2B prices to be individualized based on factors such as contract, purchasing frequency, and volume. Sometimes bulk prices are displayed, grouped according to the number bought.

There are 2 main pricing situations that your search system will need to cater for. We will explore them and discover how best they can be handled:

1.    Price Groups

Rather than allocating individual prices for every customer, some businesses have levels or groups of prices, e.g. Level ‘A’ to ‘E’. Depending on the group level allocated to the customer, they’ll see one of a number of prices. In these situations, if the customer is identified, the search system can index all prices and decide whether to pass back, for example, the Level ‘A’, Level ‘B’ price. This situation works well when there are up to about 10 price levels.

2.    Dynamic Pricing

In this situation, every customer has negotiated prices, based on volume or contract etc. and the search system can’t know the price in advance. There are several ways in which a search system can handle dynamic pricing:

The search system indexes a standard price, and a message is displayed alongside the product on the search results page along the lines of:

“This is the standard price. For your pricing, view the product details.”

When the customer accesses the product details screen, the price for that particular customer is calculated and displayed. The advantage of this method is that price attribute filters, such as price sliders or selectors, can be displayed on the search results page, giving users the option to filter results based on the standard prices.

For each page of results, the ecommerce system calculates and retrieves the prices for the products that are going to be displayed on that page. Most sites only show one page at a time, approximately 10-12 products. Once the search system has passed back the product IDs, it can calculate the prices based on pricing rules, which can then be shown on the product results page. The search system uses a pricing engine to pass back appropriate prices. The disadvantage of this method is that any price attributes on the search results page become invalid.

An advanced, flexible search system can call a pricing engine as part of its process. This means that it can retrieve exact pricing before the attributes are calculated, passing back the correct pricing for each customer. This method allows for totally dynamic pricing. For every product in the results, a price is needed, which does mean that page load time can be affected. For on-premise B2B customers, this overhead would be reduced because the pricing engine will be running on the same internal network, or even the same server that the search system is running on.

Multiple Options

When displaying search results, there is sometimes the need for the price to be more than a single number. Sometimes a set of prices according to volume is required. The search system will need to be able to pass all prices back, perhaps displaying them in a table showing quantity vs price. Does your search system have the ability to pass back multiple prices?

Flexibility is the Answer

Every B2B business has differing needs, and it is essential that search systems are flexible to reflect this. EasyAsk can index multiple prices, calling out to a pricing engine if required, and pass back pricing structures in the results, as opposed to just a simple price. It may be that a business has a set of thresholds for discounts, or individualized pricing based on customer ID. EasyAsk adapts to all circumstances. This isn’t the case for all search systems.

Why Choose EasyAsk?

EasyAsk offers the only site search and merchandising tool designed exclusively for the rigors and challenges of B2B eCommerce. EasyAsk can be configured for any platform, either commercial or built in-house.

EasyAsk has long served the B2B customer segment and over 200 B2B distributors have chosen us to power their B2B eCommerce sites, including: Aramark, Demco, Kaman Industries, Tacoma Screw, and Crown Packaging.

The post B2B Series Challenge 4: Dynamic Pricing appeared first on EasyAsk.

No Results

According to a WebLink study, shoppers using internal site search deliver a 216% higher conversion rate compared to customers that simply browse. With search playing such a crucial role, generating a generic ‘Sorry, no results matched your search’ message could be disastrous. How many shoppers would stay on your site and how many would go straight to your competitor?

Check out these common causes for ‘No Results’ pages and how to avoid them:

Spelling Errors

A large number of ‘No Results’ searches can be put down to poor spell correction. An effective Search Engine must have sensible spell corrections and synonyms, including the ability to understand plurals. For example, the adjectives ‘smoked’ and ‘smoky’ should be understood by the Search Engine as synonymous.

Most Search Engines provide a reasonably good service when it comes to spell corrections or synonyms, but can yours manage word inflection, such as changes in tense or verb conjugation? For example, will your search system recognize that ‘mattresses’ is the plural form of ‘mattress’ because it has the suffix ‘es’ rather than just ‘s’? If your business operates across different countries, and therefore languages, this becomes an even bigger challenge. Do you have a Natural Language engine that understands these complexities across multiple languages?

Specific Language

Another cause of ‘No Results’ searches is shoppers’ use of language. It could be that a search is too specific, and that by ‘relaxing’ one or more words, a relevant set of results can still be shown. A good Natural Language tool will be capable of understanding that words such as ‘over’ and ‘under’ relate to a price indication. As people become increasingly used to using voice searches, they may include phrases like ‘show me’ or ‘I’m looking for’. Although typically searches would fail with the addition of these phrases, a Search Engine that can understand Natural Language will still make sense of the query.

You don’t sell particular products

When a shopper searches for something that is not stocked by an eRetailer, it may seem that a ‘No Results’ message would be inevitable. But a successful Search System will show results for alternative searches when a company doesn’t have what users ask for. The system should allow the business to set up redirects based on likely search queries (see our first blog post in this series to find out how to best use Search Analytics in this way). For example, if a company doesn’t stock Nike products, a redirect could be set up to show Adidas products (that they do stock) instead. Redirects allow businesses to redefine the terms a customer uses in order to point to different terms and therefore different products.

Products are Out of Stock

If a popular product has gone out of stock, most Search Systems will display a ‘No Results’ message. This is no longer a reason for a wasted opportunity with EasyAsk’s unique tool, ‘Conditional Redirects’. The tool enables businesses to set up a redirection to a separate set of products when their customers search for a popular product that is out of stock. For example, if a company usually sells Gracey Recliners, they may find that they sometimes have an issue with the supply, due to the product’s popularity. The merchandiser could set up a conditional redirect rule using EasyAsk software, so that when Gracey recliners are unavailable, a message displays explaining this, along with a redirected search for ‘leather recliners’. When the Gracey recliners are back in stock, they will once again be shown in the search results. It is not necessary for a merchandizer to remove the rule; the Search System will automatically show the product again.

No Results? No way!

Avoiding ‘No Results’ searches is a simple yet valuable way to ensure your customers find, and buy, what they are looking for. It is essential to use a Search Engine that is flexible and can understand the multitude of ways that a user might phrase a query. Quite simply, you won’t resolve the underlying causes of ‘No results’ without an intuitive Natural Language engine. With flexible merchandising and search tuning, you can make sure that you’re not hard-wiring the way that search works.

The post Optimizing your eCommerce site for the Holiday Season 2: Reducing “No Results” appeared first on EasyAsk.

What is a Chatbot?

Users interact with a Chatbot via a chat or messaging interface, imitating a conversation. After a series of questions and responses, users should arrive at an answer or product. Chatbots have moved on from their beginnings as a text-based response tree to now include elements of Artificial Intelligence such as Natural Language Processing. The most common use for Chatbots is for customer service. The Chatbot can answer the simpler customer queries, leaving the complex answers to the ‘living’ employees.

So Chatbots might be a good substitution for human customer service, but are they the next big thing for eCommerce sites? Chatbots can provide an intelligent interaction, but do shoppers really want to have an interaction when they know what they’re looking for?

Not a substitute for Good Search

Chatbots can provide a novel user experience, but when it comes to customers who are searching for a specific product or service, Chatbots cannot be a substitute for good search. A Chatbot is only as smart as the engine behind it.

Take the example of a customer looking for a leather sofa. Using a Chatbot, the customer may have to navigate through a series of questions to narrow down their search. They might select ‘Living Room Products’, followed by ‘Sofas’ and then be presented with a list of attributes, such as brand or size. Eventually, the customer may be presented with the option to select a fabric, where they can choose a ‘leather’ sofa. A Chatbot is typically pre-programmed with the order in which it will ask the questions and a customer may have to answer four or five questions before the most important question is asked. Compare this scenario to a great search system, where the customer can ask for a ‘brown leather sofa under $1000’ using the search box and see exactly the right products.

Chatbot vs Site Search

We thought we’d give eCommerce giant eBay a chance to wow us with their Chatbot capabilities before it stopped taking orders on September 6, 2018. We compared the same search with its regular site search results.

A query using the eBay ShopBot for ‘1967 Corvette Seats’ initially brought back the following responses:

It was necessary to click on one of the options regarding placement before results were presented. The first result shows seat covers rather than seats and the subsequent products available to view by clicking through the carousel were seat parts or tools, and not actual seats either.

The ShopBot then required a typed response to ascertain the desired price range. No clickable answers to save time, just the messaging entry field:

Once a price range was identified, the first result is indeed Corvette seat foam. Subsequent products in the carousel are seat brackets or assembly items.

The ShopBot then asked a seemingly relevant question about which type of seat was required.  On selecting ‘driver seat’, one result is returned, of a seat base for a 1993 – 2013 Corvette:

We think we can see why the ShopBot has had relatively little take up compared to the eBay mobile App. Contrast the Chatbot experience to that of using the regular site search:

Although the first result is a seat belt, the following results are 1967 Corvette seats, as requested, and the site search also provides the customer with a choice of products.

Even though the Ebay Shopbot appeared clever, it was still quite painful compared with simply running a search. If a shopper knows what products they are looking for, their choice is to either spend 2 to 5 minutes taking part in a question and answer session or to simply use intuitive search. Do shoppers want to have a chat with a computer or find the right products the first time they ask?

As Chatbots become more intelligent with advances in Natural Language Processing and Machine Learning, it would be easy to assume that they are a cure-all, but let’s not underestimate the necessity of great search. A Search Box query combined with intuitive Natural Language search technology will speed up a customer’s route to checkout. Instead of customers spending time interacting with a Chatbot, intelligent Search creates a faster track to a product or service, and therefore improves conversion rates.

Do you want EasyAsk’s advice? Don’t rush to deploy something that will bring down your conversion.

The post Chatbots: Online Domination? appeared first on EasyAsk.

(Message from the CEO of EasyAsk, Craig Bassin)

Looks like this is the beginning of the end for keyword search.  You’ve probably seen a number of articles discussing Google’s shift to ‘semantic search’.  Anyone understand what that REALLY means?  First, the definition of ‘semantic search’ is an understanding of the ‘intent’, or meaning, of the search, rather than just matching the keywords.

Now why would the undisputed 800-pound gorilla of keyword search, change course at this late date?  Conventional wisdom says they were forced to take a hard look after Apple launched Siri.  The timing sure seems to reinforce the fact that they’ve been playing with semantic search for some time, but needed to make a marketing splash now.

So, why change?  Well, obviously it’s a BETTER way to search and they had to, or they wouldn’t have!  I mean, really, Google acknowledging the limitations of keyword search?

Quoting from Paul Demery’s recent article (to read it, click here) about Google’s adoption of semantic search in Internet Retailer, ‘“Semantic search should allow Google as well as other search engines to better understand the true user intent of a search query,” says Kevin Lee, CEO of search marketing firm Didit.

Also, quoting from the same article: “Every day, we’re improving our ability to give you the best answers to your questions as quickly as possible,” Amit Singhal, Google’s head of search technology, said in a blog post. “In doing so, we convert raw data into knowledge for millions of users around the world. But our ability to deliver this experience is a function of our understanding your question and also truly understanding all the data that’s out there. And right now, our understanding is pretty darn limited. Ask us for ‘the 10 deepest lakes in the U.S,’ and we’ll give you decent results based on those keywords, but not necessarily because we understand what depth is or what a lake is.”

Now, understanding ‘intent’ AND ‘content’ is something that is at the very core of who EasyAsk is and how EasyAsk searches.  It’s the idea that, in an e-commerce setting, you can search for ‘men’s dress shirts under $30’ or ‘ladies red pumps size 6’ and get EXACTLY what you’re looking for.  Natural language understands the semantics involved in the search.  We understand the ‘intent’ of the question, we understand the ‘content’ of the data.  In adopting a new ‘semantic’ architecture Google will start to understand the ‘intent’ piece as well.

Now, who else searches this way?  How about Microsoft’s Bing, IBM’s Watson, obviously Apple’s Siri.

Now which of these companies can help you improve your e-commerce site?

None of them.

OK, but what about the other e-commerce search providers.  You probably know a few of them.  Endeca, SLI, Adobe, SOLR.

No, no, no and no.  Strictly keyword search.  Old news. Yesterday’s tech.

So we want to be the first to welcome Google.  We like them, use them all the time for internet search, along with Bing.  But when it comes to e-commerce search, folks, EasyAsk is leading the way.  Let us show you how.

It’s what we do.

The post Google joining IBM, Apple and EasyAsk? Pigs fly! News at 11… appeared first on EasyAsk.