Tag: machine learning


Relevance only ecommerce Site Search is passé. Increase revenue with Intelligent Search results

Posted on October 30, 2017 by


Humans have the attention span of a gold fish

Or less. Time Magazine & Microsoft reported in May 2015, that “The average attention span for the notoriously ill-focused goldfish is nine seconds, but according to a new study from Microsoft Corp., people now generally lose concentration after eight seconds, highlighting the affects of an increasingly digitalized lifestyle on the brain.” If this is the state of online visitors, you need to think before deciding what products to show your visitors, cause they will lose interest in 8 seconds and you will lose a sale.

Get Smart

Site search is a high value funnel for online visitors. Visitors who search are more likely to convert to customers than those who browse. Then why is your site search engine, dumb? Yes, you read that right, if you have a site search engine that spits out all products just because it matches the query (keyword relevance), your search engine is not thinking and you are losing on revenue.

Get Tagalys

Search engines like Algolia, Swifttype and others, are known for very fast search results. But what is the point in ecommerce, when you have 8 seconds to help your visitors find something of interest? Take a search query like “Dresses” or “Sofas”, these generic searches will contribute to over 605 of your search volume and each of these searches will results in over 100 products per query. What do you show? These queries are driven from mobile devices and now, you have only 2 products per scroll. Thats over 50 scrolls to see all the products in for that query.

This is why you need Tagalys, to analyze data, learn from it and predict what products are most likely to sell for that query. Show results for any search query, where products are sorted by the probability of sale. Tagalys Search also offers other value added services like Search suggestions, Popular Searches, Pinned Searches, Merchandising, but the core of what we do is determine sort orders for products to ensure what shows on top has a higher probability of sale.

Tagalys also applies the same data driven approach across all our API’s so you can ensure your visitors engage with the products that are most likely going to bring you a customer. We apply our data driven approach across Site Search, Product Recommendations & Product Listing pages. If you want to go one step further, we also apply personalization, which is the ability to understand what products are most likely going to appeal to each unique visitor.

Signup for a 28 day free trial and find out for your self. No credit card required. Let the numbers speak for itself.

 

Posted in eCommerce General, eCommerce Personalization, eCommerce Site Search
Tagged in , ,


eCommerce Site Search – Relevant, Trending or Personalized – Whats right for you?

Posted on February 1, 2017 by


Site search is a cash generation engine when done right for your ecommerce business. But assuming you do have an engine in place, what should it do for you? Search engines are aplenty and at Tagalys we offer 3 versions for retailers to choose from that will best suit their business needs. If we equate Tagalys to a Mercedes Sedan (a bias I have over the BMW), we can rank the technology behind our search engines as Relevant = A class, Trending = C class and Personalized = S class. The technology that drives the order of the search results is the biggest differentiator across these three versions, besides a few minor aesthetic differences. So let’s dig in deeper. Please note that the search results are universally accurate across all 3 versions, the stand out difference here is how product sorting is impacted in the search results. Interested already – SIGNUP NOW

Relevant search results = Mercedes A class

This is an entry level search engine, most suitable for retailers with less than 10,000 monthly visits.  Tagalys also offers a free plan for this engine. This version of the engine will still show you accurate search results for any simple or complex search query, but it does not have the ability to personalize results (visitor analytics) or sort results by what is trending (site analytics).  You might still consider opting for the C class or S class models, but based on our data analysis across existing clientele we have some average user engagement metrics that point out that this engine is most likely to give you better ROI that the other two. But if your average cart size ($50) or percentage of visitors who search (5% – 10%) far exceeds industry benchmarks, you might want to consider the other engines. We also acknowledge that if  you are smaller business you will be starting out with low mobile traffic, acquiring analytics data and visitors are getting to know more about your online store. This a great starter engine to help you tread the waters of online commerce.

Trending search results = Mercedes C class

This mid-segment engine is ideally suited for retailers above 10,000 visitors a month. Retailers at this size start showing patterns of in data as the volume of data show patterns on what products have the best engagement. With this engine, besides showing accurate search results regardless of search query complexity e.g., dress, maxi dress, Zara maxi dress, Zara cotton maxi dress, Zara cotton floral maxi dress, the engine considers historical engagement analytics across the site to determine the best sort order for search results that increases the probability of conversion. This engine should be used as a bare minimum to sort search results which are most likely to sell based on engagement data from the entire site. These results will appeal to your larger audience but not everyone because individual preferences are not considered in product sorting. Hence while this engine may cost more than the A class, it is more likely to convert more visitors to customers. Hence ROI for this engine starts to make sense when you have high traffic leading to high search volume.

Personalized search results = Mercedes S class

The top line of our engines is ideally suited for retailers who attract over 50% of traffic coming in via mobile devices and where this traffic is mainly composed of return visitors. Why? Return visitors have shown interest in your retail business. They will share hard and soft data on their intent and as a smart retailer, you can make sense of this data. If your return visitors are primarily engaging with your commerce store via a mobile device, there is a good chance almost 55% may bounce after the first 30 seconds (Refer Hubspot). These visitors get to engage with 2 to 4 products per page screen in a mobile device and in order to find products that are personally relevant to them, they are then expected to filter, sort, scroll and invest their time in product discovery, when you already know what their intent is.  How do you keep them engaged in the first 30 seconds?

Return mobile visitors is the key driver for search personalization, cause menus, filters and sort options are typically hidden from the mobile interface to reduce clutter. This user interface leads most visitors to scroll through the results and prefer frictionless experiences. And the longer a visitor scrolls to find a product of interest, increase the probability of bounce. So to keep these visitors engaged, the personalized search results from Tagalys, considers visitor specific engagement data to showcase results that are personally relevant to each shoppers persona. Objective is to generate visitor interest to click on a product, to engage. Please note that there is a direct correlation to product views and customer conversion rate. Our analysis with personalized search shows almost 68% of the clicks to a product page happen within the first 4 pages of search results. Hence showing the most relevant personally products, on top per unique visitor increases engagement. Increased product engagement from search results leads to higher search revenue.

There are many more fine details that differentiate the workings of each engine and there are a few cases not outlined here, where a mid sized retailer might still want to use a personalized search engine.

Interested already – SIGNUP NOW and see your visitor engagement improve.

Posted in eCommerce Personalization, eCommerce Site Search
Tagged in , , ,


Image Search vs. Text Search in eCommerce

Posted on December 26, 2016 by


At Tagalys, we extract text data from the product catalog to start data enrichment, hence we are biased towards text based tagging 🙂 But let us also share some of the advantages and disadvantages of the same so you can evaluate for yourself.

VERTICAL AGNOSTIC:

Tagalys can start working with an eCommerce company across any vertical and start showing improvements from day one. That is why we never hesitated taking on clients that sell apparel, shoes, accessories, jewelry, food and building material as long as they meet certain minimum requirements of scale. This is because, our engine relies on text data that is usually available in various parts of the product catalog and unlike image recognition we do not need any training data (vast sums), to reach a 90% accuracy in auto generating tags.

DATA AVAILIBILITY:

Image recognition products that are vertical focused extract tags from the image. The advantages are that if the engine is trained properly, it has the ability to extract fine details tags at scale with over a 90% accuracy. Hence an image of a dress can lead to extraction of external visible tags tied to color, length, sleeve, neck, design etc., Text tagging engines like Tagalys rely on the content, hence if such details are missing they will not be considered in the analytics of visitor-product engagement. But this also means, if Tagalys cannot find such tags, nor can your online visitors or search engines like google that rely on text data to understand the product better. Hence it is in every retailers best interest to ensure all details about the product are clearly outlined in the product catalog to promote improved product discovery both in the website and also in search engines.

The second question is how much of an added benefit will the retailer have in terms of engagement/conversion/revenues by auto extracting tags from images? As mentioned below, costs for such a high accuracy image extraction engine are higher than text tagging, leading to higher prices for the customer.

COSTS:

As text based tagging does not require vast sums of training data to get started, but on available data our cost are much lower than image recognition companies.  The costs for an image recognition company are not only acquiring vast volumes of images to build a database but also the time it takes them to create this database as that impacts the NPV of cash on hand. As our costs are lower, these savings are passed on to you, the customer.

CORE COMPETENCY:

If a retailers problem statement is to improve personalization or showing the right products/content to the right visitors, then you need personalization engine like Tagalys, whose focus is to create data models that simulate the shopping persona of visitors. But if your problem statement is to auto extract tags from images as there is no content, then a visual recognition engine is what you need, provided it has been previously trained with millions of images in the same vertical.

APPLICATIONS:

A personalization engine like Tagalys can be applied across any channel where a visitor is likely to discover products. This could be a listing page, search results or a recommendations widgets.

Visual or image recognition engines, whose core competency is to identify the image and the meta data (tags) from the image is best applied in the Similar products recommendation widget. That being said, we recently worked with a jewelry retailer who switched from an Image Recognition to Tagalys recommendations and realized cost savings in monthly fees and higher revenues driven by better conversion from our recommendation engine.

Image recognition is also confined to verticals where there is enough data to improve the auto tag accuracy levels. If they need to be applied to any other features of eCommerce, then a personalization engine has to be built on top of the existing visual/image recognition engine.

ERROR TOLERANCE:

A combination of image recognition to extract the tags and human tagging, might help reduce the error tolerance. Product catalog data that is a must have for an ecommerce platform and is usually entered by humans. This task sometime has errors or incomplete submissions, that can impact the pattern recognition algorithms in Tagalys. In over 90% of the cases, we find the data to be complete with some minute details missing by the data entry operator, but very seldom incorrect. This is also because incorrect data can negatively impact SERP score, as the content displayed on any listing page is based on the text data behind the product. Hence while we have seen good data from the catalog itself, adding an image based system to identify tags in verticals like fashion, can extract additional data that can be leveraged by Tagalys.

Try Tagalys for a 28 day free trial, to see the results, before you decide to spent 3X more on image recognition.  SIGNUP NOW

Posted in eCommerce General, eCommerce Site Search
Tagged in , , , ,