Where to go after "Get started"

[bjnath]

The big Get Started call-to-action is wonderful. I'd like to cede more power to it.

It's in competition with the top line ("Install...Documentation"....). If the CTA is the only way into the site, people will press it gladly. If there are other ways, both alternatives lose out. Either readers won't see the top line, and they'll press the CTA and find it's not what they wanted, or they'll see both and wonder what Getting Started is if it isn't one of those other things.

The top line is hard to see on the home page, and as I argue in #238 space above the fold is at a premium. We might hand its job over to Get Started.

There would be no top line, and rather than forcibly landing the reader in Learn, Get Started would open up a menu of the options that are currently on the top line.

[bjnath]

Please allow me to take another swing at this, because I didn't express it clearly:

Rather than bringing the user to a predetermined page, pressing Get Started should present the user with a menu that has the same options as the nav links.

Rationale: The user shouldn't have to guess what Get Started does, and we shouldn't presume to know what the user wanted when they pressed it.

Forget what I said about the nav links. Web pages can and should have different ways to do the same thing. My point in passing was that the links on the home page aren't very prominent and we might benefit from saving room at the top.

[rgommers]

What I had in mind all along is a dedicated place that "Get Started" button would go to. It's indeed not clear if one wants Install, Learn, Docs, or whatever. Those pages may also not have enough context, a dedicated page would be better. We just haven't written it yet, and decided to link to Learn as the most reasonable of available pages.

[bjnath]

That makes sense. But by making users first land at a central page you're making them change planes when they could have taken a direct flight.

Remember that each of the Install, Learn, etc. pages are themselves hubs, so there'd be two indirections for the user rather than one.

Thus I'd favor a drop-down menu over a dedicated intermediate page -- which, in addition, we'd have to write and maintain.

[InessaPawson]

My point in passing was that the links on the home page aren't very prominent and we might benefit from saving room at the top.

I don't think ceding the main navigation bar is going to improve user experience. I'll experiment with increasing the size of the font though.

Thus I'd favor a drop-down menu over a dedicated intermediate page -- which, in addition, we'd have to write and maintain.

Drop down menus as well as hover effects are notoriously frustrating on mobile. UX designers usually avoid them nowadays.

[bjnath]

don't think ceding the main navigation bar is going to improve user experience.

Neither do I. I meant that if we lost it, with the goal of getting more content above the fold, it wouldn't hurt user experience, because in the layout now the bar is barely visible.

Drop down menus as well as hover effects are notoriously frustrating on mobile.

Yet I see hamburger buttons; clarify for me how a hamburger button is neither of these.

[joelachance]

@bjnath, do you have an example of a CTA that is a dropdown? I'm not sure I've seen one before.

[bjnath]

Point taken. I guess it stops being a CTA if it presents choices. But it can look like one (saying "Get started") but underneath behave like a hamburger. Or so I naively imagine.

My thought is that people will be relieved rather than annoyed. "I'd love to get started...but what is that, exactly...?"

[InessaPawson]

My thought is that people will be relieved rather than annoyed. "I'd love to get started...but what is that, exactly...?"

I see your point. Should we think of a better wording for this button to avoid the confusion?

[InessaPawson]

Yet I see hamburger buttons; clarify for me how a hamburger button is neither of these.

Some get by without using external style sheet files, it doesn’t make it the best practice.:)

[bjnath]

Some get by without using external style sheet files, it doesn’t make it the best practice.:)

I'm not sure how these are connected. I can't imagine a commercial site failing to use external style sheets, but I see expanding icons and links everywhere on mobile. (Hover, I agree, is nasty.) Where's the "Hamburger considered harmful" paper?

Also, to make a comment that may not much elevate the conversation, I'm unclear how the shell fits in with this help-the-mobile-user fervor.

[bjnath]

Should we think of a better wording for this button to avoid the confusion?

Would it be rude to ask why the button was put there in the first place?

[InessaPawson]

Would it be rude to ask why the button was put there in the first place?

Primarily to minimize the number of buttons in the main nav bar. “Getting Started” guide is important enough to be given a special button.

[InessaPawson]

Also, to make a comment that may not much elevate the conversation, I'm unclear how the shell fits in with this help-the-mobile-user fervor.

The shell can be simply ignored, with the menu it's harder.

Design is subjective. In UX usability tests help. Since we are not in a position to carry out formal usability tests we have to rely on our experience and intuition. Personally, I try to avoid drop-down menus whenever possible. In some instances, e.g. e-commerce, drop-down/mega menus are unavoidable.

[bjnath]

“Getting Started” guide is important enough to be given a special button.

Based on what?

[bjnath]

The shell can be simply ignored, with the menu it's harder.

That is true.

our experience and intuition

Mine is telling me that a big button pulling users onto a specialized page is harmful to the usability of a site that must serve

many kinds of users: students new to programming or Python, educators, researchers, domain experts in one of the areas that NumPy covers, data scientists, library developers, packagers, and more.

(from NumPy's GSoD proposal).

[bjnath]

Based on what?

Inessa, I see that in fact you've already answered this in #234, let me study that.

[bjnath]

Since

• we're expecting users seeking specific content to head directly to the nav bar
• we're characterizing the button as no more than a supporting player on the page, albeit for a large class of users, and
• we're data-driven about UX

please take a look at the first line of Nielsen Norman Group's Is Navigation Useful?:

For almost seven years, my studies have shown the same user behavior: users look straight at the content and ignore the navigation areas when they scan a new page.

To me, that says they'll head straight for the button.

So rather than dedicating the button to some subset of users and excluding every other user, I'd think of how it, or something like it, can be made to serve everyone.

[bjnath]

@InessaPawson asks:

Should we think of a better wording for this button to avoid the confusion?

Let's make the button say "Learn NumPy".

[rgommers]

Had a look at other sites as well. PyTorch links to install instructions, TensorFlow to user guides, Julia and Jupyter have multiple more explicit buttons instead:

Little consistency here. Anyway this is temporary; we'll get a dedicated "get started" page, or two buttons, or .....

[rgommers]

@InessaPawson didn't like Learn NumPy much, and I'm also ambivalent. I think the final solution will include either a custom page, or multiple buttons. In the meantime, I think we'll leave it as is. Running both out of energy and runway for launching this weekend here, and the site looks really nice.

[bjnath]

I'd like to hear her reasoning -- but realize today is not the day.

One subject we're surely unanimous on: it does look dazzling.

[bjnath]

Didn't realize Get Started goes to the Install page and not Learn, sorry. Finding oneself on an Install page after a Get Started CTA isn't unreasonable.

I would not advocate changing Get Started to a Learn button that goes to Learn. If that's what you thought I wanted, you don't have to talk me out of it.