SUGCON 2019 – Highlights
It has been a while ago: SUGCON London 2019. However, I still believe it is worth a blog post. I’m always looking forward to the international Sitecore meetings. In my years as IT professional, I attended many international meetings regarding many different subjects. To be honest: They could be boring sometimes, but this is never the case at the Sitecore meetings.
Back in the days Sitecore was created for developers. Sitecore created a good base with not that many features. The code structure was very well architected and since then there is a very clean separation between data and structure. They created a code base that is ready for extension. Sitecore is very pluggable. You can add your modules, pipelines, rules and whatever by configuration. You can hook in at virtually every event. Sitecore is simply more than a bunch of ‘bought together’ applications, skinned to look like one software package. Sitecore has a solid base upon which you can create solutions that solve challenges for customers.
Nowadays Sitecore is much more than a CMS on which you can build solutions. They build upon the base, created in the past. There are many features, modules and more out of the box. However, they are still building blocks for creating solutions.
It is always exciting to go to an event and see how the product evolves. I will not exactly rephrase what happened on the event. I will highlight subjects from which I believe they are important.
The frontend developer does not need to know much about Sitecore. They will need to incorporate the JSS SDK and after that, they can almost work as they did before they knew about the existence of Sitecore. The beauty is that virtually all Sitecore functionalities like personalization, automation, and analytics still works. In addition, the WYSIWYG Experience editor will work as it works in classic Sitecore applications. So, developers that have little or no knowledge about Sitecore are capable of building great Sitecore Solutions.
Web development teams use Sitecore Experience Accelerator (SXA) to speed up the production of websites and to reuse components, layouts, and templates across a variety of sites.
SXA separates structure from design, so front-end designers, creative designers, content authors, and developers can work parallel and you can deploy and maintain multiple sites quickly and cost effectively. Once a basic user experience plan is in place, everyone can get started on the platform. For example: the content author can start entering the content in the wireframe environment, while the front-end developer works on the theme, and the developer sets up the data templates.
SXA offers the possibility to let your team act in the agile way. Create a minimal viable product first and after that start adding incrementals. Ad small shippable products and measure the value they add to your business. It’s more than just a couple of components out of the box. The setup is completely Helix compliant. That means the folder structure will be automatically setup according the Helix rules. SXA adds a lot of bits and pieces that will help you to go fast to market. An important thing to take note of is that SXA helps you to structure and build your website in a certain technical way. It is a better approach to not try to change the inner functionalities in the backend too much. Make use of it. Everything is changeable however the more you will do that the more time it will cost you. It will hamper you in going to the market fast.
How to deal with big data? Sitecore with the XP topology activated and XConnect in place gathers many data. So how can you learn from these data and how can you do smart things with it? The answer: use Sitecore Cortex. The question is what does Cortex out of the box and how do you use it in real live. The answer is that the out of the box functionality is still limited. It…
…Suggest better personalization rules Based on internal machine learning models Cortex will suggest you to use personalization rules that might serve your needs better
….Predicts outcomes. While creating tests Cortex will based one machine, learning models try to predict the outcome of these tests. This will help you to fine tune the test
Cortex provides also a processing engine. Developers get a toolbox to create tasks that help them to implement machine learning results in Sitecore.
The Cortex processing engine as a sort of scheduler. Developers can create a task and schedule it to create a learning model that eventually can decide to which persona a user belongs. A next task could be that by using the trained model the Cortex processing engine tags new users with the persona flag they belong too.
The conclusion is that Cortex is much more than a pair of Features. You can integrate specific smartness to the Experience platform from the Cortex framework.
Monolithic or not?
One of the important goals Sitecore aims for is to move away from being a monophyletic application. The Sitecore 9.1 topology already contains nine web roles. The number can grow bigger when using other services as the publishing service and or the Identity service.
Sitecore also introduced Sitecore Host. It’s not a service on itself however; it can be used to create services to do separate tasks in the Sitecore topology.
Very good news is that the search indexing service will be moved away from the content management server. In Sitecore 9.2 search will get its own web role to do the work. Therefore, it is easy to say that the architecture of Sitecore gets more and more service oriented. Separation of concern is always a good thing. The scalability of Sitecore has improved a lot and will improve more in the future.
It was a great event with interesting updates about Sitecore! I learned so much. In my opinion, Sitecore is making interesting steps forward and I really want to be part of these developments. I’m looking forward to the next meetings.