AI Bulk Category Assignment
Categories are super important inside of
the discoverability of your app and it's
the key that everything feeds off of.
For example, meta fields get assigned to
categories. So, it's important to assign
categories for all of your products. And
here you can see in this example store
I'm looking at, we've only got two of 11
products with a category assigned. So,
we need to finish out the rest of those
categories being assigned. Now, we can
go in to the products and manually start
assigning the categories. For example, I
think this category might be women's
tops. So, let's just see if we can find
a women's tops category. Type it in.
There's women's underpants.
Maybe it's not called tops
under shirts. Maybe it's not.
See how you're trying to find the right
category and it can be hard in this
because there's 10,000 nodes in here
that it's looking for. Um, and if you
don't get the right words, it's not
going to find the right one for you. So
maybe it's apparel
and then top.
So it becomes frustrating to find the
right categories. And sometimes Shopify
and their AI tool called Magic will
recommend a category for you. But as you
can see here, sometimes it doesn't.
Sometimes that category is accurate and
sometimes it's not. And when you go
tohead to accept their recommendations,
because many times they are accurate.
When you go ahead to accept it, you have
to go in one at a time and say accept.
There's no bulk accept in the Shopify
admin for their category
recommendations. So what we've built
into datafy is the ability to get in
bulk category recommendations and then
you can assign the ones that you approve
of in bulk also. Now on a small catalog
like this where I've got, you know, a
dozen or so products that need the
categories assigned, I could do this
manually if I wanted to, but it is a
bunch of work, right? And on a catalog
of hundreds or thousands of products,
there's no way you can do it manually.
So, we're going to go in the datafy app
to the bulk assign categories section.
And the first thing you have to do is
you have to subscribe to AI
categorization. This is a paid feature.
We do have to analyze all your products,
use our AI models to determine our
category recommendations because Shopify
does not share their category
recommendations to any apps in the
ecosystem. So, there there's a bunch of
time and energy involved in us doing
this. So, it is a paid feature. So all
you have to do is subscribe to AI
categorization. Now this isn't a monthly
fee. It's a pay as you go model. So if
you only categorize your products once,
you make one payment. But if you bring
in new products and want to recategorize
them, you'll make a payment for that
amount. And that amount changes based on
the number of products you're applying a
category for. So here I'm going to
subscribe to AI categorization.
And you can see, you know, here it's
telling you, oh, there's a monthly
charge. won't exceed $25.
I'm going to approve that.
And now I have the ability inside of
bulk assign categories
of
looking at them. So here what I'm going
to do is I'm going to run AI
categorization for all uncatategorized
products of which there's nine of in
this store. So I'm going to run that
and it's going to look at each product
and the data about that product. the
product title, the product description,
and other attributes about the product,
and compare that to the Shopify category
list, that 10,000 list of categories,
and determine the best one for that
product. I'm going to pause this while
this, it's already run. Now, it's
telling us that
these products here, this is the the
product name that we put in, matches to
this suggested category. And here's the
confidence that the AI model has for
that. And you can see all of them are
high except for the frozen asah of ice
cream and frozen yogurt. It's a medium
confidence. And you can filter by
confidence level. So there's no low
confidence products. There's one medium
confidence of the category and then the
high. Usually we're fairly confident
that most of or all of the highs are
accurate. We haven't come across a case
yet that they're not. the mediums half
the time they're accurate and the lows
you definitely need to look at. Now in
this case here um it actually just
updated the oh we're on the highs and so
here and now I can you know I can select
all of these and just you know bulk
accept them. So, I'm going to go in and
bulk accept all. Or if I wanted to, I
can say, you know, take all the highs
and select the ones I want um for them.
But I'm just going to bulk accept all in
this case.
Now, like I said, for, you know, when
I'm doing nine products, you could do
this by hand. It is a little much
quicker and easier to do this through
the automated automation tools here. But
when you have hundreds and thousands of
products, it is indispensable to use
this functionality. Now, if I go back to
my product list,
you'll see that they all have oh,
activewear tops. That was that category
I was looking for before. Remember the
problem I was having trying to find
women's tops? Because it's actually
active wear tops. Um, and you know, it
would have taken me a while to figure
out that out in the taxonomy, but the AI
tools and datafy figured that out for
me. So, I don't have to spend my time
understanding all of that. So, you can
see here all the categories are now
assigned to those products. And if I go
back to the top level of datafy
and rerun my uh discoverability score,
right now it's a 5.6.
I'm going to recalculate that.
And we went from a 5.6 to a 6.7 because
now all of our products have a category
assigned. So that's how you do bulk
assigning of categories using AI to
determine the right category for your
products. Thanks for watching.