‘We’re particularly excited to be an open platform’: How BigCommerce is investing in AI with the help of Google

But BigCommerce CEO Brent Bellm said that even before the GoogleCloud partnership, the company was already investing in AI thanks to its acquisition of Feedonomics in July 2021. Feedonomics is a software startup that takes brands’ product data, and uses it to create the most optimal product listings depending upon whether a merchant is selling through Amazon, eBay, Walmart or other marketplaces.

“Even before the AI wave began, late last year, we already were the global leader in AI for e-commerce by far by virtue of the AI behind Feedonomics,” Bellm claimed. He added that the integration with GoogleCloud is already helping BigCommerce’s Feedonomics clients improve ROAS by 10% to 20%.  

Bellm joined BigCommerce in 2015 with a focus on expanding the company’s efforts to serve mid-market and large enterprise customers. “That’s really what we’re most designed for now. And we continue to serve small businesses, too. But you know, we’re best for small businesses of size and complexity,” said Bellm.

Prior to joining BigCommerce, Brent served as Chief Operating Officer and President at HomeAway during a period in which the company went public. Brent also spent more than nine years serving in a variety of executive leadership positions at PayPal and eBay.

In an interview with Modern Retail, Bellm talks about upcoming AI product launches at BigCommerce, best use cases of AI for e-commerce and what’s next for software-as-a-service tech providers. This interview has been edited for length and clarity.

What can you tell us about BigCommerce and its AI strategy?

Even before the AI wave began, late last year, we already were the global leader in AI for e-commerce by far by virtue of the AI behind Feedonomics. The whole basis for Feedonomics, among other things, begins with an AI engine for taking merchant or brand catalog data, and then optimizing and matching it for Google and Google schema.

Typically, the catalog data coming out of, you know, whatever a brand, or merchants source of truth is only 60% optimized for Google schema, which is not good about because you usually get rejected by Google Shopping, when you only have a 60% match. Or if you are accepted, it’s just going to perform terribly and your return on ad spend will be really bad, your performance will be bad.The AI that Feedonomics has gets that to 95%.

Why is it so important to get merchant listings right on Google?

A third or more of all e-commerce sales in the U.S. on virtual websites were last click source from search engines like Google. The Google experience is now optimized around Google Shopping at the top. And then ads and listings — it’s not SEO any longer. You can’t make up for keywords in your title description, you have to actually follow the Google schema for items that need to be appearing in Google Shopping, ads and listings to perform well. Even though Feedonomics is a managed service for brands and retailers, the AI gets them to 95%. Now, the AI is optimized for Google. But the good thing is that the Google schema is more or less what all the other ad channels have copied. So, social networks, affiliates, display ads, they mostly have copied the Google schema. And, so, if you do the first step of really optimizing your catalogue to perform on Google, then it’s quite easy to have other channels and leverage that same work and then add the other channels.

How is this changing the way BigCommerce operates tangibly speaking?

Everything we’re now doing in AI since the Google and the OpenAI and the Microsoft announcements is additive to that. And we’re particularly excited to be an open platform so our merchants and our agency and tech ecosystem can use whatever underlying AI models or algorithms they want. But, we really like working with Google, in particular, because Google has a lot of data privacy controls built into it, meaning the merchant can make sure that the data that they are using to train their AI models for their brand, are not being used by others.

What are some of the other tools or announcements that we can expect before the end of this year with regards to AI?

There is a whole bunch of very obvious applications of AI that we expect to be rolling out before the end of this year. I mean, things like auto generating descriptions. It’s that sort of thing that you can train an AI model to do to try to write your descriptions for you.

I’m particularly excited about giving AI access to our knowledge base. If you imagine the way customer service agents serves merchants, we have a whole bunch of knowledge base in the background — answers to questions, information about BigCommerce. And if somebody calls up on the phone, or starts to chat with one of our customer service agents, if the agent doesn’t already know the answer to that question, they’ve got to go in and use the knowledge base. But that can be time consuming. That can be hard.

Whereas AI if it gets access to all the knowledge base, it can then spin out answers to customer questions so fast, and maybe without ever even requiring an agent. You put it into a little chat bot, you type in your question. AI accesses all of our internal knowledge base gives back the answer. And you may not have to talk to the right person. Or if you do talk to a person, the person is going to use AI, because it’s such a faster way to pull the right answer from the knowledge base.

And what does that mean in terms of investments?
In terms of ranking our product priorities now and going into next year, AI and AI capabilities is going to be top five or six. It’s not number one or number two, but they’ll be top five or top six.

What would be the top product priorities for BigCommerce?
Things that are ahead of it include, well, just always enterprise level quality like uptime page speed. Enterprise level qualities is always sort of number one, but others that would be on that list would include multi storefront capabilities, B-to-B capabilities. As an example, this year, in May, we released our new B-to- B buyer portal, which is an extremely good user experience. And now we’re adding invoicing capabilities to that buyer portal.

What are some of the best use cases of AI in e-commerce?

The way I look at it, like every team, on your team within your team should be focused on driving profit and efficiency. And I think AI for customer service should help turn that service into a profit center. It’ll help turn shipping into a profit center. So, there are efficiencies with time and money and returns that I think will become more strategic. For example, we’ll probably see use cases around “where should I return my products to?”

What do you feel like is next for the Saas technology platforms?

I see less fragmentation. I think the big platforms are going to just increasingly focus on us and Shopify as the two major SaaS platforms at scale for collaboration. And then for merchants who aren’t on one of those two platforms. They’ll look at Feedonomics as an intermediary.

Are you suggesting it’s going to be a BigCommerce vs Shopify world?

The big difference between the two of us [is]: they’re very opinionated. Shopify wants you to use their payments, their point of sale, at one in point time their fulfillment. Whereas we’re open, we give you the choice of the best payments provider for you, with a best of breed integration out of the box. You can choose which point of sale is best for you. You’ve got Feedonomics and you’ve got integration into all or any of the ad channels and marketplaces you most want to use. So, it’s that openness and that flexibility. And then [the] big difference, number two is our focus on mid-market enterprise, and complex use cases, we have a lot more enterprise functionality. And that’s where we’re wanting to be a market leader.

Do you feel Shopify is encroaching upon your business with its latest tools?

Not so much taking business away. We don’t see a lot of companies who would be on one or the other platform and then leave it to go to the other. That happens. But not that frequently. Usually, companies that get up and running on us or Shopify are pretty happy. It’s unusual for somebody that will leave us to go to them or vice versa.