Current website audit

[buzznick]

This builds on the target audiences (#19264) discussion. It reviews the current orchardcore.net and shows where the site fails to reach those audiences. It is a starting point for discussion, so please comment, disagree, and add anything it misses.

We have defined two audiences for a redesigned orchardcore.net. The primary audience is .NET developers, and the secondary audience is business decision-makers, the people who fund a project or approve the choice of platform. This audit looks at the current site and asks one question for each of them. Does it give this person what they need?

It mostly does not. The deepest reason is that the current site does not recognise that it has two different audiences at all. Every visitor gets the same single page and the same single path.

What is on the site today

Element	What is on the page today
Hero	"Orchard Core" / "ASP.NET open-source, modular and multi-tenant application framework and CMS"
Primary nav	Getting started · Portfolio · Features
Sections	Hero → Getting Started (6 resource cards) → Portfolio (3 case studies) → Features (6 capability blocks) → Partners
Primary buttons	"Try Orchard" (demo), "Orchard Harvest" (event)
Proof from users	3 portfolio entries (TOKYO Dance Music, Stimata/Kythnos, Property Brokers) and a partners strip

Where the site fails the .NET developer

The primary audience. They evaluate a tool quickly, leave quickly, and trust other developers more than marketing.

There is no clear first step. A developer who is interested wants to get something running fast. The current "Getting started" section is a row of links, to the docs, the source code, Discord, and more. The links are useful, but they are not a guided start. There is no command to copy and paste, such as dotnet new, and no obvious path from "interested" to "running on my machine".

The signals developers trust are missing. Developers judge an open-source project on a few quick signals, and the site shows almost none of them.

• Is it alive? Release frequency, the date of the last commit, and the contributor count are not on the site.
• What does the code feel like? There are no code samples, no recipe example, and no screenshot of the admin.
• Is there a community? The Discord and the podcast exist, but the site gives them no prominence.

The project has all of these things. The site simply does not surface them.

Where the site fails the business decision-maker

The secondary audience. They rarely write code, but they decide whether OrchardCore is considered at all.

There is no business case. Nothing on the page helps a CTO, an agency owner, or a content lead argue for OrchardCore over the alternatives they are weighing, whether a paid platform or another open-source project. There is no sense of total cost of ownership, no clear explanation of licensing, and no hosting or deployment guidance. The site never sets out, in business terms, why a team would choose OrchardCore over the other options.

The proof a business looks for is missing. A business reader wants to see that real, recognisable organisations rely on OrchardCore. The site shows three portfolio entries and a strip of partners. There are no customer logos, no named organisations, and no numbers that show scale, such as sites built, downloads, or deployments. OrchardCore reached its 1.0 release in 2021, and it builds on Orchard, an open-source project running since 2009. The project has a real history of production use, and the site shows almost none of it.

Problems that affect both audiences

The page does not answer either audience's questions. This is the deepest problem, and it is the reason the two sections above are both so short of what their audience needs. A .NET developer evaluating a framework and a business owner evaluating a CMS arrive with very different questions. The current page is written for neither of them in particular. It gives every visitor the same general description of the product. A developer's questions and a business reader's questions are both left unanswered. How the redesign should answer them is a decision for the design work, not for this audit.

The headline names the product, not the value. The hero headline is the product name. The subhead below it is a list of features written as one sentence. A developer does not learn what they could build. A business reader does not learn why it would matter to them. A strong headline leads with a result the reader wants, not with a description of the architecture.

The visual design looks dated. A few things stand out.

• The hero is heavy with text and has no product imagery.
• The card layouts work, but they look flat.
• There is no motion and no live preview.
• The orchard idea is right there in the name, but it never appears in the visual design.

The site is hard to find through search. The site does not appear to compete for the search terms a developer or a business would actually type, such as "open source .NET CMS", "ASP.NET CMS framework", or "multi-tenant .NET CMS". A redesign is the chance to compete for these terms.

The site gives AI answer engines little to work with. Many developers now ask an AI assistant before they open a website, with questions like "what is a good open-source .NET CMS?" or "what is OrchardCore, and what is it for?". The answer those tools give is assembled from clear, factual, well-structured content they can find across the web. Making sure that content exists, and that it is accurate, is sometimes called Generative Engine Optimisation, or GEO. The current site gives these engines very little to work with. The homepage is a short feature list, with no plain-language explanation of what OrchardCore is, who it is for, or when to choose it, and no question-and-answer content or clear "why" narrative. So when an AI assistant describes OrchardCore today, it is working mostly from third-party sources that the project does not control and that may be incomplete or out of date.

There is no way to measure how the site performs. The current site has no analytics or tracking code. There is no Google Analytics, and no privacy-friendly alternative either. Because of this, no one can see how many people visit the site, which pages they read, how they arrive (search, a link, or social media), what they search for once they are on the site, or where they give up and leave.

This matters in two ways. First, the findings here are based on judgement and on what visitors generally expect, not on data about real OrchardCore visitors. Second, once the new site is live there will be no way to tell whether it is actually better than the old one. A redesign should be measurable. The project should be able to see, for example, whether more developers reach the "getting started" step.

Some form of measurement is therefore needed. Which tool to use, and who owns and reviews the data, are decisions for the design and build work, not for this audit.

What the current site does well

Not everything is a problem. These are worth keeping.

• The portfolio cards use real screenshots and clear industry tags. This is a strong pattern to build on.
• The feature areas (Modular, Multi-Tenant, Multi-Lingual, Security, Workflows, Media) are the right list. They are just not promoted well.
• The Orchard Harvest event shows a living community. The site should make it more visible.
• The public podcast and Discord links show an active project. For developers, these belong near the top of the page.

In short, the current site reads like a project README turned into a web page. The goal of the redesign is to make it a website that respects the intelligence of developers and gives business decision-makers a clear reason to consider OrchardCore.

[Piedone]

I don't think there's much to keep :).

BTW whenever I prompt an AI tool for .NET CMS platforms in some way, OC is there. Just for illustration, ChatGPT (left) might be biased, though, since I use it a lot, but much less Gemini (right):

[agriffard]

Here are some slides to summarize the discussion:
OrchardCore_Website.pptx

[Piedone]

Was there some kind of discussion on the meeting, or did you generate this just from this thread?