Every few weeks we’ll be focussing on one of our amazing partners - asking questions about the industry and getting to know more about their product.

This week we’re focusing on Klevu.

In our minds Klevu is one of the best product discovery tools available for consistent ROI for clients. Speed of implementation, fast learning curve for merchants, and an extensive roadmap of enhancements make this a must-have. Plus they have an amazing team of lovely people who are passionate about what they do!.

We sat down with Rachel Tonner, Global Head of Marketing, to learn about Klevu and its products, as well as getting some insight into eCommerce challenges ahead and their expert advice for merchants.

Who is Klevu?

I am going to answer this a bit differently than expected. Klevu is actually a group of people across several continents who collectively speak more than 15 languages. We care deeply about each other, our work, and our customers. It’s this entrepreneurial spirit and dedication to improve and innovate that drew me to Klevu in the first place, and that’s also something I hear constantly from our customers and partners. So, regardless of what Klevu sells, the company will succeed by putting people first - we all want to make sure every customer reaches the best possible results, and we want to help each other grow.

If you could describe Klevu’s product in a paragraph, what would you say?

Klevu is a composable microservice for eCommerce on-site search, category merchandising, product recommendations, and personalization. It uniquely leverages AI, Natural Language Processing and User Behaviour Analytics to elevate the search experience, and automatically re-merchandise category listing pages and product recommendations.

Your latest Benchmarking Report has some really interesting findings. Which part stood out to you the most? And how can people get hold of the report?

Absolutely! What stood out to me is that there is an obvious product discovery gap. Shoppers get inconsistent experiences with brands, and yet the brands don’t know it. They blame other things - like recession, stock issues, etc for high bounce rates and lower conversion, when actually, fixing some basics in product discovery can drastically improve conversion.

One of my favourite insights from the benchmarking report was when we did a search on Hotel Chocolat for ‘chocolate’ the website produced no results. You can’t make this stuff up.

The Product Discovery Benchmarking framework takes into account how well the search engine works on-site on eCommerce websites -- natural language processing, misspellings, usability. It also then tallied up how product recommendations were displayed, and with which strategies. We look at personalization and whether or not behavioural personalization is fully utilised, as we find that it's best to not only look at past purchase history but also add in the behavioural layer and use intent data driven by the search engine. (This also helps not to put your shoppers in a filter bubble, or drive them away with recommendations based on a potentially biased machine, but I won’t dive into that massive topic here!).

The research then looks at general usability in the product discovery journey, such as the depth of filters on product listing pages, and how much information about products is surfaced there.

The features that can get brands to the top spots in the research can be implemented over time, and Klevu can help speed that up. For more information on the research, including case studies and statistical results, download the report.

B2B eCommerce is seeing rapid growth, is Klevu well placed to serve the B2B market just as it does B2C or DTC?

B2B eCommerce is definitely getting more fun to talk about. For B2B retailers, it’s no longer enough to simply have a functioning eCommerce website. Instead, you must be able to replicate the same B2C website experience that shoppers are familiar with in their personal lives.

They want robust product information. They want advanced search capabilities, so they can quickly find the specific products they’re looking for. And of course, they want to do it all from their smartphones.

Part of this change in expectations can be attributed to the fact that there are more millennials in the workplace. In 2020, close to half of B2B buyers were millennials — nearly double the amount from 2012. Millennials are also much more involved in making business decisions. A recent report found that 44% of millennials are making purchasing decisions while 33% say they are key influencers or recommenders in the purchasing process.

Klevu has recently been included in the Gartner Market Guide for Digital Commerce Search. In it, Klevu scored amongst the leaders in B2B product discovery because of the AI technology that enriches product catalogues with searchable data automatically, the ability to handle SKU searches with grace, and its inbuilt automated dimensional transformation which means that shoppers can search and filter by measurements easily, no matter if they use inches, millimetres or centimetres, for example.

Some of our B2B customers are even utilising Klevu’s merchandising tools to turn their eCommerce site into a lucrative revenue stream, selling ad space or sponsored product positioning to suppliers.

Learn more about Klevu and B2B here.

New challenges hit eCommerce every year, what do you think will be the biggest challenge facing merchants going into 2023?

Analytics will become a hot topic and high up on the project lists. We have already begun projects to help our customers improve analytics - connecting with GA4, reviewing how we get and give data, making that scalable. This becomes even more important within headless eCommerce tech stacks where businesses will likely centralise data into one tool.

Cookies, privacy, personalization - Profiling shoppers by demographic data is not only risky because of the ever-changing legal requirements, but also can be biased. Today, people have many preferences that don’t fit within the standard binary,  which many businesses still use to make assumptions about their customers. Klevu believes that personalization doesn't require personal data and that behaviour and action on a website speaks louder (and is more sustainable) than profiling and segmentation.

Margin, margin, margin - it’s tough out there. Post-covid slumps in profit, looming recession... There are so many factors that could prevent stasis, let alone growth. Klevu merchandising tools help retailers to merchandise based on any available attribute, so many customers are using Klevu to boost products with higher margins.

Are there any other industry tips you can offer merchants to help with product discovery?

Yes! Thanks for asking - I love helping people with some quick wins, whether you use Klevu or not:

  1. Make your search bar bigger, and expose the actual search box.
    Over 40% of shoppers go immediately to the search bar when on a new website, this is especially relevant on mobile. Exposing the search bar on your website can lead to a 44% increase in orders, especially on mobile.

  2. Make sure typos can still drive revenue.
    Using an automated tool like Klevu can help, but if you’re not, make sure you’re reviewing your ‘no results’ terms multiple times a day. Shoppers will abandon a website if they can’t find what they are looking for, even if they misspelt it! For 56% of retailers, their search experiences are just not surfacing stock they have because of simple typos. Processing natural language, including mistakes, is vital to improving conversion rates.

  3. Add auto-suggest on search overlay.
    For an optimum experience, show rich content within the search overlay including product recommendations, text, images, price, offers, and content - even on mobile. Doing this increases search-led conversion by 54%.

  4. Add natural language processing to search.
    NLP helps search engines process what shoppers mean, rather than just what they search for. Adding this will result in measurable revenue gains - an increase of 12% in search-led conversion, and brand affinity uplift.

  5. Add comprehensive filtering on product listing pages.
    Adding filters on the search or category page allows shoppers to quickly narrow down their options by selecting relevant attributes such as size, material, or colour without leaving the results page. By doing this, conversion rate increases by at least 5%.

  6. Make blog and other content searchable.
    Making blog content or other reading materials accessible through the search bar is a great way to provide customers with the information they are searching for more easily. Doing this could lead to a 36% increase in use of search, 8% increase in search-led revenue and 21% increase in search conversion rate.

If you’re interested in a detailed product discovery audit of your ecommerce store, apply for one here (it’s free).

You already have an amazing suite of product discovery tech, how is it going to get even better? What’s on your roadmap?

You may have heard, but Klevu has just become MACH certified - only one of three discovery platforms that are.  Klevu will continue to experiment within linguistics and machine learning to find more ways our technology can help retailers and shoppers alike.

We have an in-house AI lab which consists of research scientists that continue to experiment within linguistics and machine learning to find more ways to delight shoppers. We are exploring image processing, conversational commerce, natural language processing, and predictive analytics. We use this team to find the most cutting edge use cases and technology that we then apply to our existing products, or make new ones.

Klevu’s most recent and impactful innovations:

  • Headless SDK - Klevu's new Headless SDK makes it easier for developers and solution integrators to build custom storefronts and connect headless front end experiences with Klevu's open APIs. Written in Typescript, the Klevu SDK library also provides helpers to manage filters, cache requests, simplifying pagination, and more. This library has been built for all browser based frameworks and the Node.js environment, meaning it is agnostic and available to use with most of the JavaScript libraries in both traditional and headless implementations.

  • Visual Discovery (in beta right now - schedule a demo to see it) - Powered by Klevu’s intelligent and proprietary AI and image processing technology, Visual Search and Visually Similar product recommendations are available out-of-the-box and require a simple image sync. For Visual Search, shoppers can use real-world photos to search for similar products. For Visually Similar Product Recommendations, Klevu AI processes not only the image, but also the available product data.

  • Behavioural Personalization - With Klevu, the Personalization AI is customer-centric, driven by behaviour and not demographics. This allows for the shopper to drive the journey, without stereotypes or manual segmentation.

  • Klevu AI - Klevu is the most advanced product discovery technology available on the market today. The AI underpins the Klevu Discovery Platform providing semantic query processing,  product catalogue enrichment on import real-time user behaviour analytics based on shopper searches, product clicks, checkouts, product ratings, and the sentiment within product reviews. Recent enhancements to the AI include advanced linguistic support for most languages.

Customers and partners can see our roadmap and vote on items within the Klevu Merchant Centre, but I’ll list some of the main items here that are launching soon - Advanced filtering options for product recommendations, facet management in categories, analytics improvements and GA4 support, and native connectors with many tech partners including Klaviyo, Contentful, Shogun, Akeneo and more.<should there be a bit more text here?>

Finally, nothing wrong with finishing on a small bit of self-promotion. How does Klevu stand out when it comes to product discovery?

Klevu is the most advanced product discovery technology available on the market today. And, it’s easy to use whether you’re a brand on Shopify looking for a plug-and-play connector, or you want to build a fully customisable and composable eCommerce tech stack.

The AI underpins the Klevu Discovery Platform providing semantic query processing, product catalogue enrichment on import real-time user behaviour analytics based on shopper searches, product clicks, checkouts, product ratings, and the sentiment within product reviews.

Brands that use Klevu can control their merchandising as much or as little as they want to. Brands doing the best with Klevu have a balance of both automation and strategic merchandising. Klevu is the only MACH-certified product discovery solution that uses advanced natural language processing (NLP) and product ranking algorithms that allow retailers to generate 37% more revenue per web session and ensures a unified customer experience for shoppers.

Learn more about Klevu AI by watching this short video:

Klevu AI explained! Time to revolutionize your e-commerce store.

Or if you prefer to read the summary - here it is:

Uniquely, Klevu AI is trained on seven years of shopper intent, making it the most advanced and comprehensive eCommerce AI.

When the product feed is imported, Klevu AI reads and learns about the product along with all of the available product data. It learns the unique context of a merchant’s store. Klevu AI then automatically enriches the product catalogue by 3x - without the merchant team lifting a finger. And the whole process takes a few hundredths of a second.

Then, when a shopper makes a search on an eCommerce website using Klevu, Klevu AI uses advanced semantic query processing to understand and enrich the query and capture the shopper intent. Once the query has been processed, Klevu finds all relevant products, applies any business-level logic, and one-to-one personalization settings set by the merchant in the easy-to-use Klevu admin panel.

Klevu also takes into account what other people who like that item have gone on to click and purchase and then serves up the products most likely to be purchased by shoppers in search results, product recommendations and on category pages.

Klevu’s machine learning abilities develop an ongoing understanding of how  shoppers are engaging with a website specifically, improving with every click, purchase, and product review.


Learn more about the Klevu Product Discovery platform, and either sign up for a free account, or schedule a walk through demo here.

Rachel Tonner is an experienced marketer and researcher with 15 years of experience within eCommerce, retail, and UX, helping businesses identify and bridge experience gaps. Rachel’s blend of seemingly incongruous specialisms - including branding, copywriting, UX, PR, SaaS product and growth marketing, and even opera and vocal performance - give her unique perspective. Currently leading global marketing at Klevu, an AI-powered discovery suite.