Thursday, January 20, 2022
HomeDigital IconsInterview with Hamish Ogilvy from

Interview with Hamish Ogilvy from

Team eCommerce Next interviewed Hamish Ogilvy from to get more insights on how brands and retailers can ready their search functionality in 2022. Following is our interview with him:

Tell us about the current state of e-commerce search and why it’s important for brands and retailers to pay attention to it?

With disrupted global supply chains, people are working even harder to find what they want online before stepping into a store. E-tailers need to be everywhere their buyers are shopping — there is a shift towards multi-channel selling to reach consumers through channel partners, social media sites, or on their websites.

We’ve seen it with our customers as they scramble to re-platform to meet the challenge. They’re looking to plug-in site search to a variety of architectures and build a data-first approach to leverage data across channels for personalization, recommendations, and merchandising.

Search is really the tip of the spear for online sales. More than 42% of visitors immediately use an on-search box when they land on a retailer’s site and are 1.8x more likely to convert, so it’s an important place for retailers to focus.

What’s a common search misconception in the e-commerce search space?

That search still requires teams of engineers and data scientists to manage. Many companies invested in search solutions that required a lot of hand-holding. Or they spent two or three times the cost of the software on outsourcing management and upkeep.

Nowadays, SaaS-based search solutions have built-in data modeling to optimize results automatically. And, they offer user interfaces built for non-engineers to easily adjust search results for different campaigns. Gone are the days of writing thousands of rules to ensure good search results for just the top 20% of a product portfolio. Now, 100% of a retailer’s products can be optimized automatically without the huge investment in specialized teams of consultants.

How can the average retailer or e-commerce site manager strengthen their search engines today?

By embracing newer AI models. The search industry has been dominated by providers built on 20+-year-old search technologies. These keyword-based technologies simply don’t work for today’s sites. A simple example: if a customer searches for a USB C, USBC, or USB-C adaptor, keyword search engines often fail. It’s a frustrating problem. To circumvent the problem, companies have written hundreds of rules, created synonyms, or stuffed their sites with keywords to account for every possible combination of key terms.

This can be solved with newer AI solutions that don’t just look at keywords but also understand context and user intent. A problem that has dogged sellers for more than a decade can be solved in minutes, freeing up their time to focus on other business-critical areas.

How can advancements in AI help e-commerce professionals create a better search experience?

There are two ways AI can be leveraged by sellers: better relevance and better ranking.

To illustrate the idea of relevance: If a customer searches your site for a “kettle bell” or “kettlebell” or “kettelbel” (yes, it’s misspelled!) they should get exercise equipment, not a teapot!

Relevance is how close the search results are to the query. Many sites still fail to offer relevant results, likely because they are still based on old keyword search engines which require extensive training to handle every edge case. Newer neural search technologies understand context and intent to drive better results regardless of which keywords a customer is using.

Ranking is how search results are ordered. Technologies such as reinforcement learning will automatically optimize rankings based on buyer signals – clicks, signups, conversions, ratings, etc. It takes the guesswork out of ranking and ensures results are converting.

Is there a way for the average e-commerce brand to spot the difference between true AI platforms and ones that are rules or keyword-based?

This is a tough question. Frankly, it would be very difficult for most retailers to know whether something is true AI or fake AI, but there are a few general principles.

Rules, synonyms, and keywords: if your search engine still relies on you manually typing in relevance rules, common synonyms, or lots of keywords, it’s probably fake AI. Real AI engines don’t rely on these things anymore.

Common typos fail: Vector-based search handles common typos out of the box, so if your search engine is still confused that “vaccum” means “vacuum”,  it’s likely fake AI.

Longtail searches are poor: Most e-commerce companies optimize search results manually for the very common queries (head queries) on their site. Optimizing for the bottom 80% (long-tail) was too time-intensive. Real AI search works on all 100% of your products and search queries. If you’re still not getting great results on the bottom 80%, it’s probably fake AI search.

A vendor calls it “explainable AI”: AI models are inherently extremely complex and often include millions or even billions of data points during inference. They aren’t explainable! The vendor should still be able to explain what models are in use; if the answer isn’t clear, it’s probably not real AI. just launched a new product, Neuralsearch. How is this platform different from what’s already available on the market?

Many companies claim to offer AI search, but they are either simply tacking AI capabilities on top of an existing search solution (which slows results and can be difficult to train) or not delivering AI.

Neuralsearch is the only true hybrid search service (combines both keyword and AI search simultaneously in the same index) that we are aware of that delivers single-digit millisecond query times no matter the scale or query throughput. Additionally, Neuralsearch has a unique one-click improvement UI that allows anyone to teach relevance as needed.

What can the e-commerce industry look forward to in 2022 and beyond as it relates to site search?

Consumers are getting more and more sophisticated in their buying habits and demands for great shopping experiences. To us, it begins with search. AI is now matured to the point where it can help retailers provide better multi-channel experiences. Using the data and new machine learning models, things like personalization, merchandising (searchandising!), and better recommendations can all be delivered through a better on-search experience — whether it’s mobile or desktop.

About Hamish Ogilvy, CEO and co-founder

Hamish is CEO and co-founder at, a company that pioneered the use of machine learning in site search to provide unparalleled results. Prior to, Hamish crossed over into data science from his original career as a laser engineer. He has managed the data analytics and intelligence for many large corporations before moving into product management and machine learning. He holds a PhD in laser physics from Macquarie University.

About offers a search and discovery service with powerful AI engines, including Neuralsearch™, the world’s first instant AI search technology, and reinforcement learning for self-optimizing results. Businesses of all sizes use to build site search and discovery solutions that maximize e-commerce revenue, optimize on-site customer experience, and scale their online presence. currently serves billions of API requests every month on more than a thousand sites globally including Catch, Lockheed Martin,, Rentpath, Sennheiser, and Unity. For more information, visit


Most Popular