August 24, 2026

Why make a B2B SaaS marketing website?

Pondering my role, purpose, and experience relative to a rapidly evolving software industry stemming from the LLM singularity.

For the first time (maybe) ever, the future of B2B SaaS marketing sites is unclear to me. I have been growing continually skeptical of their value relative to the investment needed to build and maintain them. Whether or not users need them to adopt a piece of software is questionable, at least in their current format.

Personally, I find myself spending much less, if any, time on software provider's websites. I have not subsituted the medium. As in, I do not evaluate new software in a different place—rather, I find myself rarely evaluating software in the first place.

This makes me wonder if a B2B SaaS provider's marketing website ever truly influenced my decision in adopting it. When it comes to the tools I have recently adopted, I don't think that they had a role. For example:

OpenAI: I know what their website looks like, but I do not recall scrolling their homepage or any subpages for a purpose outside of design reference and understanding their marketing approach.
Anthtropic: I'm not sure I even know what Anthropic's website looks like outside of the Claude login page or app dashboard.
Cursor: I have viewed Cursor's marketing website a handful of times and scrolled it—not because I was curious about the product or features but because their design team prominently shares and markets their own work.
IA writer: This tool is the exception of the bunch in that it's not an AI tool. That said, I had heard positive reviews from coworkers and only viewed its app store page before adopting.

As a person who makes a living designing and building these sites, this is a concerning thought. When I first had it, my initial reaction was to attribute it to the rise of LLMs and chat interfaces. After consideration, it could be that the shape of B2B SaaS marketing sites, in part, acted as a tailwind for these new tools.

Marketing and user needs

As designers, we seek to create things that meet a user's need/s. However, who needs marketing in the form of a multipage website? While it is true that users willfully visit any B2B SaaS marketing website, are they looking to be sold to? Do they care about the latest features? Do they need abstracted pictures of the software itself and a bunch of icons? Surely, they want something from this website—but let's be honest, we generally have little-to-no idea of what that is.

Do users even want to be here?

Being that you are here reading this, chances are, you a psycho like me who prioritizes screen time over meaningful things in real life. Not that I can relate, but my guess would be that most people would rather be doing just about anything other than working.

When a user visits a B2B SaaS marketing website, we can assume they are putting effort toward completing a work-related task. At which point, we can also assume they would rather be doing "just about anything" other than this (i.e. memes, spending time with their kids, exercising, petting dogs, eating, staring at the sun, etc).

The current user experience

Despite not wanting to really be here, we greet them with gradiose, unclear benefits-based messaging about work-related software paired with a decorative interface that has a Javascript-powered background. It's why we designers get paid the big bucks, jazzing up some insanely boring content. Surely, the only blocker between them using the software we sell and the problem they seek to solve was this...

Now that we are actively wasting a person's precious time to dole out an incoherrent sales pitch, let's think about where they can go from here:

The user can scroll to see more of this bullshit
The user can navigate pages to see more of this bullshit
The user can contact sales to get a call from a new grad
The user can sign up to give the software a test drive (and receive drip campaign emails until they reach their grave)
To avoid all of the above, the user can hop on over to an LLM and ask it to gather the information they seek almost instantaneously

Looking at this medium from this lens, is it that shocking most B2B SaaS marketing websites have abysmal scroll-depth measurements and, on average, <2 page views per user?

Harsh, but capped in potential

The above may come-off like it was written in anger or resentment. Rest assurred, I am sitting here in a peachy mood and and genuinely, solely, curious. After writing it, I too asked myself the following question:

If you think this about these websites, why are you working on them? Why not go do something else?

I truly love the internet and the process of making websites. While I tend to love personal websites a lot more than business websites, I do think that businesses have potentially more impactful pieces of content and information to make available on the world wide web. That said, at large, I don't think the B2B SaaS marketing websites we see today have yet figured out what that content is or looks like.

Limited by design

The shape of digital business surfaces has not evolved over the years, this contributes to the problem. A B2B SaaS typically has four surfaces:

Their marketing website
Their documentation website
Their support website
The app itself

It is easy to imagine how the first three may be better if rolled into one. Essentially, this results in a singular space that tells users:

What the product does
How to use the product
How much the product costs
How to get help with the product

The above is all critical info. The amount of information held in the above warrants a 2,000+ pixel tall webpage and is worthy of a user's time. Decoupling "what a product does" from the rest to construct a marketing webpage has resulted in fluff. If marketing websites were to solely capture this in a succint way, they would be awkward in shape.

Easily replaced by docs

Most businesses could document mission critical marketing information on one page without even needing users to scroll. This information is only interesting if it solves a real problem the user deeply feels. Though, it's unlikely a user feels deeply about something work-related. More likely, a user just wants to know enough to complete their task.

At this point, we would have weird, tiny websites that deliver users a fraction of the information they need. Large web design agencies and teams would lack justification of their size. If a business were to come to this conclusion, the easiest solution would be porting the few important bits of content from a marketing website over to their documentation site. It is important to clarify this because it would not be easy to do the inverse. Marketing websites were not designed to simply present dense content in an easy-to-navigate way.

This is why LLMs love docs websites. Many docs sites are already doing this. If chat interfaces continue to be rapidly adopted at work, marketing websites very well could become a nicety rather than a necessity—only serving the few who lag in chat adoption or need pricing information.

Inherently boring

The above caps the potential of B2B SaaS marketing sites by siloing information that would be useful in one place across at least three different surfaces. Potential is also capped by the subject itself.

B2B SaaS product are inherently boring relative to what we see on the rest of the internet daily. For instance, yesterday, I was watching a collection of clips where people pranked their dads into thinking they broke their family's TV. The entertainment value of it was superior to any B2B SaaS site I've ever seen—even the best of the best: Stripe, Vercel, Linear, PostHog, you name it. None came close.

Again, this is inevitable. These sites deal with work and, day to day, most work is not compelling or solving generally intriguing problems in itself. Another for instance: the excitement of hosting and deploying an enterprise website, at scale, pales in comparison even to videos of cats reacting to water in funny ways...

Paths forward

Thousands of pixels deep into this article, thus far I have entirely, unconstructively spewed negativity toward the work that I do every day. Before solutioning, I needed to think through where I believe the industry I work within is at and where it is headed.

Since I'd prefer to retain my employment by offering meaningful value to businesses, I will now pivot into exploring potential solutions. It's worth noting anything I come up with here may have glaring holes that I simply miss as I write. Please feel free to tell me if there is.

Let's start with the most obvious based on what was written above.

Merging surfaces

We talked about how B2B SaaS companies tend to have four distinct, silo'd web surfaces: marketing, docs, support, and the app itself. We also mentioned how three of these might be easily brought together into one (merge everything into docs). However, might it be even better to make this all happen in just one place?

A dashboard with everything

I'd imagine the ideal user experience for any tool is to quickly:

See what it looks like
Understand how it works
Understand the capabilities
Know what it costs
Find specific support

To put this in context, ideally, rather than indirectly learning about a feature, then having to go sign up, then re-find it in a dashboard, and then try it for myself—I should be able to try the feature as I learn about it relative to the business task I'm trying to accomplish.

Describing an all-in-one experience

The first challenge in this approach would likely be discoverability. If a user were to be dropped into a dashboard experience prior to account creation, how are they going to find the specific feature within this dashboard that is relevant to them?

A minimal amount of intent capture outright can solve for this. In application, this could be a brief modal that appears upon landing that allows a user to simply describe or pre-select what they are hoping to do. Categorizing this input is a simple task for any LLM.

From there, a user can be guided through how the product serves their task via coach marks, inputting their own info along the way. This eventually amounts to the user seeing exactly how the product accomplishes the task they set out to complete.

If there was a hangup along the way, reaching support should be efficient and contextual. 99/100 times, a half-way decent LLM widget trained on the code/docs can likely help resolve snarls in an instant. In the case that it doesn't, live support should be available even to prospective users.

The outcome and/or output of the task should be visible in some fashion but likely gated from production use by account creation and (potentially) payment. At this point, the user should see the explicit pricing for the task they completed or the feature/s they used alongside the sign up form. No more trying to correlate marketing naming conventions with in-product experiences on pricing pages.

Example all-in-one experience

The simplest example of this type of flow is likely for website deployment and hosting. It is especially relevant in an era where a bunch of non-coders are vibe-coding new websites with no familiarity of how this process works.

Imagine you just created a new website using Cursor and needed somewhere to deploy and host it. You know a tool like Vercel can do this sort of thing but haven't the first clue of how it works or what it costs.

You head over to vercel.com and are first asked what you are trying to do. Since the task you are trying to accomplish is common, "Launch a website" is one-click on the form.

At this point, the screen asks, "How did you create your website?" Let's assume you vibe-coded it in Cursor, so you select "Vibe coded in an IDE (i.e. Cursor, VScode)". By default, Vercel assumes you don't understand Git or have a Git provider, so it then gives you a click-to-copy prompt you can drop into Cursor to use Vercel's CLI to sync your code to Vercel.

After an agent executes the prompt, its reply contains a link to a temporary domain. You can now see your website live, on the world wide web. At the same time, there's a small banner annotating the domain is temporary. To have it persist and/or connect it to a real domain is gated by account creation.

You click the button to create an account. The form that appears includes the pricing information and tiers. Since you are launching a new site, you realize this is all going to be free for a while and happily proceed.

Because you were so satisfied that the website was live, you hadn't considered how to update it going forward. Fortunately, Vercel anticipated this. After creating an account, you are met with a next steps screen that captures exactly how to do this. It also provides a few other features you may find cool, personalized to you based on your site's code/content (upsells).

All-in-one recap

The user in the above experience did everything that typically happens across three surfaces (marketing, docs, and app) in one productive flow. While they didn't need support, it would have been triggerable from the bottom or top right of the screen along the way.

Yeah... But, enterprises.

This captures self-serve flows well—but what about big business that has a plethora of security rules and evaluation procesesses? I have yet to see a top-down eval process for new software. It almost always comes from the bottoms up. As in, an engineer goes and tries it before suggesting adoption. At this point, the above still works. Sure, there may need to be a static privacy policy page, but that stuff takes little effort.

Blogs, thought leadership, and changelogs

The first solution only captured product-specific marketing applications, which only covers a fraction of what most B2B SaaS marketing sites contain. The real doozy of marketing sites is their blog content, thought leadership, and changelog. There are usually hundreds if not thousands of these pages attached to every B2B SaaS site.

Most often, these are unnecessary

I have designed and built a lot of blogs and supported numerous thought leadership initiatives during my time as a designer. They are a complete pain in the ass. Aside from every CMS being hot garbage, for the effort that goes into producing content, the visits and business that stems from this work is underwhelming.

Again, this is not shocking. The content shared here is rarely, if ever, required in a prospect's journey of accomplishing a task. Which means, to read this content, someone must choosingly do so. We already covered how most people aren't psychos like us who do this shit for fun. We also covered how the subject itself is relatively unamusing.

It is also the case that most companies writing blogs have no business writing about business. Is being told how to work or do something by a startup that is likely going to fail in the next two years reasonable? Few care about the newest feature of a piece of software that is irrelevant to them. Content marketing is a shotgun with bird shot, not a precision tool. It is commonly, by definition, spam.

Sure, sure, there is a scenario where the company writing content is established and/or on unicorn trajectory. The thoughts and ideas they are sharing could very well make a meaningful difference in your own business, if adopted—but how often is this content focused on that versus something self-serving?

The subject needs to change

We see a plethora of unaffiliated business content influencers doing quite well for themselves. Take for instance, Lenny's Newsletter and Podcast. In these instances, more often than not, high-level employees at ultra-successful companies are being cast for interviews.

It makes sense that people want to hear from this persona as they too would like to find themselves in such a role. Me, personally, I find it largely unhelpful and unattractive. If the persona is not one of the primary business operators or someone who deeply understands both the strategy and how to do the work itself, why bother?

In my own role at a reputable company, I realize there is a certain, large degree of luck involved. There are many designers as talented as I am who are putting in just as much effort with less reward. I believe the same to be true about most director, head, and managerial level roles. In many instances, a series of fortunate events played just as much, if not more, of a role as personal merits in success.

Rarely do we see this being disclosed in any content. Rather, the focus is on what "we" can do to have similar outcomes. To be fair, telling people to hope for luck is not very constructive. It is also unlikely that operators and people actually doing the thing have time to continuously churn out well-formulated new content.

Meaningful business content

As you've witnessed in this article, I am by no means an author or content strategist. Please take my opinions here with a grain of salt. I am solely exploring solutions based on both my experience as a user and a designer with observed user behavior in mind.

Since almost all businesses don't have a Jenson Huang or Patrick Collison, there must be another way to contribute. It's worth noting the term "contribute" is intentional. We already covered that nobody needs (or, let's be honest, wants) content marketing. If we choose to continue with content, it cannot be entirely self-serving.

Other people must see value in it for themselves—trading their time to read your perspective. This may not directly contribute to business growth as the act should aim to be selfless in nature. The shape of this is likely transparent—sharing successes, failures, and chance that may be applicable to anyone or any business. Publishing at a business-level, it is unlikely to be industry-specific. I.e. an article about how spending millions on trad content marketing turned out to the wrong move...

If that doesn't sound compelling—can we simply reallocate those resources toward contributing to the first part of the solution 😂?

In the case this still sounds interesting, what does this look like? There are few examples to pull from here. I have not played any role in it, but Stripe has always been a leader in the B2B SaaS content space and may be a north star.

Stripe Press, Works in Progress, Increment Magazine, and Cheeky Pint contain compelling content in themselves, even when decoupled from Stripe's business specifically.

Here, I am going to stop exploring path's forward for this type of content. Mainly, I don't know what else I have to add that may be meaningful. Let's hop to changelogs.

Changelogs

I'm not sure changelogs were ever meant to live on a marketing website. The idea of announcing and explaining new features mostly intrigues existing users, not prospects. Since dashboards generally require authentication to access, my guess is they were placed here to account for the fringe chance a prospective user was curious.

Now that the first part of this solution takes care of that (allowing prospects to access a dashboard). I think it's safe to say moving changelogs into the dashboard also makes sense.

Beyond meeting users where they already are, this placement of changelogs might allow for more reasonable promotion of the features being announced. For example, if a user who uses features A, B, and C is likely to care about a new feature being talked about in the changelog, serving them a small notification for it should be no big deal. This could have occurred prior, too. However, the coordination would have been much more challenging. Likely, this would involve staffing from two separate teams (one built the feature, one helped announce it). In theory, the same team who built the feature should also handle announcing it. Bouncing users from a dashboard to a marketing site is also less than ideal, assuming a user was in the dashboard to get some work done in the first place.

So, should a B2B SaaS marketing website be made?

To conclude, if I were to bootstrap a new B2B SaaS myself, I don't think I'd invest in a dedicated marketing website. Instead, I'd direct those resources towards making incredible, un-gated onboarding.

The question of whether my skillset is still relevant is unclear. Though, I'd like to believe that great brand or web designers are exactly what are needed to turn complex, ugly product experiences into something more approachable, efficient, and delightful.

That said, we must not simply drop traditional marketing experiences into products. Demonstrated by the rapid adoption of LLM interfaces at work, it is clear users are not looking for fluff.

What is left

Essentially, this whole article's purpose was to help me better understand where I must adapt to continue providing value to retain employment. After writing, I believe that the next era of my career will be defined by my ability to conceive and execute screen design that fosters elegant efficiency while being highly functional.

To be precise, when it comes to work, users will continue to seek and expect higher levels of efficiency from the screens they view, in even the most basic of tasks. They will not long for decoration as much as an escape from work and their computer screen entirely. Here lies their motivation. It is unlikely they will feel that they can escape without increasing their output. This feeling is driven by, and will increase, as LLMs continue to decimate (or at least consolidate) white-collar employment opportunities.

Similar to my motivation in writing this article, I believe the next era of design will be driven by people's own sense of business-related performance and employment security. Anything slower than an LLM interface will fall by the wayside.