Ohh yeah that’s right… we got Vanilla Ice today.

The last time I heard anything about Vanilla Ice was last year when I saw a billboard off Hwy 217 for a concert he was doing at a casino in the middle of nowhere between Portland and Salem. That billboard usually advertises their all-you-can-eat seafood buffet, but I guess sometimes they have soft serve on the menu too…

Anyway…

Stop.

The ability to stop and process a thought in our head is increasingly becoming a skill to be practiced. AI has given us too many automatic, magic tools that make us feel productive even when we don’t produce anything.

But you have to stop moving so you can take inventory of the data you’ve collected before the next undertaking.

And our stopping point for now is with the Snapshots.

- We did a storewide audit with the first Store Snapshot.
- We ran a page snapshot on a collection page.
- We ran 2 page snapshots of top product pages.

Mouse Whisperer got us to this point…

And now we can move onto the next step in the CRO workflow…

Corroborate.

Analytics tool are not designed to give you everything. Many apps will sell you on that idea, but they know its not true either. My goal with Mouse Whisperer is to give you what you need to gain direction & measurement for your CRO initiatives… which on its own is a lot, but it would behoove us to not discount the value of qualitative data — that juicy feedback and anecdotal evidence you get straight from the customers.

- Store reviews
- Amazon reviews
- Store / Amazon reviews of competitors
- Customer service emails (Gorgias)
- Customer surveys and/or interviews

Fleshing out the quant data with qual data will spark more ideas, help you write better messaging, solidify your positioning, and all in all help you understand your customers deeper.

So the goal here is to corroborate the findings in your data with customer feedback. And vice versa.

Listen.

I will listen to my own and vice and STOP.

Yesterday, today, and tomorrow, I will be doing my own research. Not like a tin-foil mad hatter uncle who “did his own research” on facebook, but actual research and analysis or CRO.

Then we will come up with some hypotheses.

Then we will build some stuff.

Then we will test the stuff.

Then we will do it again.

One more parting thought…

Feasible rhymes that you can vision and feel
Conducted and formed, this is a hell of a concept*

*actual lyrics from ice ice baby, no idea what they mean, but can’t argue with it…

Until Friday!

- Shep

We are only through 5 of the 30 Days of CRO, and we gotta keep moving on so let’s get back to it…

On Friday, we left off by starting a few page snapshots. If you recall, it was for 2 top product pages and 1 collection page, and I want to start with the collection page first.

The Collection Snapshot

Most traffic is going to land on your Homepage, a collection page, or a product page. The last one - product pages - are closer to checkout, and therefore, more important for CRO than other pages (typically).

The collection page, however, shows you more about your customers shopping habits and intent. It shows you:

a) what they are looking for
b) how they are looking for it.

We will look at the results page for a MW collection snapshot in a minute, but first, let’s hit the peace pipe.

Benchmark

That image above is what you see when you click on the Collections tab in MW. And right off the bat, you’ll see the high level overview

- 500 human visitors
- 283 clicked to view a product (59%)
- 3 people used the search bar

These will be our own benchmarks for subsequent snapshots.

Snapshot Results

What I like about these snapshots the most is that it is Sequential Testing. It is not an AB test because there is no B. It’s just your A game…

AB Testing is great when you have a variable to test, but you might not have a variable to test because you might not know what people are doing on or clicking on that page.

Or you might have just launched. Or traffic is too low to test but you still want to know what’s working or not.

These snapshots results in a bunch of fun signals for you to analyze.

This video walks through all the data that is collected. It aims to be a comprehensive look at your on-page engagement based on data (signals) that you can actually act upon… depending on what you are trying to optimize

Snapshot Results

1. Traffic Quality (Real Users vs. Zombies vs. Bots)

2. Traffic Source Segmentation

3. Engagement Funnel

4. Link Clicks

5. On-Site Searches & Top Queries

6. Exit Rate & Exit Paths

7. Product CTR & Top Products Clicked

8. Average Time on Page

9. Average Scroll Depth

10. Top Countries & Cities

11. Device Breakdown

12. Filter Usage

13. Engagement by Zone (Body Click Map)

14. Header & Footer Click Maps

15. Recent Visits (Session-Level Journeys)

Woooooahh that is a lot. Just kidding. When you run this, you will glean insights and ideas from several metrics, and some others will just be good to keep an eye on as changes are implemented.

Here are a few ideas to go with each type.

CRO Ideas for Each

1. Traffic Quality (Real vs. Zombies vs. Bots)

This shows the conversion rate using only real users for accurate benchmarks. If you have a high number of zombies, that’s a problem to work on, but that is a separate line of optimizations that may lie outside the page.

2. Traffic Source Segmentation

Compare product CTR by marketing channel; if one channel is underperforming compared to the others, there is an issue to dig into.

3. Engagement Funnel

Find where the big drop-off is and focus on that step. Then fan out.

4. Link Clicks

If you get low link clicks, your page might be a dead-end. You might need internal links or other calls-to-action. Or you might remove links that get ignored so they dont confuse shoppers.

5. On-Site Searches & Top Queries

Searches are gold. They show what people want but can not find, and a really high search volume means the merchandising sucks.

6. Exit Rate & Exit Paths

Closed Tab exits are double-plus ungood. So if that is high, test around that first, then work your way down. Except for idle sessions… someone may have spilled coffee on their lap while shopping.

7. Product CTR & Top Products Clicked

Feature your most-clicked product above the fold, and test new product photos for products with low clicks despite good placement.

8. Average Time on Page

Short time with a low CTR means you got no hooks; short time with high CTR is is good and you should move on.

9. Average Scroll Depth

Good to know, and good to keep an eye on. Because sometimes you might want to move sections high up if your scroll depth has no depth.

10. Top Countries & Cities

Ashburn or Counsil Bluff are usually crawler traffic… the rest is good if you sell internationally.

11. Device Breakdown

This is obvious

12. Filter Usage

Filters are great IF they are needed. Not all stores need them. But if you have a huge catalog, this is a huge component to keep an eye on.

13. Engagement by Zone (Body Click Map)

This is #13 on this list but its #1 or #2 on my list… because you want to know where the action is taking place… so you can be there.

14. Header & Footer Click Maps

Separated into their ow boxes for quick reference.

15. Recent Visits (Session-Level Journeys)

These are there for when you need to get granular. One of the least insightful areas because it is so anecdotal, like watching session recordings, but nevertheless handy when you need it.

16. Snapshot Comparison

Run subsequent snapshots to measure changes and see if your optimizations actually worked!!

That is all for now!

- Shep

Just some work…. not work work. I am not doing regular week work on the weekend. Usually I will work on something fun like writing a comic, but since we are mousing around for 30 days, I think some housecleaning would be good…

So we are going to head to the Shopify Admin > Settings > Apps

To quote the little kiddo, “That could be fun…”

Laying bare a screenshot of your apps feels so personal. I know you are going to judge me. But whatever, let’s see what we got. And to be clear, this little project is something that I have needed to do for a while. I recently removed an expensive app (it was made redundent by Mouse Whisperer), and thought that I should take inventory and just see what I can cut.

Withouth Further ado…

That is 31 apps… lets take inventory

Shopify Apps (Native apps made by Shopify)

1. Forms

2. Messaging

3. Search & Discovery

4. Flow

5. Collabs

6. Simgym

7. Knowledge Base

8. Shopify Graph QL

9. Order Printer

10. Bundles

Actually used Shopify Apps

1. Forms — good for simple pops and forms; limited but solid

2. Messaging — email marketing sans the predatory pricing; good editor, clunky automation. Start app, not for someone who’s job title is “email marketer”

3. Search & Discovery — Necessary for all stores; controls filters and recommended/complimentary products

4. Flow — Necessary; I should use this more. It’s powerful and these guys get into the nitty gritty on the Unofficial Shopify Podcast.

5. Knowledge Base — Set up because everyone was barking about AI and I thought ok why not, let’s add words there and…. then lost interest (if anyone else knows more about this app or wtf I need it for… please, share it with the class!)

6. Shopify Graph QL — if you like your data raw…

Unused Shopify Apps

1. Collabs — it is like all influencer networking apps: solid tech, shitty shitty shitty product. I effing hate influencers and influencer marketing.

2. Simgym — Stupidest app that Shopify has ever made. Almost comical to watch it throw up insights based on synthetic users.

3. Order Printer — We stopped using it in favor of the shipstation templates.

4. Bundles — We are about to add bundles on the store, so this one might move to the list above in a couple months.

Vibe-coded Shopify Apps

1. Alt Texter

2. Media Files Editor

It was back in February when I made these. There was another one I made too but it was deleted. I don’t remember what it was for though.

But I clearly remember why I made them.

We were under attack

When this feature in their ai Sidekick was released, I think we all tried it out. And if you did try it, I would love to hear about it.. comment or email me

So I would have tried it out anyway, but the reason that these 2 apps were made was because of one of those frivolous ADA lawsuits.

If you haven’t been through one, here is the lowdown.

Shitty Lawyers

Apparently the firm behind out suit had filed +1000 lawsuits. And their tactic was simple. Automate the filings, scare you, get you to settle. They were looking for a $7500 settlement.

That number is not so scary. I mean, it would completely fuck over a small Shopify store still struggling to get off the ground, but they don’t go after those sties. They wait until you are bringing in a certain amount of revenue or traffic (I don’t know what they track exactly, this is my assumption), and then slap you with one.

ADA Compliance

Trying to make your website accessible to everyone is a good thing. The things we need to do better as website developers for accessibility is not very visible.

Alt Text on images helps screen readers and their users to understand your products better. That is great and easy to do, in theory, but in practice, it can be a huge undertaking. Especially if you didnt take the images yourself… or if you inherited a store, and no one added any Alt Text, you gotta go one by one?

Matrixfy

Best and most used app for bulk editing. You export what you need into a spreadsheet, and then you can edit them easily. Import the file back into Shopify via Matrixify again and it syncs up the changes. Doneskies.

Great for migrations and saving your ass.

Media Cleanup

This was an app I found to help with the Alt Text. It helped me get through a lot of files.

I only used it for that month though before downgrading. Well worth the $14.99. Some apps are for projects and not for an on-going need. I wish there were a way for apps like that to have a separate billing type so you dont have to subscribe and then uninstall or downgrade.

So yeah, we were supposed to be talking about the vibe-coded Alt Texter and Media File Editor apps that shopify’s Sidekick shittily made.

I tried to open them to get a screenshot, but its saying I don’t have permissions, even thought I made them…?

Whatever. Shopify Sidekick has never NOT let me down. So not surprised.

This is not over!!!

I have more apps to get to, but my coffee is cold, and little duders just got up. So let’s pick this back up another day… I gotta jet…

Until tomorrow

- shep

Subscribe now

According to the story, in 1914, an architect named Joseph N. Bradford used a hot air balloon to fly up and over the trails that had been created in the snow by students during the winter. They then used that to build out a network of paved sidewalks.

Desire Paths

These criss-crossy sidewalks are the shortest path from a to b…

Wikipedia — A desire path is an unplanned small trail formed by erosion caused by human or animal traffic. The path usually represents the shortest or the most easily navigated route through a natural area or greenspace, and the width and severity of its surface erosion are often indicators of the quantity of traffic it receives.

Desire paths may go on to become formal trails and even established thoroughfares, such as Broadway in New York City, which some urban planners believe to follow a convenient trail developed by the Wecquaesgeek people, predating American colonization.

(I heard this story years ago, but it wasn’t until today that I heard that NYC’s Broadway was a result of it.)

Desire paths are not only about trampled vegetation. The UX design community uses this example a lot to help new designers wrap their head around the fact that not everyone uses software in the way it was built/intended.

So it naturally came to mind when I started this yesterday. Because of two things:

1. I wanted to explore using Mouse Whisperer as a user instead of a builder, and that is how desire paths are revealed.

2. Part of Mouse Whisperer is for evaluating the desire paths on your ecommerce store. It’s actually the first step in the CRO workflow…

I didn’t even make that correlation when building that part of the workflow, so yeah, worked out well.

Let’s look at that now…

Store Snapshots

When you install Mouse Whisperer, you see that we call audits “snapshots.” We use that name because we don’t look back at past data to drum up insights. There is nothing wrong with that, but in my experience over the last 12 years (and esp in the last 1 or 2 years), websites change too fast. So instead of pulling data, we let our mousetracker.js see what your visitors are doing over a short period of time.

Because not only websites change… the market and your competition change too. So this gives you data that is more relevant to right now.

And here it is… a Store Snapshot in progress. I set this one to track 2,500 unique human visitors. Because I want to know what real people are doing on the site.

I’m trying to optimize the site, so I only are about the hooooomans. The Zombies you see in that image are people who came to the site but did not engage with anything or click anything. They are not customers in this story.

Some could be accidental clicks. Some could be meta-allowed click fraud. Some could be aliens. I don’t care. They weren’t shopping, and our task at hand is CRO not traffic quality. That’s important, but it’s for a different day.

Anyway, we got over 1600 so far. That is plenty to work with. so let’s get to the good stuff…

Alrighty… we got some numbers.

• CVR = 4.5%. Not bad

• ATC (add to cart) = 10.6%. That is definitely something. A drop from 10.6 ATC to 4.5 CVR. Work is needed here.

• Searches = 21. I need to dig into what those keywords those are. That could be show us something.

These don’t really tell us what to do or where to do it, but they are important to gauge the health of your store and set benchmarks.

This image below, though… we are starting to get to the good shit.

I have it set to show in order of highest to lowest sessions.

The top page by both session count and revenue is the homepage. In general, I would suggest that you focus your CRO efforts on just about all other pages before you start tweaking your homepage.

The next 5 pages on that list better us our purposes. 3 product pages and 2 collections. I like to keep it tight so I don’t get scatter-brained, so I am not going to work on 5 at a time. But I also like variety and ideas tend to bounce around more when I’m juggling more than one at time.

I will go with 3.

And because I am just looking at some high-level numbers for each, I need to know more. So let me go and make 4 clicks to move onto step 3.

I can choose these 3 with the checkboxes, and then click Create Snapshot - which will prompt me for a few settings.

I started the Store Snapshot on Thursday morning early. It’s Friday now. I want this to be done by Monday morning so I can continue on, so I set the target visitor count to 500 for each.

You can also just start a Snapshot for any page at any time by going to that section of the app, but as part of the workflow, we build this into the Store Snapshot phase because… why not. It’s smooth.

So now, this new section appears to let me know that these Snapshot audits are up and running.

That is all for today. On Monday we will dive into the page snapshots… which were the first things we designed in this app. And kind of the whole point.

So that is exciting, but not nearly as exciting as tonight, because I get to go see Blind Pilot… if you haven’t heard of them, here is my favorite song.

Have a great weekend.

- shep

Subscribe now

Anyway.

I am challenging myself and my Shopify app, Mouse Whisperer. We just put the final touches on the app last week, and it is time for a rigorous test.

So for the next 30 days, I will use MW, which is for CRO & AB testing, to track how we use it and the results we get.

This series of posts are under a separate section than my usual posts because the theme is conversion rate optimization, and that is different enough from the usual writing.

So if this one isn’t you thing, just unsubscribe to this “section” or whatever they call it here, and then you can still get the others.

But if you do enjoy CRO and follow along, I plan on posting each week day.

To kick things off, lets set the baseline with yesterday’s metrics:

We will look at lots of data and their fluctuations over the next month, and tomorrow I will show you what we gathered from the first step in this journey with our mouse friend

And also tomorrow, I will talk about why this photo below sparked the idea for this challenge. If you know what this photo is, put it in the comments.

Until tomorrow!!

- Shep

Subscribe now

I especially had fun creating the pixel zombie on this one.

And I don’t know why, but this zombie is wearing a haggard gray suit. It’s probably because movies depict zombies wearing nice clothes all torn up. Like they just left the office and going home to eat dinner and/or your face. I don’t know why they are never wearing a band shirt or that shirt with Che Guevara on it. It would be cool if one was wearing a “Kiss me I gave blood” shirts you used to get at a blood drive.

Anyway....

Zombies are a interesting class on traffic in my books. I define them as visitors who come to your site, never interact or engage or click anything, and then leave. They are not a customer nor were they ever planning to be... or maybe they could be a customer but that session was a fluke because they clicked on your ad, and in a fit of absolutely excitement over finding your store, spilled their coffee on their lap, causing them the throw their arms up, sending their phone flying out the window of their car, going straight through a metal grate into the gutter.

In those scenarios, they will come back later. Maybe. We cannot control the wind or human spasms. Regardless, it does not matter. No reason to count that in your analytics.

Conversion Rate typically is counted by Orders and Session. This is oversimplified on purpose. When you are a platform (like shopify), you cannot afford to have a confusing analytics dashboard. That would be bad for their biz. So lame, standard metrics that work for all stores are what we get.

Anyway, here is the basic formula -- CVR = (Total Orders / Total Sessions) x 100

Easy peasy...

Ecommerce stores on average hover around 3% CVR with this. And it works. Kinda. For the most part. And good for reports. And covering your ass... aka reporting. So now I’m being redundant…

On the store I currently manage, when you strip out the bots and zombie traffic, our hero product is over 15%. And our 2nd most popular product is around 12% last I checked. Some products are super duper low.

And that brings us to the point..

If we discount non-customer traffic, we can actually see which pages are performing well and which need work. We can study what the real customers are clicking on, where they go after the product page, or which products they click on a collection.

Remove the noise and watch what the humans are doing on your website. That is how you get a sense of what your conversion rate optimization plan will look like. Less noise. Easier to identify roadblocks and missed opportunities. Easier to spell our actual opportunities. And less time jumping around trying to fix everything because your analytics says you are 2.2% today instead of 2.5%.

Analytics that tell you where to focus your attention are way more useful that analytics that just tell you what to report.

Analytics for the sake of analytics is for data scientists... analytics for ecommerce operators needs to be directional and informing your next moves.

If you want these types of analytics, get Mouse Whisperer: CRO Analytics & AB Testing for Shopify Stores. It’s free (for now). Link to Shopify App Store

If you want me to pixelate you as one of The Real Heroes of Ecommerce, reply with a photo and your superhero name.