[MUSIC PLAYING]

SUZ HINTON: There
is one term that I’m

going to mention
to you just based

on this is the reason why
I had to submit the URL

to be re-indexed.

And that’s microformats.

MARTIN SPLITT: Oh!

All right.

SUZ HINTON: So can
we talk about–

are they still a thing?

I haven’t really had to do
a lot of SEO optimization

for a while.

And I knew microformats
was such a huge thing

because let’s say you’ve
got a product page

and it has reviews
on it and you want

to show the little stars and all
of that kind of rich content.

And every time I made a
tweak and we deployed,

I would have to then submit
to get re-crawled and see

if the results got more written.

And that was definitely a
very slow feedback cycle.

MARTIN SPLITT: Yes.

SUZ HINTON: So
what is the state?

Is microformat still a thing?

And are there better resources
out there right now for us

to be able to pull
that rich content out?

MARTIN SPLITT: You’re
going to be very happy.

And we have much better things.

SUZ HINTON: Yay!

MARTIN SPLITT: They
are still a thing.

But they are now
called structured data.

SUZ HINTON: Structured data.

MARTIN SPLITT: And we
are using JSON-LD, so

JSON for Linked Data.

SUZ HINTON: Yeah, this
is all new terms to me.

MARTIN SPLITT: Right.

And you probably used
literally the microdata

attributes in HTML.

SUZ HINTON: Yes.

Yep, exactly.

Yeah, we were using them.

And they were very hit and miss.

MARTIN SPLITT: Yes.

SUZ HINTON: It was very easy
to just mess up one tiny thing.

And the validator
didn’t catch it.

And then the stars
would disappear.

And we’d be like [GASP].

MARTIN SPLITT: And we
have moved on from there.

SUZ HINTON: OK, that’s good.

MARTIN SPLITT: So there is now–
schema.org is an open source

organization where people can
submit or discuss or change

or do stuff with
the semantic data

that they want to
put on the web.

SUZ HINTON: Got it.

MARTIN SPLITT: And people
that’s participating–

there is much more
semantic data out there

than we are supporting
in search results.

But a bunch of it is supported
in the search results.

So for instance, if you
have an event that we want

to have showing up
with the location

and if you can get
tickets and who

is the performer and
all that kind of stuff–

if you have a recipe
where you might

have an image or the
instructions on how to make it

or the time it takes
to make it and reviews,

how nice this recipe might
be, articles, books, and TV

series, all sorts of things,
we have documentation on that

specifically as well.

If you go to
developers.google.com/search,

you find all the
supported types.

And they show up nicely
in the search results.

So you get a little
preview picture.

And then you get the stars
and all that kind of stuff.

SUZ HINTON: Oh, this
would have been amazing.

MARTIN SPLITT: It’s fantastic.

And it’s JSON.

SUZ HINTON: Which
is so much easier.

MARTIN SPLITT: It’s the
script tags with JSON in it.

It’s so much easier.

SUZ HINTON: It’s just not
little meta attribute things?

MARTIN SPLITT: Correct, yes.

So you have your JSON block.

And we have what’s called the
Structured Data Testing Tool.

That is a little dated by now.

But it supports– generally,
basically everything

that we know of shows up
as either valid or invalid.

And then we have the
Rich Results Test,

because the Structured Data
Test, while being very generic,

is also not very specific
to what you want to achieve.

You want to probably achieve
the nice little stars showing up

in the search results.

This is what we
call rich results.

And there’s the Rich
Results Test for it.

And that even
gives you a preview

of how that might look
like in the search results.

There’s no guarantee
that it does

look like that in
the search results

because people have been
using it to spam stuff, like–

SUZ HINTON: Yeah, true.

MARTIN SPLITT: I have
a bazillion reviews.

And then we’re
like, yeah, you just

have some JavaScript
generating fake reviews.

That’s not really–

SUZ HINTON: Well, how do
you actually use the tool?

Because I remember you used to
have to dump your entire HTML

file in there.

MARTIN SPLITT: You
[? don’t. ?] [INAUDIBLE]

SUZ HINTON: And if you
did it too many times,

you got timed out.

MARTIN SPLITT: Right.

SUZ HINTON: Yeah.

MARTIN SPLITT: But that
doesn’t happen anymore.

SUZ HINTON: Oh, OK.

That’s pretty exciting.

MARTIN SPLITT: So
you have two options.

You can dump a URL
in it, which is nice.

And you can even use ngrok or
something if you have a local–

SUZ HINTON: Oh, you
could do local host?

MARTIN SPLITT: Yes.

SUZ HINTON: Oh,
this is very fancy.

MARTIN SPLITT: Or
you can even also

still do like you dump
your HTML in there.

We execute the JavaScript.

So if you’re using JavaScript
within that code dump,

that’s fine.

SUZ HINTON: Oh, wonderful.

MARTIN SPLITT: If
you’re running it– yes.

And you can basically
live debug as you type.

You press a button and
it goes like, nope.

And you’re like, oh, damn it.

And you get the feedback here.

And it’s like, missing
performer for your event.

And I’m like, OK, sorry, sorry.

And you write it in.

And then it reruns it.

And you’re like, OK, cool.

This is what I want.

And I can take it
back to [INAUDIBLE]

SUZ HINTON: That is awesome.

MARTIN SPLITT: And
yeah, we have that tool.

We have Search
Console that gives you

a live view of what
happens on your page,

also for structured data.

Yeah, microdata is not
that much of a thing.

But the structured data
is still going strong.

SUZ HINTON: Well, it sounds
like it’s come a long way.

That’s very exciting.

MARTIN SPLITT: It does.

SUZ HINTON: If I’m ever
working for a large retailer

ever again, then I
feel like I got this.

MARTIN SPLITT: If you have a
blog, add the article markup.

You might get [INAUDIBLE]

SUZ HINTON: Oh, so OK.

I’m going to look at
the schema for that.

That would be like
author and stuff.

MARTIN SPLITT: And other sources
might pull the data as well,

right?

It’s an open source format.

So theoretically,
voice assistance

could use it as well.

So just imagine if
you have a recipe blog

and then you stand
in the kitchen,

go like, hey, assistant
thing– whatever

it is, whatever company
you’re choosing.

There’s a variety of
options these days, right?

And then the thing goes like
yeah, Martin’s apple pie.

First step– take some
apples and peel them.

And you’re like oh,
OK, fair enough.

That can come from the
structured data as well.

So that’s pretty cool.

SUZ HINTON: That is really cool.

I didn’t even think
of those use cases.

I just always thought
about search results.

[MUSIC PLAYING]