I don’t blog too frequently, but when I do it’s usually about technical SEO or web development; this time it’s a little different.
Over the past few years, digital marketing has become more data-driven than ever before. Here at Impression, we love having data to analyse so we can make better informed decisions that produce much stronger ROIs.
But with this increased reliance on data comes a requirement that that data be reliable and complete. We’re increasingly faced with new clients who need to improve their on-site Analytics and over digital campaign tracking, and we enjoy getting to the crux of the problem to reveal what’s really going on. That said, technology has moved on so much that there really isn’t a need for any marketer or website owners to be operating without clear and complete data.Too often we see marketers or website owners operating in the dark, and for no good reason.
The issue
The main issue with tracking that we see is inconsistency in the user journey. Of course, this is far more than just a tracking issue.
What I’m talking about is examples such as one we saw with an ecommerce store recently; the buyer clicks ‘buy’ on that domain, and is then sent to an off-site cart to continue their journey.
Because of this, is it very difficult, if not impossible, to track what the user does after they leave your primary domain. For the ecommerce store in question, that means they have absolutely no visibility on whether that person went on to buy, or if they dropped out of the sales funnel.
But this problem isn’t exclusively an ecommerce one. Lead generation websites can sometimes fall foul of similar mistakes;
- Forms hosted on subdomains or elsewhere
- Over-utilisation of third party applications use for data capture
- Probably, some other stuff I haven’t even ever seen before
- Etcetera.
If there’s a point in your customer journey where they leave your domain, it can be a huge issue.
Which leads to
This setup often leads to a number of issues, but they can always be summarised as a ‘complete nightmare’ scenario. Typically, you’d find your website Analytics data aren’t making sense, you see huge drop-offs and large swathes of mystery traffic you’re not sure even existed.
As a result you being to lose confidence in your upwards reporting and ability to generate a good return from your marketing channels.
When you lack clear and complete data, you miss out on opportunities to learn about your users. If the ecommerce store I mentioned did have complete tracking information, they’d be able to improve the sales funnel to increase conversion rates. They’d also be armed with hugely powerful data about cart drop outs, which could feed a very successful retargeting campaign.
It’s not just bad for you
This doesn’t just make for a confusing day buried in Analytics and Excel for you, but this is also a terrible user experience (UX) for your customers. And that’s where I’m going with this:
It doesn’t have to be this way any more!
The modern landscape
With the huge range in open source ecommerce software available to anyone seriously looking into a new venture, it’s difficult to learn of clients, friends and other (typically new) webmasters buying into anything but a well-integrated software solution with great UX to boot.
And with so much data to produce and harvest, it’s never been so important to own the pages where your prospects convert.
1. Software and platforms
There are so many platforms available to ecommerce store owners these days. There’s really no excuse for using a platform which requires customers to checkout on a separate domain or service.
Even if you’re on a tighter budget, you can find a great platform to power your shop. For example, for only a few hundred dollars you can install a WooCommerce (WordPress-powered) ecommerce store, install a theme, customise it to your liking and install a payment gateway. If you get some professional help, which even these days isn’t always required, then you can make your site secure with an SSL certificate too.
I won’t go into the detail of that here as it’s wide open to a lot of conjecture, especially around security, but you get the point; a little now goes a long way so long as you’re looking for a mainly out of the box solution.
If you have a larger budget and want something that’s suitable for a larger shop with a high volume of products, platforms such as Magento are brilliant. They offer the functionality you need alongside a level of customisation that means you can get exactly what you want in a way that works well for everyone.
2. Gateways and payment processing providers
In the UK, SagePay, PayPal and Stripe are some of the most popular payment platforms.
Ebay corporation has previously been an owner of Magento and PayPal and as a result their integration is very slick. And for most other gateway-software combinations there’s a module or plugin available between £0 and £100 that makes the integration almost seamless.
Another benefit of these types of payment gateways is that they are open source, meaning that any bugs or requests for improvement usually get sorted very quickly, making for a highly robust platform – which is really important for an ecommerce store.
This is all possible due to the advancements in the APIs (application programming interfaces; communication pipes allowing or secure data transfer, basically) in these services and the more open nature of the businesses powering them to allow just about anyone to connect successfully through their APIs.
Again, there’s no excuse for webmasters to host their payments on a different domain, when integration of these high quality gateways is so straightforward.
3. Tracking
Whether the technical implementation itself is new or not, old approaches to ecommerce often lead to the same old tracking issues. For example, we’ve come across some gateways which still force users to only use the now-legacy ga.js Classic Google Analytics.
Without getting into too much detail here, the newer “Universal Analytics” features are much better, require less code maintenance and also allow for extended functionality, such as Enhanced Ecommerce. Check out Google’s blog on the release here to understand why you’d want this data at your fingertips!
In addition, Google has in the last few years, released a product called ‘Google Tag Manager’ — a nice user interface allowing non-developers to insert tracking tags conditionally into specific (or all) pages. With these tags, you can send any piece of data that you can gather from a site to any third party application that’s ready to accept it.
Without a huge knowledge of code, you can now integrate your store with basket abandonment apps, call tracking solutions and social advertising platforms. But to do all of this, you can’t rely on third party hosted solutions — you need to own (read: manage) the server where this sits in order to edit the underlying codebase in the first place to make all of the above possible.
If you don’t already have this in place…
If you don’t already have this in place, then I’d really be starting to ask questions around the importance of being able to understand your site’s usage. If you don’t have benchmarks set in believable data, then how can you improve?
If you are in this situation, I’ve come up with three steps to help you move forward; it’s not going to be easy but the end result is extremely rewarding.
Step 1: Analyse what you have
Even with issues, you will still have something. Check for the following to ensure your Analytics is up to scratch. This will help you identify the issues you can then check off one by one;
Check for cart/gateway referral issues — if you’re seeing referral traffic from these pages, then you can be sure that your tracking is either not configured, or configured poorly. As well as being a nuisance, this actually has a huge impact on your being able to follow your customers’ journeys around your site, on a per campaign source/medium basis.
Identify any self-referring issues — similarly, self-referring issues describe a scenario when your own website or other properties are marked as referrers within your Analytics. In an ideal world, these individual sessions should be stitched together to give you a holistic view on your marketing efforts. This is usually caused by a single page or page template not being properly fitted out with tracking code.
Exclude all irrelevant IP addresses — the IP addresses of your office and your developers will likely lead to a number of false purchase data which doesn’t tie up with your bank. If you’re testing on the same Google Analytics property, consider setting up another for configuration and gateway testing so that your numbers aren’t skewed. In the long run, this approach will also ensure you session and user data isn’t inflated by your own visits.
Check your ecommerce and goal reports — sometimes “creative” agencies don’t consider your ongoing business needs, such as revenue tracking through anything other than the platform itself. Check that you have ecommerce reporting enabled so that you can track your platform’s output, and also consider setting up conversion goals for actions within your website, such as newsletter signups. This data will begin to form a benchmark for moving forward.
Step 2: Put in place any missing tracking
Once you’ve identified your issues, then put in place a plan to resolve these. Some may just simply be misconfigurations in the Analytics interfaces, however many sites which span multiple domains or services will require an element of development work to patch them up.
This is a really good guide from Google on measuring campaigns and traffic sources.
Step 3: Set in place a plan to forward integrate your cart, checkout and aftercare process
It’s certainly my opinion that you should have in place a long-term plan to completely forward-integrate your outsourced services so that you begin to own more and more of your customer journey, and the data and tracking benefits which come with it.
If your cart is currently located off-site, then it’s likely you aren’t yet running an ecommerce platform. For you, a migration to a cable platform should be high on your list. So long as you don’t have any complex requirements, an open source ecommerce platform will suit your needs perfectly.
Following this, or if you already have your shopping cart functionality, but not an on-site checkout (for example, you use PayPal’s off-site checkout page), then the next step in the puzzle is to bring this onto your own website.
There is an important factor to consider here — web security; you should speak with a developer or your web host and seek help in installing a security certificate. This gives you a green bad in your browser and peace of mind that your user’s connections with the site are encrypted.
You then have a decision to make around the level of payment gateway integration; through an iframe or through your server. For simplicity I’ll skirt around the most integrated option — typically called a server integration as you then need to seriously think about your card detail handling protocols.
A ‘direct’ integration — through an iframe — allows you to retain almost all of the shopping experience on your website with just the ultra-important card handling handled for you by the experts. Once this step is completed, your users are then typically forwarded on to your on-site thank you page, along with the transaction data that your Analytics package needs to populate a successful transaction.
Click here to see how SagePay allow differing levels of integration.
Finally, let me know how you get on!
If this guide has helped you put in place an improvement strategy, then please let me know in the comments below.
We do also offer advanced Analytics consultancy, so if you think you may benefit from a bespoke audit and migration/implementation, then please get in touch.