Optimisation Report – 16th Feb to 1st March 2017

I wanted to start writing about my journey to grow Epic Themes monthly sales figures, if you didn’t see the latest numbers they are over on the Epic Plugins Blog (transparency reports are grouped there, as that’s where they started out and will stay).

Here on the Epic Themes Blog I wanted to write about the ‘process’ and journey that I’m going through to increase those numbers. This will fall into a number of different ‘areas’ each with their own series and progression of blog posts.

New Product Development (such as the ThemePax Framework)
On-site optimisation (the focus of this report)
SEO and Inbound link generation
Paid traffic and acquisition channels

HotJar Funnel Analysis

These reports will focus on the HotJar funnel analysis (with info from heat maps) to see where my efforts should focus. First up is what I cal the full funnel. Towards the end of this post I’ll also share the ‘Shopper Behaviour’ from Google Analytics which confirms (and also shows some holes) in the HotJar way of analysing the funnels.

Full Funnel

The period shown in all the funnels is the 16th February to the 1st March 2017. This is what I call the ‘Full Funnel’ and it covers any entry to the site and tracks the flow through to a complete purchase via the specified required steps below.

The full funnel shows a large drop off at the following pages:-

Home Page (80.2% do not go to the Store Page) i.e. they enter the site and do not visit the store
Product Page (83.6% do not add to cart)
Checkout Page (75% did not complete the purchase)

I’ll delve into the potential reasons for these a little later, but first I wanted to share the other two funnels which I’ll be monitoring in these reports (and why)

Store to Purchase

The next funnel is the Store to purchase. Since the conversion rate above (0.2%) is for the required steps, above I thought it would be useful to also see the information for those who enter to the store and then proceed to checkout.

For example, in the full funnel I may be recording people as exits who visit the blog to read these posts, but do not purchase from the site every time they visit (which may go towards explaining the c80% drop off at the very first step).

Looking at the next funnel it is looking at those who have intent to purchase. They’ve visited the store (and not just casually reading the blog – because it’s so frickin epic 🙂 ).

It’s worth taking a look at the numbers above.. 89 sessions enter at the Store Page, vs 81 in the full funnel. I’m taking this to read that there’s 8 sessions where the visitor entered at the Store Page (the rest came to the store page from elsewhere on the site and captured in the first funnel).

What is nice in the above funnel is how a large proportion who visit the store, do indeed head through to the product page..

The general pattern is the same when looking at the above funnel though, when they actually make it through to the product page(s)

Product Page (84.7% do not add to cart)
Checkout Page (75% did not complete the purchase)

Finally, onto the last funnel that I’ll be monitoring, this is because I know I made more than 1 sale. So people must be entering another way.

Product to Purchase

This is slightly better. I’ll explain why below. The Product to Purchase funnel has more sessions. This is because people are more than likely finding my products demo sites (e.g. http://pluginhunt.com) and then clicking the buy now. This would bring them to the product page itself.

Also any referrals coming from external websites, are likely to be sending the traffic directly to the product page (rather than to the store page) such as a theme review, for example.

So this is now showing 7 completed purchases (rather than the 1 above). This is because we have a LOT of people enter at the Product Page.

201 sessions in this funnel (compared to the 59 who came from the Store first)
Still a large drop off at the Product Page (89.1%)
A relatively large drop off at the checkout page

This is really interesting. It shows me where I should spend my time next. The two areas that are seeing the most drop offs. The question to ask is why are people dropping off at that step.

The HotJar tool is great in that I can watch what’s happening at the drop off steps.

Why are people dropping off

I wanted to cover both the areas above in this report. The *home* page (which is any entry to the site, since it only has to begin with https://epicthemes.com) as well as the Product Page.

Home Page Drop Offs

Don’t ask me why I decided to share a little interlude to this post… but whenever I write about things like this, the following song springs into my head…

OK, I won’t share this video again, but sing it in your head each time I talk about ‘drop offs’… 🙂 or start it playing and read along with the music in the background 🙂

As I suspected, when analysing the home page funnel, it’s certainly counting everyone who visits blog posts and then leaves (after they’ve read the content).

Is this what I really want to monitor with a funnel, it makes sense to see how many can be converted from reading on the blog etc…

But is it really what I should be focussing on. Taking a quick look at the funnel, when the first step is EXACT home page match shows the picture below

This is a much fairer picture of the homepage and how it has the intent of filtering people down the funnel.

The same 87% drop off is showing here at the Product Page again

For any of the other pages (like the blog page) it’s not really clear what the goal is here. I think from the high ‘exits’ that these pages can be improved to send more traffic into the rest of the funnel. The following is the current markup of the blog and where I’ll look to improve this to send more to the site.

OK great. Those changes will be being pushed live today with the aim to convert more ‘casual’ visitors into sales from the blog pages.

Product Page Drop Offs

This one certainly is an interesting one to delve into. It’s clear from the review above that this is an area that needs more information to push people further through the funnel (if you’re one of these people, hi 🙂 please do feedback your thoughts).

Looking into the analysis which HotJar gives me for the Product Page drop offs there’s some very interesting insights.

drops offs due to visiting the demo (and then potentially coming back as a new session)
drop offs, after clicking the features, spending some time, (perhaps not seeing what they need to see)
drops offs, reached bottom of page (not buy now button so they scroll around back to the top and leave)
Need help pop-out pissing people off (causing them to exit after closing the slide in)

For the drop offs at the product page due to visiting the demo is something which has been hard to see what’s going on in HotJar. I have eCommerce tracking also setup under Google Analytics. For the same period I get the following information.

The difference (16 conversions vs 7 in HotJar) is due to the back-filling. HotJar counts the drop off, however the person may return at a later step in the funnel to complete the purchase.

This would be a tick in Google Analytics (and it would fill in the earlier steps) but a drop off and a 0% conversion with HotJar.

This can be seen in the following comparisons (for the same time period, 16th Feb to the 1st March)

HotJar – 409 sessions, 0.2% conversion rate
Google Analytics – 597 sessions, 2.7% conversion rate

Both funnels have their uses, however the number which I’ll be measuring is the Google Analytics metric (and using HotJar to help drive the ‘why’ it’s happening).

So, Google Analytics is also confirming the following:-

There’s still a high proportion of sessions with Product Views but without Adding to Cart (87% of sessions that viewed a product didn’t add to cart)

So the improvements that I’m looking at making to the product pages are:-

Make a section at the bottom of each product with details a money back guarantee
Add a contact form to the bottom of the features list (to ask us about any “missing” features)
Make sure all products have a buy now button at the bottom of the product description
Turned off the ‘Need Help’ slide out which was interrupting visitors when they were on the product page

Once these changes go live (today), I’ll then be revisiting the above analysis in the middle of March to see if it has helped at all with the conversion rates and whether there’s any further insights that I need to adapt the individual product pages for.

Don’t forget, if you want to follow along with these reports, sign up to the newsletter using the form below. I’m sure they’ll come in useful if you’re looking to optimise your own businesses conversion funnels.