How to Improve Shopify SEO and Appear on Shopify x ChatGPT (Interview with Tom Gandhi)

(0:00) Let's go. Awesome. Okay.
Hi, everyone. I'm Joy, the marketing lead here at Fera. (0:05) I'm really excited to kick off today's interview and welcome Tom Gandhi, an SEO whiz who's worked (0:12) with some truly incredible brands.
One of my personal favorites being Pangaia, along with (0:17) Wild Nutrition and many others. Tom has a rare ability to turn search into a real growth engine, (0:23) and I'm so excited to deep into his insights, experiences, and practical strategies with you (0:30) all today. Tom, I know I gave kind of a bit of an introduction, but did you still want to introduce (0:35) yourself? Yes, I'm Tom.
I've worked in SEO my whole career, pretty much fell into it as you often do (0:43) with SEO. Worked at agency most of my career and about four, five months ago, went full-time (0:51) consulting. Working in a plethora of different businesses, not just e-commerce anymore, but also (0:58) in like across border solutions, some SaaS businesses.
But my bread and butter is SEO. (1:04) I'm staying very close to the ground with the changes to search and still working with some (1:09) really great brands, evolving their search strategies as well. So super excited to be (1:13) part of this conversation.
Awesome. Well, we can kind of get started right away in the questions. (1:19) I know this is probably something that you will review when you're working with new (1:26) clients, but I'd love to know what are some of the most common technical SEO issues you (1:32) encounter on e-commerce marketplaces, whether it's Shopify, Wix, and how do you typically resolve (1:38) them or provide a solution to them? Yeah, so Shopify, I guess I worked through the platforms (1:45) that I've worked with most.
Shopify generally doesn't have too many technical issues, just purely (1:50) because you get what you get and it's a very strong solution for a lot of merchants. (1:56) There's a lot of technical optimizations you can make. That's not to say that without them you (2:01) won't be able to perform, but Shopify is probably one of the better platforms for (2:06) general ongoing technical problems.
Once you fix them or improve the site, you generally don't (2:11) need to keep revisiting it unless it's like a headless build, for example. I haven't worked (2:18) too much with Wix, only a couple of times in my career. It seems to be a very good entry-level platform.
(2:24) Again, you can improve it as much as you can improve it, but I have worked with Magento (2:29) probably second most following Shopify and that's, I guess, the sort of platform where (2:34) there are a lot of improvements to make technically from what you get out of the box. (2:40) Again, it's not to say that you can't perform without those improvements, but (2:45) you run the risk more frequently, I would say, with Magento to actually have technical issues and (2:50) that could be from like a filter indexation, callability, how your products are configured, (2:56) and then also in the background of that is that you have these catalog URLs, which (3:01) is essentially how you'd stipulate your categories. Then you have a front-facing (3:07) version, which is the optimized version, and so brands sometimes don't completely remove (3:13) that back-end version and it becomes a bit duplicated.
So that's just a quick example of it, (3:18) but I would say Shopify is very good generally. Magento needs some tweaks and some refinement, (3:25) and then I guess like the Wix and Squarespace of the world are at a slightly more entry-level, (3:33) but as I said, I don't have a huge amount of experience with them. (3:36) Yeah, and I know you mentioned that at least for SEO on Shopify, it's probably set out extremely (3:43) well, but do you notice that there are maybe any SEO opportunities that sometimes Shopify stores (3:50) overlook or maybe are missing? Yeah, so technically they're built really well, but from an optimization (3:58) point of view, there's always that there are technical optimizations and on-site optimizations (4:03) from an SEO standpoint are where a lot of brands that have just entered into say the Shopify store (4:07) that haven't worked with anyone on their SEO and on their configuration before, (4:12) they'll often not be well optimized.
So sometimes brands don't like to have, (4:16) except for example on their category pages, collection pages obviously on Shopify, (4:20) they might not want to show customers what the collection title is from like a styling point (4:27) of view, but often if you like neglect that, then you don't have an h1 for the collection pages, (4:32) which can work against you in ranking the collection page for those key terms, (4:37) like which are very commercially focused. Breadcrumbs, so Shopify has the long product URL, (4:44) the short product URL, it always defaults to the long product URL because that also (4:48) allows you to build out the breadcrumb, it follows the path of the customer essentially, (4:53) but that means that if a product sits in multiple categories, then it will have multiple different (4:57) breadcrumb paths. So we'd always advocate to go to the short version, but before you do, (5:03) you want to manually and statically build out your breadcrumbs for every product, (5:08) so that a product will only ever have one breadcrumb trail, which is a really good advantageous (5:12) way to link more to your key category pages and to emphasize them more, which normally transpires to (5:19) improvements in performance.
And then it's just like really the blog on Shopify is a really (5:25) challenging one, like it's not great, it's not so commercial, so what a lot of brands end up (5:30) doing is they start to use the page template rather than the blog template, because it becomes more (5:35) of a landing page, way more customization opportunities, you can actually add a product (5:39) carousel, a category carousel, and it becomes way more commercially focused and transactional (5:45) than the blog can ever be. So that's one of the biggest limitations I would say currently with (5:49) the Shopify blog, I'm not against it, it's good for just if you want long form content, (5:53) you want to sell products at the same time, then it doesn't do the best job of that. (5:58) I mean, there's probably more things I could go into, but I'd say they're the ones that I would (6:04) say we most commonly fix immediately, just because once you've got them nailed right, (6:11) it makes a difference.
Yeah, and just curious, I have a feeling that a lot of merchants, (6:18) one of their common questions or the issues that they might be facing is, (6:23) hey, Tom, I'm posting all of this blog content, I'm doing what I'm supposed to do, (6:28) but none of my content or none of my information is showing up on, for example, Google search (6:33) results, right? I feel like that's always a common issue that a lot of merchants like to talk about. (6:39) Is there a kind of a rough timeframe or timeline that you kind of give to merchants? I know a lot (6:46) of factors are involved in it, but when a seller does kind of come to you with this issue, (6:53) how do you typically respond to it? Yeah, so I think it really does, (6:59) and it's an SEO answer, but it does depend, because ultimately, if your site is really big, (7:05) it gets called really frequently, it's often that you make a change and it gets picked up (7:10) super quickly, like within a couple of days. If you're a smaller store, you get visited less, (7:14) you generally have less authority, because you're less well known, it takes a lot longer.
And (7:18) obviously, also, at the same time, if you have a really unclean page profile, and from that, (7:27) I mean, you have a huge number of non-indexable pages, and a really small number of indexable (7:31) pages, Google and other bots are spending so much time calling those really low value pages, (7:36) because you've either not known to index them, you've got a parameter that is just infinitely (7:41) causing a loop of discovery, or you're not using the robots.txt well enough, it means that you're (7:49) not prioritizing those key pages, and those key pages might be the blog, for example, as you've (7:53) mentioned there, and that's why that's then impacting the discovery of it, because every (7:57) site technically has like a half-life of time that Google will spend crawling it, and the bigger the (8:02) site, the more time it will spend. So if it's spending all its time calling really low value (8:06) pages that you don't want to be crawled, then you're going to impact your ability to actually (8:10) rank those key pages. So yeah, the big one is site cleanliness and healthiness.
I do say, (8:18) obviously, earlier on that Shopify is pretty good, but the more apps you add, the more complex it (8:23) becomes, the more parameters start to be crawled and indexed, or crawled and non-indexed. So it (8:29) really is a case of like continually just reviewing the page report in Search Console, seeing if (8:34) they've currently like spiked massively, you need to understand why and what you need to do to stop (8:38) that happening, so that then you can prioritize those key pages, and then you can actually make (8:42) sure that the pages you want people to actually interact with are being interacted with. (8:47) Yeah, absolutely.
I always feel like, especially with SEO, it's a marathon, not a race, right? (8:53) It is, but honestly, sometimes it's not, and sometimes you can make a change, and it's like (8:58) super instantaneous, and you wish it was always that way, because obviously SEO isn't really (9:03) always that way, not like the paid team has it, like, you know, turn on the money tap, (9:09) and you start to make money. Yeah, but it's nice when that does happen. As I say, the bigger the (9:14) store, generally, the quicker the implementation is picked up, and the quicker you start to see results.
(9:20) Yeah, that makes sense. Absolutely. Now, I know this is kind of an interesting, exciting topic, (9:26) but back in September 2025, Shopify and OpenAI made an announcement on how Shopify merchants (9:34) are going to be able to sell directly through chat GPT conversations.
I'm curious, how were (9:40) your initial thoughts when you heard this? I mean, I've not seen, you know, I think it's a very (9:46) exciting announcement, and it certainly was at the time, but I think it's kind of been shrouded now by (9:51) the announcement from with Google and Shopify two days ago, which seems way more exciting, (9:58) seems way more current, seems like they're going to be able to implement it much more quickly, (10:02) and you're already seeing MCP code, sorry, UCP code on the Shopify sites. So that shows that (10:09) they're already pushing forward with this relationship. There's already a product (10:12) preview feature in Shopify to show you what the agentic protocol will be displaying, (10:17) and if it's displaying the right information.
So OpenAI had a really good early mover advantage (10:23) there, but they've just not capitalized on it, and they still haven't got the shopping feed (10:26) into chat GPT live. So obviously the product, when you actually try and transact and shop in (10:32) chat GPT, the outcome isn't very good. It's not very in line with what you've searched.
(10:36) The product selection might not be great. It links to out of stock products, which obviously (10:40) isn't very helpful. So I think there's a few key things that have slowed them down, (10:48) but obviously Shopify have gone ahead and also now are in relationship with Google.
So that (10:53) could really, you know, accelerate the agentic commerce future much more quickly, and we might (10:59) start, you know, we might all start using AI mode more frequently to shop, and same with Gemini. (11:04) And it's a very exciting time, obviously, for search. We haven't had such a big shake-up in (11:10) ever, well I guess since you could start shopping online versus just retail, brick and mortar.
(11:18) I have no idea how long adoption will take for agentic. People have to trust the bots, (11:23) and right now I don't think people would trust the bots to make a shopping decision for them. (11:26) Yeah.
At least I know even with, at least on Amazon, when I'm shopping now, (11:33) they're really pushing you to use Amazon Rufus. I'm not too sure if you had the chance to use it, (11:38) but it's also their AI shopping assistant. If you log into your Amazon account, it's probably (11:43) right there, and you can actually type in like, you know, best hiking shoes that are waterproof, (11:50) that are popular amongst women with 30 to 40-year-olds, and they actually give really (11:54) accurate results, right? So I know ChatGPT has been doing quite a bit as well.
Whenever I'm (12:01) shopping, they'll actually give me product recommendations, but probably not as advanced (12:06) as to what Google is already doing right now. And I guess, you know, it's probably still in (12:14) beta mode at OpenAI, but I'm curious, is there anything a Shopify store or seller can do in (12:22) preparation for this, whether it's with ChatGPT or Google? Essentially, how can they make sure (12:28) that they show up on these AI search results on ChatGPT and Google? Yeah, I think the biggest one, (12:36) fundamentally, and it's going to be the same with the genetic shopping, is product data, (12:42) depth of product data, how far do you go down that road? And by that, I mean, you know, the brands (12:48) often just tick the basic box in product data, like where's the product come from, what's the fit (12:53) type, but do they go and say what activity is this particular product appropriate for, how do (12:59) you wear it, what do you style it with, what's it made out of, what level of protection has it got, (13:06) it's waterproof, like how waterproof is it, is it suitable for selling with a bad back or (13:12) dodgy knees, you know, that sort of, that's going very granular there, but just to give you an example of (13:17) the depth people should be going to, the reason I say that is because the number of five plus word (13:24) searches people are searching now since ChatGPT went live last year has increased (13:31) so significantly that people are searching way more specifically now, so people would search for (13:37) I'm a man looking for a hiking backpack in colour black, it needs to be 100% waterproof, (13:46) it needs to be suitable for someone with a bad back, those sorts of questions people are asking (13:51) LLMs and ultimately, in order to appear, your product to appear for that, you need to have the (13:56) right product information on the product page, if you don't and it's not clearly laid out, you're (14:00) not going to appear for those specific terms, you're just going to keep trying to appear for (14:05) those really basic searches that people are used to doing, like the traditional searches, but not (14:11) really these much more long tail versions of it, which are way more specific, they want a specific (14:15) type of product that's only suitable to their criteria. Yeah, so it seems like really merchants, (14:21) what they need to be doing now is kind of going back into their, I guess, product listings, (14:26) making sure any accurate data is there, I guess, more accurate detailed information as much as (14:32) possible.
Yeah, and I'm curious, do you think, when it comes to customer reviews, do you think (14:41) now it's also even more important than ever for, I would say, Shopify SEO or any, I guess, AI tools (14:48) out there, would you say it's even more important now? Yeah, I mean, I think reviews have always (14:54) been really important from a customer conversion standpoint, but from an EAT, which is expertise, (15:02) authoritiveness, and trust, experience, expertise, they added the extra, from that perspective, (15:09) social proof to say a product has got 3,000 reviews, five stars, is a pretty significant (15:16) indicator that that product is good. If that's not there, and ultimately, we all shop, if their (15:21) product reviews aren't there, and I'm a pretty, I'm bad at this, but I guess, good for it as well. (15:25) I will always do my due diligence.
If they haven't got reviews, you know, you're always way more (15:29) sceptical. So, there's two sides to it. Google really likes it because it shows the value of a (15:35) product, people like to engage with the product, and that is a good product, which is a good tick (15:40) box from brand advocacy from Google, but then from a customer standpoint, it's also really essential, (15:45) like, and obviously, these days, with customer reviews, you've got the AI highlights of the (15:50) reviews, you've got, like, there's categories within the reviews, so if we're talking about, (15:55) I don't know, trainers might say, comfort, fit, you might have three, four different categories (16:01) that actually fit the review.
Is it suitable for a wide foot, for example, and then that just adds (16:07) even more value for customers, but also for search engines to say, okay, this product is really (16:12) good for people with wide feet, really soft cushioning, it's good for road running, and that (16:19) that's just so valuable, and this would also tie really well into shopping and LLMs, because (16:24) ultimately, if you're being shown 10 different products related to your search, and of that 10, (16:31) only two have actually got reviews, you're going to look at the ones with reviews rather than the (16:36) ones that don't have reviews. So, really valuable there, and the same for agentic as well. Ultimately, (16:42) reviews are still going to be really important there.
A brand might have an agentic storefront, (16:46) and on that agentic storefront, you might be interacting with customers that have always (16:52) shopped with you, but they also still want to see product reviews, because otherwise, why would they (16:55) trust what the agentic shopping experience is showing them? So, I think ultimately, what I'm (17:00) saying is, they're very important, they're not going anywhere, and they're probably only going (17:04) to become more important. Yeah, and would you say there's any other strategies that Shopify stores (17:12) can do to, I guess, optimize their reviews to make sure that they're maximizing all of the (17:18) benefits that they can get from SEO? A big one we often do, and I think because reviews platforms are (17:26) often apps, and they're often client-side rendered, is we try and get clients to connect via API to (17:35) customer review platforms, so that they can host the most recent five, say, blogs, reviews, (17:42) sorry, directly on the page, because otherwise, what happens is, they're only discovered once the (17:48) JavaScript's rendered, because it's an app that has to render JavaScript to then see the reviews. (17:53) So, that's often a big one, because then Google will crawl the page, sees the HTML (17:57) review straight away, and that's immediately way more efficient experience than them having to (18:02) render the JavaScript.
Additionally to that, the AI review summary, if you can have that, (18:07) and if it's got the functionality, that's really valuable, because customers don't want to always (18:11) read loads of reviews, they might just want to see what is the summary of all the reviews that this (18:16) particular product has, which is obviously really helpful. We're all really lazy now, (18:20) we don't want to do anything too laborious, apparently, so that's a good one. And then, (18:25) finally, to that is something more technical, but having your structured data, making sure that (18:29) you've got aggregate ratings implemented, and that you're using the review section.
So, (18:34) the review section is where you actually include two or three actual live reviews in your schema, (18:41) and then the rating is, if you've got a thousand reviews, and they're 4.6 stars, (18:47) you can list that in the schema, and then that will help you appear in search results with a star (18:52) rating next to your product name, which is obviously really good, especially from a click (18:56) point of view. If you're up against other products and other brands selling the same product, (19:02) but yours has got a thousand reviews, and theirs has got none, you're obviously going to click on (19:05) the one with the reviews. Right.
I guess this is just kind of a follow-up question to that, but (19:11) do you think... I know there's no perfect answer to this, and to be honest, the answer is probably (19:16) both, but do you find that star ratings are more important, or, for example, text reviews are more (19:21) important for a Shopify store? Definitely both, yeah. I think star ratings, you can illustrate (19:27) on Google directly if you have the right structured data in place, but then people (19:31) want to also read the reviews. They want to understand how customers have found that product, (19:34) and they want to actually get real-life insight and understanding.
I know I read the reviews, (19:39) and if I see a set of reviews, one after the other, that are negative amongst those reviews, (19:49) you've got to assess it, because I know there's also some dodgy dealings going on with reviews (19:54) sometimes, and people like to slate a product. So if you've got a thousand reviews, and three of (19:58) them really recently have been bad, but everything else is great, you can use your common sense maybe, (20:03) but I find them personally, when I'm shopping, that it's super valuable. (20:09) Yeah, absolutely.
Well, Tom, I think those are the questions that I really wanted to ask. I think this (20:15) has been such a great conversation. I really appreciate you sharing your experiences working (20:20) with different Shopify stores, and really breaking down things in a very, I would say, practical way.
(20:28) I would love to see if you have any final comments, where maybe our audience can find you. (20:39) Mm-hmm. Yeah, so the final comment is, if there's anything you want to be doing from a LLM, (20:46) an agentic point of view, is go as deep as you possibly can, put yourself in the customer's (20:51) shoes.
What product information would you want to be shown when you're trying to convert? Like, (20:58) go as far as you possibly can. A good place, starting point, is also to use like ChatGPT to (21:03) ask it, what are some really deep-level product attributes I could associate with this particular (21:08) product? Because it does come back with some pretty good ideas, and then it'll give you some (21:12) inspiration also to like, based on the brand, based on who you are, based on your characteristics, (21:17) based on your brand values, what attributes, what product information should we be illustrating (21:23) on our product page, to really sell our products, but also to help give us the best possible chance (21:27) to show up in LLMs. And then how to find me, find me on LinkedIn, Tom Gandy, and yeah, you know, (21:37) I'm working, as I mentioned, with loads of different brands right now, whether that's on a (21:42) project basis, retained basis, yeah, I'm always happy to have those conversations.
(21:49) Awesome. Well, I think for everyone watching, I really hope you found this as helpful as I did. (21:56) Please don't forget to subscribe for more conversations with industry experts, and (22:00) I'll make sure, Tom, that we include all the necessary information in the description box, (22:05) so you can easily find him.
Awesome. Perfect.