The conversion funnel is a tool used by ecommerce professionals to help optimise the usability of their online assets. This is achieved by monitoring and analysing the path to conversion through a set of core stages to see if there are any roadblocks to completing a transaction. When I say online assets, I am typically referring to any standard website, mobile site, responsive design site or anything else that has multiple unique steps involved in the path to transaction.
The ecommerce conversion funnel monitors all stages on the path to a completed transaction. It doesn’t have to be a transaction per se, it can be any goal which you wish your visitors to complete, like signing up to your blog, which if you are using inbound marketing as a strategy, could be of great importance.
However, I digress! Traditionally, the most common use is focused on a completed transaction as the end goal or conversion. This is defined as when the consumer has been through the checkout, entered their billing details and paid for their basket of goods.
Typical conversion funnel stages
The stages described below outline a typical customer journey to a transaction after landing on the homepage. The conversion funnel stages may differ slightly for certain sites, but the following are optimal for nearly all sites:
Visited the site - someone visited the site.
Viewed listing - someone viewed a product listing, category or sub-category page.
Viewed product - someone viewed a product page.
Added product to basket - someone added at least one product to basket.
Started checkout - someone began the checkout process
Completed checkout - somone performed a transaction and completed the checkout process.
Applying the above stages to your historical data will bring to light some very interesting insights for one of the most important customer journeys on your site. You will be able to quantify the reduction in volume of visitors at each stage of the funnel up until conversion. Using this information you can now see if there is any abnormal drop off at a particular stage of the funnel. For example, if there is an abnormally large exit at the ‘viewed product’ stage, you can use this information to suggest a number of hypotheses about why the product page is not performing well, like:
- Your images don’t highlight the quality of your products.
- Your descriptions don’t resonate with your audience.
- People don’t add to cart because your social proof isn't high.
Following this you can put forward suggestions on how to resolve these problems through comprehensive testing:
- Create Multiple Hi Def Photos that highlight items.
- Create product descriptions that use language more relevant to your target market.
- Incorporate testimonials and customer reviews into your product pages.
For further inspiration and advice on product page improvements, see our ebook: The Ultimate Product Page - A Guide to High Converting Product Pages
Now you may be asking "how do I know if a certain stage of the funnel is performing badly?" and "How do I determine what is good or bad?" The answer is benchmarking - this information is typically quite hard to get ahold of however I'm sharing with you the figures I use based onthe numerous companies I have worked with in the ecommerce industry.
The figures below are pass-through percentages from one stage to the next for all ecommerce visits. An ecommerce visit encompasses any visit which views a category or product page.
Viewed Category - 100%
Viewed Product - 80%
Added Product to Basket - 20%
Started Checkout - 60%
Completed Checkout - 80%
Remember these are generalised figures and will inevitably vary based on specifics such as business model and industry. However what they do provide is a comparative guideline to quickly find the best opportunities for improvement.
Once you have resolved the obvious problem areas you should strive to achieve your best possible levels of conversion for each stage of the funnel. So if you're hitting these benchmarks don’t relax just yet - do more!
Continue working towards improving your stages through testing different hypotheses. Use a monthly or weekly comparison using your ecommerce analytics platform to determine your ongoing performance. It will be useful to set certain achievable goals for you and your team to work towards. When comparing each funnel stage and setting goals it is worth looking at the % of conversion based on the total people that complete the previous stage of the funnel. We call this the pass through percentage, which is in contrast to the absolute value for each stage. This is very useful because the quantity of visitors over two different discrete time periods is almost never the same.
Another useful tip before we move onto the next section is to always focus on improving problems at the bottom end of the funnel first. Resolving issues here can have a prominent effect on revenue very quickly. Relatively simple changes such as adding guest checkout can have a greatly beneficial effect on your conversion rate.
The conversion funnel described above is very good for an aggregate overview of all your visitors but it's not enought to only monitoring this funnel. The next step is funnel segmentation and it should be applied to your conversion funnels in order to get more granular.
Common segmentations you may use include:
- Only mobile users
- Only International Visitors
- Only Visits from a specific marketing channel
Mobile is a particularly interesting one; the recent explosion of smartphones and tablets has left quite a few sites scrambling to create the right assets for a great user experience on mobile. Consequently, there are still a lot of usability problems with mobile traffic for a lot of companies.With a funnel analysis you can very quickly determine if there are problems at a specific or multiple stages of the funnel on mobile.
To get some perspective on segmented funnel performance I would use two forms of analysis:
- Simply compare the performance of your chosen funnel segment to the average performance across all segments. With this information you are able to quickly see any abnormalities that may be unique to that particular segment. A common example we come across is language difficulties for international visitors. This will only affect a specific segment of your traffic and can quite easily go unnoticed unless you perform more granular analysis.
- Use a grid format to discover any problems for particular segments in comparison to a group of highly related segments. One good example is for marketing channels. Are there any abnormalities across marketing channels? Does traffic from paid search have a high quantity of visits but has significantly lower conversion from home page to a listing page? Are we advertising to this audience in the correct way?
A good habit is to prioritise improving the segmented funnels that have the largest amount of visits or revenue. Fixing these will likely result in the highest uplift.
Different customer journeys
The path outlined above is only one of many potential paths a visitor may have on your site. Therefore it is a great starting point but to analyse your whole ecosystem we need to go into further depth on other key customer journeys.
Other common paths to consider are:
Straight to a product page - this is likely to occur a lot from marketing channels such as PPC.
Straight to checkout - this is common for abandoned cart email reminders.
Analyse the above in the same fashion as described previously for your main customer journey
I welcome any ideas, feedback or more examples. Please comment below and let’s have a discussion!