How To Define, Measure & Test Conversion Events
It is a truth universally acknowledged, that a single website in possession of good traffic must be in want of conversion events. As much as this is a dreadful hack of the famous opening line to Pride and Prejudice – it is sadly, a truth not universally acknowledged among site owners.
Even worse, large numbers of new websites fail because their owners have failed to clearly define what they want visitors to do once they land on their site.
At both a global and page level, improving the performance of your site requires the following steps:
- Defining a conversion event.
- Collecting data on this event.
- Making changes that can improve the relative frequency of this event.
While all three steps are key to producing profitable websites, site owners that clearly define their conversion aims while omitting steps 2 and 3 – measurement and testing – will at least provide themselves focus and direction as they develop their site.
However, embarking on a web venture without envisioning precisely what actions they want their visitors to take is an almost guaranteed recipe for disaster.
I am using the phrase “conversion event” rather than “conversion goal” deliberately and pointedly. Your website might have a single conversion goal, such as the purchase of an item on an ecommerce website, but there is likely a number of conversion events required at a page level to realize this conversion goal.
In the ecommerce example, this might include viewing a product detail page, adding an item to a shopping cart, filling in contact and shipping details and, finally, completing and submitting a page with credit card details.
A conversion goal might also be achieved by a user taking one of two or more actions that facilitate the realization of this goal.
For a site focused on lead generation, for example, the global conversion goal might be contacting a company representative.
However, this goal might be achieved by users taking one of several actions: opening a live chat session, submitting a contact form, or phoning a customer service representative. All three of these outcomes represent a successful goal being reached, but in each case, the actions taken by a user to reach the goal are quite different.
Having a crystal clear vision of your conversion goal is imperative to designing and improving a successful money-making website. But my reason for focusing on events here is that, even if it sometimes only a semantic difference, a goal may be conceived of abstractly, but an event by its nature is always related to an action, and is never passive.
By viewing your website through the lens of conversion events, you are forced to assess what the existence of any given page is meant to achieve.
Defining Conversion Events
Conversion events are first and foremost going to vary according to the nature of a website. For an ecommerce website, the conversion funnel (the sequence of conversion events to reach a goal) is relatively easy to define.
For other types of websites, however, conversion events might be more difficult to define; in fact, the more difficult it is to uncover these events, the more important it is to nail them down.
Take the example of a content-rich site that relies on display advertising for revenue, where the site owner is paid according to the number of banner impressions that are served. A conversion event on this website might be users clicking on a link that brings them to another page on the site. Knowing this helps guides page design, including such elements as effective links to related pages.
A secondary, or complementary, conversion event on such a site might be a user clicking on an advertiser’s banner, as a site with a proven record of high banner click-through rates can command a higher rate per thousand banners served.
Here, a conversion event-focused design might include placing banners in places where users are most likely to click on them, such as in the upper left hand corner of a page. I use this example because a very common response from webmasters (particularly on advertising driven sites) when asked about their conversion goal is “to get more traffic.”
Traffic itself is never really a conversion goal, let alone a conversion event, and a site owner who is focused on attracting more visitors without a very clear idea of what those visitors should do is squandering any traffic he or she has managed to amass.
Conversion is about getting more value from the traffic you have, rather than generating more traffic.
Measuring & Segmenting Conversion Events
A critical reason for clearly defining conversion events is that doing so facilitates putting metrics in place that allows you to measure your site’s success. If you are not recording actions that constitute a successful conversion event, how are you going to assess whether that visit is successful?
Here, especially, site owners need to think of what different conversion events need to be measured on different parts of the site.
For a site focused on generating leads, different mechanisms might be employed get users to provide their contact information. From the site’s blog, a user might be urged to download a free white paper; from the site’s corporate information pages, a user might be urged request a free consultation.
Even if this involves filling out an identical form, the conversion event that takes them there is different, and the two events need to measured separately. Improving the placement and call-to-action of the “free white paper” button on a blog post page will not impact the performance of the “free consultation” button on the about us page.
In Google Analytics, measuring conversion events by page type (setting up different goals for different types of pages) is easier if you are able to conceive of these page-specific conversion events when planning your website structure. This is because you can assign identifiers to URLs that make it easy to use match types when setting up funnels in Google Analytics.
For example, if the primary conversion event for an ecommerce category page is a user clicking through to a product detail page, tracking a user’s path through the funnel will be made a lot easier if all category pages contain a string in the URL that identifies them as a category page.
Why not simply cull data on page-specific conversion events from the complete purchase conversion funnel?
This might be possible if there is only one path a user can follow to complete a goal, but this is not always the case. A user might reach a product detail page through other paths besides a category page (such as onsite search), or even make a purchase without visiting a product page (such as by clicking on an “add to cart” button from another location).
By setting up product detail page views as a specific goal for category pages, you will then have the metrics available to make improvements to category pages in order to drive more page views to product detail pages.
Testing Conversion Events
The chief benefit of accurately measuring conversion events is that it doing so enables website owners to assess whether changes made to a site improve or depress conversion rates.
Website testing is, of course, a huge topic in its own right, so here I just want to stress again the importance of understanding your site’s conversion events as a jumping-off point for improving your site’s performance.
Whether you wish to conduct a complex multivariate test or simply try out two different button colors, your efforts will only be made possible if you know what visitor actions you are trying to encourage.
The flip side of this are pages for which no test can be conceived.
Are there pages on your site that have no conversion events associated with them, and so cannot be tested? If so, you may want to consider the time and effort you’re putting into creating and maintaining these pages, versus the reward you achieve from their existence.
While some of the points I’ve made may seem rudimentary, I’ve stumped experienced ecommerce webmasters by asking simple questions focused on desired visitor actions.
- What’s the conversion event for this page?
- How is the success of this page being measured?
- How might this page’s performance be improved?
These are questions I ask myself all the time, and there’s almost certainly profitable insights to be had if you ask these same questions while browsing your site.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.
Analytics news and expert advice every Thursday.