The Future of B2B eCommerce with Optimizely Configured Commerce

Integration Tech: Then vs Now

When I started, I was seeing more SQL connections to other SQL connections, sometimes flat files being dumped, like CSVs full of data. And now we're seeing a lot more APIs, which is great.  As well as real time data instead of the schedule. We’re seeing webhooks that can be triggered when something happens, such as an order's placed or you create a new customer in a system.

Seeing a lot more integration platforms as a service. Those are gaining a lot of popularity. A lot of micro services. So, I was using AWS Lambdas recently, which is like a serverless application. It only starts up when it's needed, when it's called, and it uses minimal resources. So, it's just really cool to be able to just get coding and not worry about all the hosting.

And there's also a lot of good tools that are going low code, no code. Kind of allows non-developers to create integrations through a more visual interface. So that's definitely the future, trying to make things doable for everybody and not so specialized.

The Truth about eCommerce Development

The complexity of integration is really something that I don't think stakeholders often understand. That's not really plug and play with the data.

Each system has its own data formats, protocols, constraints, sometimes different structures, completely in terms of the child-parent relationships in the data. So it's not always a one-to-one mapping.

Development does take time and the testing phases are super essential. Need to go through every scenario that we expect to see in real life to make sure that we identify and fix issues before a launch.

There’s a cost vs value element to it to figure out what can be MVP, a minimum viable product. What we can go live with at the beginning and what things can wait maybe to a later phase. This helps us just get up and running and not have to delay things if we don't need to.

When it comes to eCommerce sites, security is a non-negotiable. In every stage of development, integration, front-end, API work,  security really should not be treated as an afterthought. It really should be the main concern. There's going to be ongoing maintenance. That should be an expectation. We're going to be needing to make updates on it to keep it secure and to keep it efficient and also to keep it aligned with the current business’ requirements.

Testing is really critical. You can make adjustments before you go live so that you don't have to have those pain points while you're already live.

By appreciating these different challenges and respecting the different people on your teams, you can better support the project and really come to a mutual understanding that contributes to delivering a really robust, efficient, user-friendly solution.

AI in eCommerce: What You Need to Know

AI algorithms are able to analyze customer behavior and their purchase history and browsing patterns and offer different product suggestions, which just can really increase your sales because when somebody's on your site, you have a limited amount of time to try to get them to want to buy something.

Dynamic content personalization, too, that's just nlogs and articles that the user would possibly be into based on literally what they're hovering their mouse over the longest and things like that.

Optimizely has an AI assistant plugin for the CMS user who creates content for a site, creating pages and creating little reusable blocks within the system. You can basically just give it commands, like, “Hey, create this block for me that explains what our product offerings are on the site”. And it's able to go through and create that whole block for you. And then of course you can go in and tweak it to how you really want it.

AI's being used for fraud detection to figure out different types of purchases that are a little fishy. There are tools using AI to make sure that your site is up to the regulatory compliances.

Those are the key things that come to mind when it comes to e-commerce. It's just allowing people to be more efficient. As with everything else with AI, it's just helping us be more efficient and get things done quicker.

What Not to Use AI For

I would not advise using AI to create a block or a page in and not go over and review what they wrote. There definitely can be some really badly worded phrases that come out of AI when it comes to writing your content.

I would just warn, anything with AI, just always double check it and make sure that everything is how you would see yourself actually creating that content.

What is Optimizely Configured Commerce?

It's a really powerful eCommerce platform. Think about a WordPress, but on a little bit more complex level. Really specialized for businesses with complex product offerings and really intricate purchasing processes. Commonly, in the B2B realm where there's more complex company hierarchy, as well, and a lot of different decision makers doing the buying.

The main benefits are the complex product configurations, supports products with a lot of different attributes, options, and dependencies. Really allowing customers to configure products to their exact specifications. We can tie different pricing structures to different quantities, different customers, and different customer segments.

There's my favorite part, which is the integration capabilities. Great APIs to really be able to do whatever you need to do within the database.

It's tailored for businesses that require more than just your standard B2C online storefront. And whether a company is looking to streamline ordering processes or do more customization with their products, I think it provides all the necessary features that a B2B company would want to achieve their goals.

Biggest Mistakes in Optimizely Configured Commerce

I've seen data that isn't completely normalized. You know, you have missing indexes in the data, which the integrity of the data is basically not there. I think getting that cleaned up first and foremost is important. I've seen more often with a lot of companies that have homegrown built ERPs.  Because they built them themselves, and this data has really evolved over the years for them, and it can get messy.

Other mistakes I've seen is a company trying to go on prem, but not having the necessary infrastructure to support it. That or the planning on how they're going to do the hosting. They procrastinate on it and they're unable to execute in time, so it delays the project and changes launch dates.

How to Nail Your eCommerce Migration to Optimizely

It's a significant undertaking that just requires the right type of planning and execution.

Different departments of a company need to be in sync on the goals so that they all know where we're eventually going. The stakeholders need to be aligned, essentially. You also need to check that the infrastructure can support all the things that we're trying to do. Are the servers going to be able to handle the amounts of data that are going to be flowing? Is it going to be able to handle that and not time out, not take a long time?

You really want to look at the inventory and catalog data, the customer data, the order data, and make sure that it's clean and that there's not unnecessary duplicates and things that could really make the data migration challenging.

Also, looking at the ERP and CRM systems that you use, how easy are they to integrate with? Are there APIs on those systems that could make this easy?

I've seen different departments have different expectations on what certain things are going to look like, and then there's disappointment on one end or the other when they see the result. Getting that aligned as well.

I think it's really important to have a partner that has experience doing these because they have a process and they have different things, experiences, documentation, and templates to be able to jumpstart these integrations.