The Most Important Tool to Deliver a Quality Ecommerce Experience: Search On Your Store — Arv Natarajan | Why Search Quality is Key for Ecommerce, How Product Data Affects Search, What Businesses Gain from Advanced Search (#339)

In this podcast episode, we explore one of the most crucial tools for a successful eCommerce experience: your store’s search function.

Joining us is Arv Natarajan, Director of Product at GroupBy Inc. Arv shares his insights on improving search capabilities to enhance customer satisfaction and conversion rates.

We discuss data catalog quality, the role of AI in search engines, and how to optimize merchandising with analytics. Tune in to learn how to boost your online store's performance with smarter search tools!

Topics discussed in this episode:  

• Why improving search quality is crucial for ecommerce success
• How product data quality impacts search performance
• What advanced search capabilities like natural language processing can offer
• Why analytics and understanding user search behavior is important
• How to balance AI-powered search with human curation and merchandising
• What types of ecommerce businesses benefit most from advanced search
• How the search solution integrates with custom ecommerce themes
• What the future of AI-powered ecommerce search and product discovery might look like

Links & Resources

Website: https://www.groupbyinc.com/integrations/shopify
Shopify App Store: https://apps.shopify.com/groupby-ai-search-discovery
LinkedIn: https://www.linkedin.com/in/arvnatarajan/
X/Twitter: https://x.com/groupbyinc

Get access to more free resources by visiting the show notes at
https://t.ly/GECxQ

SPONSORS:

Is your tracking setup really ready for the holiday season? Most brands rely on out-of-the-box tracking that miss out on valuable data, limiting your ad channels’ performance. With Elevar advanced tracking, eCommerce brands capture 99% of conversion events, even after cookies expire. Brands like Obvi have boosted abandoned cart revenue by 34% without lifting a finger—just smarter data! https://www.getelevar.com/BFCM/ 

Ask The Ecommerce Coffee Break → Submit your question here: https://ecommercecoffeebreak.com/ask/


Sign up for our free newsletter

Become a smarter online seller in just 10 minutes per week. 100% free. Delivered every Thursday to your inbox. No Spam. Unsubscribe anytime. Sign up at https://newsletter.ecommercecoffeebreak.com


Rate, Review & Follow on Apple Podcasts

Enjoying this episode? Help others like you by rating and reviewing my show on Apple Podcasts! Your feedback supports more people in achieving their online business dreams. Click below, give five stars, and share your favorite part in a review!

Click here: https://podcasts.apple.com/us/podcast/ecommerce-coffee-break-digital-marketing-podcast-for/id1567749422

And if you haven’t yet, follow the podcast to catch all the bonus episodes I’m adding. Don’t miss out—hit that follow button now!

Transcript

Welcome to the eCommerce Coffee Break Podcast. In today's episode, we discuss the most important tool in your tech stack to establish and deliver a quality eCommerce experience. Joining me on the show is Arv Natarajan, Director of Product at groupbyinc.com. So let's dive right into it. 

This is the eCommerce Coffee Break.

A top rated Shopify growth podcast dedicated to Shopify merchants and business owners looking to grow their online stores. Learn how to survive in the fast changing eCommerce world. With your host Claus Lauter and get marketing advice you can't find on Google. Welcome to the show. 

Hello and welcome to another episode of the eCommerce Coffee Break podcast.

Today we want to talk about the most important tool to deliver a quality eCommerce experience and that is the search on your store. If people can't find your products, they'll probably leave very, very quickly because as you know, Google. The attention span is very, very short. So you want to make sure that people will find your products.

And that's exactly the topic we're going to dive into today. Joining me on the show is Arv Natarajan. He is the director of product at GroupBuy. He is a passionate entrepreneur currently responsible for product management. Arv has over eight years of experience in the oil and gas industry, including five years at Walter Lee Parsons.

So let's welcome him to the show. Hi Arv, how are you today? Hi, doing well, Claus. Glad to be here. One of the experience and one of the reasons for frustrations is poor search on e commerce sites. And give me a bit of a background. What makes a good search a good search? I 

mean, I think this one's really easy and you kind of mentioned it when you were giving your introduction.

If a customer can find the products they're looking for, you've done your job as a retailer. You've served the customer and you've shown them what they're looking for. They're able to make a purchase on your website. That is really, to me, the definition of a good search. And the obvious is also the same where if they cannot find what they're looking for, then that's a bad search.

Now there's many reasons why people might not be able to find what they're looking for. And we can get into that, but search engines and the quality of your search engine is a big part of it. Now 

at Groupby, you are helping with improving the search on sites. What do you see? What are the biggest challenges or mistakes people do, merchants do when it comes to implementing search on their store?

There's a few different things. So the first I would say is the quality of your data catalog. As a retailer, if you have a bad quality data catalog, or your product catalog is is not as complete as it could be, there's missing information, there's incorrect information, then obviously no matter how good your search engine is, it will still not be able to find the right products.

For that person at the right time, so you have to start with your, with your product catalog. How is the quality of that data? And, and you may work with manufacturers or third party marketplace sellers who sell you their products, who sell their products on your website or on your Shopify site. Then it's important to make sure you work with those partners, with those vendors to make sure that the quality of your product catalog is, is good.

That's your starting point. So you definitely want to start with data. Once you've looked at that and your data governance strategy is solid, then you can look at your search engine. So what kind of search engine are you looking at, or do you use on your website? I mean, technology that has advanced so much in search over even the last year or two that, I mean, advances in gen AI as everybody has heard is, has improved a lot of things, one of which is the quality of search engines.

So legacy search engines, We'll look at very basic ways of identifying what the person is looking for. There's limited capability to understand the intent. There's less, uh, or limited natural language processing or semantic understanding. You're basically looking at the key words that are in the query and trying to find relevant parts.

But it doesn't take into account personalization. It doesn't take into account what products are likely to be purchased for a given query. If someone searches for jeans, but they have an affinity for a brand or a size or a color, it's important to take that into account in a search engine. So, the latest generation of search engines do all of that.

They use large language models. They use AI to the best of its capabilities to return the right products at the right time. 

So he said you need to have a good data set. You need to have all the information is probably metadata. I can highlight a little bit for me. It's like, what kind of data do I need to maintain in a background?

And then you have the AI supported search engine feature, which helps you on search. Now how does that work together? Um, obviously you cannot index every keyword that people are thinking of or typo they can make off. I'll give you an example on how that would work in real life. 

Yeah, for sure. The example I actually like to use a lot with our, uh, especially our apparel customers, people that sell clothing type products is date night dress.

So that's a very interesting query because it means something to the user that's looking for it, but it may not mean something to a very basic engine. So date night dress, obviously we all understand what that means. It's, it's some kind of evening wear that, that someone would wear on a date night. The intent of what the user is looking for is a very specific kind of dress.

You're not looking for maybe summer dresses. You're not looking for wedding dresses. You've got a very specific type of dress in mind. And the engine has to understand that in legacy search engines, if you search for the words, date, night, dress. You might hit on dresses that have date patterns on them.

Like they may have patterns of date arms or all the types of things, because it picks up the word date and then it picks up the word dress. And then it tries to find that, or you might find night dresses, which you're also not going to be wearing on a date because that's obviously pajamas. So it's very important that the engine understands.

First of all, what the user is looking for with that semantic understanding and natural language processing and then applies that to your catalog. And then your catalog also has to be tagged properly. If you don't have occasion, for example, as a attribute for your product catalog, it may not understand very easily that there are these types of addresses that belong to a certain occasion.

So that's, that's a very easy way of showing, well, you have to have the right data that's tagged properly. And then your engine will also be able to work with that data to return the right results at the right time. That is applicable in many industries. That was obviously an apparel example, but you could apply that to hardware or, or any other types of electrical products, any, anything else as well.

It's equally important that the engine understands what the user is looking for. 

Is there a balance between AI and manually controlled merchandising? So how it shows it to the visitor? 

Yeah, that's a great question. Um, because in, in, in what we've seen in the recent Yes, is that the role of the merchandiser has changed now that the A.

I. Technology has advanced to this point. It's able to do a lot of the manual or busy work that merchandise is used to do before. So if you're familiar with legacy search engines, there's typically a lot of curation that needs to be done. And if you currently have to create a lot of rules, business rules to say, you know, In this case, do X, Y, Z, and show them these results.

That is very manual work that an ANI can automate for you. So it's not replacing the merchandising goal. It's changing it in a way that allows merchandisers and retailers to focus on more strategic things. Growing their catalogue, growing their customer base, reaching a new market, things like that.

Whereas these, these Rules that needed to be curated or pages that need to be curated no longer need to need to be done because AI can handle that for us. And now that's not to say that we give up all of that control either. There is obviously a human in the loop step where there's a review process that needs to be done.

So once we can use the AI to generate this kind of results, we can always go back and review it before it hits our customers so that we're not. We were just double checking the work of the name, basically. But we're thinking of it as a way of really automating and making more efficient the day to day work of a merchandiser so that they can then be free to focus on the more strategic things.

I wanted to have a little bit of a topic of analytics of searches. I think that's a very interesting topic for a lot of merchants because you can learn a lot about your customer coming to your website by the searches they do on your store. Give me a couple of examples. What kind of analytics can I look into?

Yeah, analytics, to be honest, is something that every retailer should be looking at every day. It's so important because it gives us a lot of insight. For example, you might be able to see that there's a trending search on your website. You might be able to see that people are searching for, let's say, other kiddie backpacks.

And that is there because there's a social media post that went viral and now a lot of people are looking for that type kind of backpack. If you either A, don't carry those kinds of backpacks or you're running out of inventory, Or it's not a search that is meant for your website. You can help them create the next action for your customers to either find the products they're looking for or guide them to other experiences that you would like to guide them to.

So understanding what is trending, what is important for your customers is key. And that's already done through analytics. The, on the other side, you might be able to also help improve some of the, um, the logistics or pricing information that, or a data that you have on your website. If you can. through analytics, understand that these products are running out of inventory or stock.

And then you might be able to then make sure that those are ordered in time for your next week sale or something like that. Or if you have a campaign that's running next week, analytics engines are able to tell you, these are the products that you should put in your, in your email campaign next week, because this is what your customers are looking for.

trending. So there's a lot of insights that we can get from analytic. And it's so critical that. Retailers consider these analytics every day. 

From the search queries, you might have search queries where you don't have the product for, is that something that helps you to optimize your store? 

Yeah. I mean, so the concept of null search rate is a key metric that retailers look at.

So null search rate is the amount of times that a user. Ends up seeing a null search results page, so no products are shown on that on the page. That is obviously something we want to avoid because if you don't see any products, it's a very easy way for consumers to just bounce from that website and go somewhere else.

So that's, that's definitely something that retailers want to avoid. One thing that this latest generation of search engines can do is Understand what the user is looking for and show related products to that particular query. So maybe if you're looking for a particular brand of coffee maker, it's a KitchenAid coffee maker, instead of showing no results because you don't carry KitchenAid, you might want to show a different brand of coffee makers that are also relevant to the user and they can then end up making a purchase of one of those things.

So, so being able to understand the intent of what the user is looking for. And expand the result set to be showing not just very specific results to that query, but also somewhat related results will help improve the number of conversion opportunities you as a retailer will have. And so that something like that is, is also key for retailers to explore how this search engine works and how they can improve it.

Hey, Claus here. Just a quick one. If you like the content of this episode, sign up for our free newsletter and become a smarter Shopify merchant in just seven minutes per week. We create content from more than 50 sources, saving you hours of research and helping you stay on top of your e commerce game with the latest news insights and trends.

Every Thursday in your inbox, 100 percent free, join now at newsletter. ecommercecoffeebreak. com. That is newsletter. ecommercecoffeebreak. com. And now back to the show. It's sort of proven that people who use a search feature on a store convert much higher than somebody who's just browsing around. My experience is that people get the term search and filter wrong all the time.

So they mix it in one head and whatsoever. Can you? Explain to me what the difference is between a filter and a search. 

Yeah, absolutely. And, and you're right, it can be confusing at times, especially if this is not a realm that you've spent a lot of time in. So a search basically means that there is a query associated to that particular part of the shopping journey.

So someone is searching for the word gene or the searching for the word bat. That is a search. If you are browsing or if you are applying filters to a result set, you're trying to narrow down your result set. To something that's more specific to what you're looking for as a consumer. So for, say, for example, you searched for jeans.

So you can then apply a filter of a brand, let's say KuberBoss, and now you will only see KuberBoss genes and the results set because you're narrowing the results set from a thousand genes down to maybe 50 genes. So it lets you narrow the results set down to something more specific and relevant to what you're looking for.

And Browse is the other side, we kind of alluded to it when you brought this up, is, you know, A browse or navigating through a series of browse pages or browsing to a certain landing page means that there is no query associated with it. So you're not searching for the word genes, you're navigating to the landing page for genes.

So there's no query term associated. And so your search engine may deal with that differently. You may, you may have different logic for searchers where people are looking for something a bit more. So you might want to search for crafts for my 10 year old niece or something like that. And that's now a very specific term that you're searching for, and you may not have a landing page built for that.

You might have a landing page for jeans, but not for those crafts. So there's a reason why some customers or consumers would search for something versus browse for something. It really depends on what the user is looking for. 

Now, you have a Shopify app grouped by AI search and filter. What kind of results do your clients see after implementing?

You don't need to tell me a brand or something, just give me an example of what kind of progress they had. 

Yeah, for sure. I mean, what we've seen, and the reason why we built this application is one, the default Shopify search is not for everybody. I mean, if you don't sell that many products online, if you only have a catalog of, let's So if you have a couple of hundred products, maybe you can live with the default Shopify search.

However, there are a lot of improvements that can be made. It's obviously not using the latest in technology because it's also free. And so they haven't invested as much time into that as some of the other applications that are out there. And so that's why we brought to market the move by search engine, which is powered by Google technology, actually.

So using Vertex AI, we're able to now bring back revenue optimized results. And so this is specifically targeting kind of the SMBs or the mid market type companies that have thousands or tens of thousands of products. That's really where it starts to show value. And the idea is to maximize your revenue.

So the results that are shown to your customers are shown in revenue maximizing order. So for a given context. Let's say it's, again, a search for jeans or it's a browse to the landing page. We will show the results that are most likely to be purchased by that individual on that particular context. So it's personalized to the individual automatically, and it looks at the aggregate of revenue.

For that website, for that search term and returning to those results. So it's maximizing revenue for retailers on that side. And on the other side, it's helping customers find the products they're looking for. 

Now, I think that's a very nice implementation of, um, one of the conversion rate optimization principles is bring your best sellers in the front because they have proven that they sell.

And, um, if the AI helps with that, there is basically a guarantee that it will improve your revenue. Who's you already mentioned a little bit, but who's your perfect customer? 

It's quite a large range with a group. I start off by looking at enterprise. So this is the big, big box stores or, or retailers that are in a niche, uh, kind of vertical, but they are quite large enterprises with the release of the, uh, the Shopify app.

We're now looking at also supporting mid market and SMB type retailers, but really it starts at. Adding value at around a few hundred to a thousand products. That's kind of the small B the smallest range for something like this. This is a, a premium product. And it's like, um, using a Ferrari to drive down to the corner store to buy a bottle of milk.

Like you don't need that kind of product to add value in the end, because you'll be paying for it and it may not be worthwhile for your shop, but if you have, In the ranges of thousands to tens of thousands of products and above, then it really starts to show that because now you're doing a lot of business online, your customers may struggle to find the products they're looking for.

And an engine like this will help them with that. 

Walk me through the typical onboarding process for a new client. What steps are involved and how long does it usually take to get up and running? 

Yeah, it's actually super quick and easy to use it in our opinion. So all you have to do is go to the Shopify store and find the app, the group by app.

Once you click install, you'll go through a series of steps to upload your catalog. So now depending on how big your catalog is, it might take anywhere from a couple of hours to maybe five or six hours. And that is just uploading and indexing your catalog for the first time. Every subsequent upload will obviously be a lot faster, but the first time it'll probably take a couple of hours to get uploaded.

And that's in the range of tens of thousands of products. If you have less, it'll obviously take less time. Once your catalog is uploaded, that's it. You just have to click deploy and your search engine will be updated with group by search engine. Now, day one, we will, you will have all of the insights from google.

com and Google shopping and YouTube to understand what people are looking for, but you won't have any of the insights of your. Users behaving on your website because those are the two key inputs into the AI engine one is the actual catalog itself Um, but then it's all of this customer data that, that is feeding the AI and training the data, the AI and how users behave on your specific site and in general.

So in general, that data is already inside the engine because those interactions, there are trillions of interactions that happen every day. And we are one of maybe one or two apps that are actually on the Shopify store that have this technology with Google to power search engines. Um, so feeding in that aggregate data of here's what people are searching for on Google.

com and YouTube and so on. adds a lot of value to the A. I. And then on top of that, we like to layer in actual data from your customers on your website. So how do people interact on your website? What are they searching for? What are they clicking on? What are they buying? All of that information is also fed into the A.

I. And then we provide this. personalized search experience for your customers. 

Now, with all this data that's coming in, you said it takes a little bit of time, but is there any kind of homework? I mean, I think you mentioned it in the beginning that I need to do before I really can get started. 

I mean, it's important to think about your, your specific use cases.

Nine times out of 10, the AI will handle it for you automatically. If you just want to show the regular search results, there's nothing that you need to do on top of that. It will return revenue maximized results automatically. But say, for example, you want to build a landing page. You've got Halloween coming up in a couple of months.

That's a landing page that I would like to build for my Halloween focused products. Well, that is something you can do in our merchandising platform and then override the AI to say, I don't want you to show just default Halloween products. I want to show these specific ones because I sell pumpkin scented candles to Halloween.

So that's going to be a page that I want to build. And so then you can, you have these specific business objectives that you might want to, you know, To show on your website and that is, that is something that you as a retailer or a merchandiser can do on top of what the AI does automatically. 

You mentioned before that you're now with the Shopify app also targeting smaller medium enterprises for your app.

What's the pricing structure like? 

So it's a really a pay as you go type structure. So you'll be looking at. The number of queries that occur on your website. So the number of searches that people perform on your website. So it's dependent on traffic. If you go on a, on a, on a app site, you'll be able to see the actual exact numbers for the, for the pricing.

But let's say it's in the realm of around 3 or so for a, for a thousand queries. So for every thousand, Whereas you'll be, you'll be paying that amount of license for that particular amount of traffic. 

One question that often comes up when it comes to, um, to apps, Shopify apps, how does it work with my theme?

Does it work with every theme or is there, are there ways that I can basically customize the design and implement it? 

Yeah, that's a good question because We have, uh, this is a new market for us as of a year ago. And so now we've, we've run into some interesting use cases where, where a lot of merchants have very customized themes.

So we've got it, got our app working with the default theme. You'll be able to use it. If you use the default theme, no issues at all straight in and you can start using it. If you have a custom theme, it really depends on the level of customization. If you've modified your front end to look very specific to your branding and so on.

The theme should work by default, but if there are issues, we have a dedicated support team that will work with you on any of those particular issues you might or challenges you might run into to help you understand what the issue is and how to fix it. But in nine cases out of 10, it should absolutely work out of the box.

Okay, plug and play. Look in your head. Can you give me an idea on how AI is shaping the future of e commerce search and product discovery? Looking into the glass bowl. 

That's such a tough question to answer with confidence because the technology is changing so quickly. Where I see this going in the next few years is really Helping customers find what they're looking for, but in different ways to regular keyword search.

So everybody's very familiar with going onto Amazon or any shopping website, retail website and typing in the search that they're looking for, the products they're looking for. But that's a very simple type search as you're looking at. It's a very specific product type. I think where the future of e commerce especially with AI is going is through other channels like social media and other ways of searching similar to what you might experience in a store where you're talking to an associate.

I have to do a bathroom renovation next weekend. Help me find the products I need. Like that is a very nuanced question to ask a surgeon and it's not really going to be able to help you. In its current iteration of technology, so I think in the future, being able to handle those kinds of searches and conversations will be important for retailers to make sure to stay on top of.

And then also with social media to drive traffic from these different channels, if someone's looking on Instagram or TikTok and they see a celebrity wearing a jacket and they'd like to buy that jacket, there has to be an easy way for them to be able to get from that social media post to your website.

So being able to handle someone clicking on a picture, taking them to relevant websites that have that same jacket will, I think, be the future of where we go in retail shopping, really making it easier for consumers to find the products they're looking for. 

Yeah. And I think that's key. Having a way to find things easy has direct impact on your revenue.

Before our coffee break comes to an end today, is there anything that you want to share with our listeners that we haven't covered yet? 

Yeah, I mean, I think it's important to evaluate where your search experience is currently. I mean, this is maybe something that not everyone pays a lot of attention to, but it is a critical factor in driving revenue for your website.

So if you're a Shopify merchant and you don't really know how your search is performing, you haven't got the analytics to support any of that with data, that is important. You need to understand how your search is performing, what your users are doing on your website. Start with that. Then you also need to look at your, your, the quality of your product catalog, do some, some analysis to understand, is it complete?

Is it incorrect? Is there any bad data in there? Is that adding my search experience? That is another thing that retailers should be considering. And then lastly, the search engine itself, if you're using default out of the box kind of plug and play. Type searches that are not kind of optimized for revenue or other key metrics.

That's something to consider as well. It's not cost prohibitive to jump across to a new search engine. It should be very easy to also migrate to something like this. You don't want to do a lot of work for it. So that's, that's where apps like ours will, will help you do something like that. 

I absolutely would recommend to have a AI search in there.

Searches convert much higher and you want to keep your customer happy. And search is a crucial factor for that. Can people find out more about you guys? 

So you can go to groupbuying. com, which is our main website. Or if you look for Group Buying on the Shopify app store, you'll be able to find our app there as well.

Okay. I will put the links in the show notes as always, then you just want to take away. Thanks so much for giving us an overview of what the status is right now in searching and filtering for your Shopify store. I hope a lot of people will click on the links and check it out. I think it's a must have for every store.

Yeah. I hope that AI will help us further with getting better searches. Thanks so much for your time today. Thank you for having me. Hey, Claus here. Thank you for joining me on another episode of the e commerce coffee break podcast. Before you go, I'd like to ask two things from you. First, please help me with the algorithm so I can bring more impactful guests on the show.

It will make it also easier for others to discover the podcast. Simply like, comment, and subscribe in the app you're using to listen to the podcast, and even better, if you could leave a rating. And finally, sign up for our free newsletter and become a smarter Shopify merchant in just seven minutes per week.

We create content from more than 50 sources, saving you hours of research and helping you stay on top of your e commerce game with the latest news, insights, and trends. Every Thursday in your inbox, 100 percent free. Join now at newsletter. ecommercecoffeebreak. com. That is newsletter. ecommercecoffeebreak.

com. Thanks again, and I'll catch you in the next episode. Have a good one.