On The Need For Data-Driven E-commerce Funnel Optimization
Working with a retail analytics platform — Insense — has taught me a lot of things that I have shared through a number of blog articles in the past and today is yet another one that most e-commerce platforms know but don’t take seriously — conversion funnel optimization. I spent a good part of my yesterday going through over 20 e-commerce platforms in Kenya and most of them were eerily similar not just in design but also in content and process flow and that means that the challenges that one of them faces is probably reflected across most of the others.
E-Commerce in Kenya isn’t a small thing. According to Statista, we expect a total revenue of $1.003m this year (2020) after corrections for the effects of Covid-19 — that’s a 51.6% YoY growth. By the end of the year 2024, the revenue is projected to reach $2.057m; double the value that it is this year. Kenya is roughly ranked number 59 globally in e-commerce with an ARPU of $64 but we can take these impressive statistics even a bit higher if we did a bit more optimization on the e-commerce platforms. For instance, as I was going through these e-commerce sites, one of the things I also found out is that nearly all of them require their users to enter a postal code/zip code. The presence of the word “zip code” in itself is the very first menace as most people will start wondering what a zip code is — we are Kenyans and we don’t have zip codes; for the few who get to understand “zip code” or see its alternative — the postal code — the next challenge is what exactly they should enter in that field. As we all know, Kenya’s postal address system is mostly reserved for companies and institutions and common citizens don’t have postal codes for their residences. It becomes a nuisance and most people would either enter fake values or postal codes for their former schools or if in formal employment, of their employer company — in most cases, a rarely useful (for delivery) but extremely disturbing field but we still insist on making it a mandatory input field.
By the end of 2024, revenue from e-commerce is expected to hit $2.057m, twice as much as this year’s (2020) projections.
On the Insense Platform, Funnel Analysis, which is one of the solutions under the Clickstream Analysis product, measures performance and drop off rates at four key stages: when products are viewed, when a product is added to cart, when checkout process is started and when an order is completed. This is done in relation to other parameters such as the device type, browser type et al to be able to understand important KPIs such as the Add to Cart Rate and Cart Abandonment Rate but most importantly, to show you every important stage and the bottlenecks in each stage so as to allow you to perform data-driven optimizations on your platform.
Add to Cart Rate is a measure of the percentage of users who add products to cart out of all the users who visit the site while cart abandonment rate is a measure of the percentage of those who don’t complete their orders having added products to cart out of all those who added their products to cart.
Let’s use Masoko as an example. From SimilarWeb, Masoko received a total traffic of 150,000 in the month of May 2020. We seek to see the approximate sales Masoko would make under different scenarios or levels of optimizations. According to Littledata.io, the average add-to-cart-rate for e-commerce industry is 10% ie only 10% of those who visit your e-commerce site would end up adding a product to cart. The industry cart abandonment rate is also roughly 70% which means that out of all who add their products to cart, 70% of them will leave before completing the purchase:
- Average Add to Cart Rate = 10%
- Average Cart Abandonment Rate = 70%
- ARPU (Kenya) = $64
For Masoko, if their website has the industry average values mentioned above, it means that out of the 150,000 site visits they received in the month of May, only 4,500 of them ended up completing a purchase. With an ARPU of $64, that translates to a total of $288,000 in total revenue earned for the month. What would happen if Masoko had better conversion rates, say:
- Add to Cart Rate = 15%
- Cart Abandonment Rate = 50%
By lowering their cart abandonment rate and increasing the add to cart rate, then Masoko, having received the same traffic of 150,000 users in the month of May at the same ARPU of $64, they would have a total of 6,400 of those users making purchases; an increment of 42% in number of completed orders. Their total sales would increase to $480,000, an additional revenue of $192,000 representing a 67% increment in revenue for the month.
The numbers are surely impressive and every e-commerce platform would be yearning to get that extra $192,000. The process to achieving that is, however, not a walk in the park. First, every platform has its own challenges so there is no one solution fits all; the sole reason why every e-commerce platform needs to identify their own bottlenecks. As earlier mentioned, some bottlenecks may only affect mobile users while others may be affecting desktop users; some may be affecting Safari users while others affect Chrome users and others Opera Mini users. That’s what motivated us to add the Funnel Analysis on our Insense platform — to help e-commerce platforms to better understand their KPIs and be able to act on specific as opposed to generalized solutions.
A recent study by Baymard Institute, on reasons for cart abandonment, reveals industry-level insights that are worth highlighting when it comes to cart abandonment. For instance, 58.6% of US online shoppers abandoned their cart in the last 3 months (from time of study) because they were just browsing and not ready to buy. This is reflected on the SimilarWeb stats on the Masoko platform whereby the bounce rate is indicated to be 59.48%. Bounce Rate indicates the number of one-page visits as a ratio of the total page visits. Whereas the bounce rate doesn’t necessarily indicate users who were just browsing, a large part of it is made up of users whose intentions weren’t really to buy but just to browse. It however, doesn’t include users who, in their browsing mission, conducted search and maybe didn’t get what they were looking for. Leaving those who are just browsing aside and focusing on those who actually want to buy, why do only 3% of them end up actually buying? From the Baymard Institute study, half of them quit because of high extra costs they hadn’t anticipated eg shipping costs, packaging costs, taxes et al. 28% of them quit because they were asked to create an account and a further 21% quit because of a complicated checkout process. The chart below highlights the full reports summary from the study.
58.6% of US e-commerce visitors just want to browse and are not ready to buy. Only 10% of the total users will add a product to cart and out of this 10%, 28% will quit as soon as they see a log in or sign up form to complete their order. Unnecessary form inputs, complexities or errors in the checkout page will cost you another 21% of the potential customers.
Personally, I felt irritated by all the over 20 e-commerce sites I visited during my tests mainly by their complex checkout forms. Nobody likes entering lots of form details and I am no exception. Masoko and Naivas, for instance, require one to log in or sign-up before proceeding; if you have forgotten your password, you can imagine the complex process you’ll have to go through to complete your order. Tuskys Online allows one to add products that are out of stock to cart (with no indication) and then when you go to the checkout process, it becomes impossible to proceed due to “errors on your cart.” Other sites do not show you when you add a product to cart and one goes through struggle to know where the cart is for them to proceed while others just have way too many form elements to fill in at the checkout process. For instance, Tuskys Online will tell you that there is either only town delivery or pickup from any of their branches but still require you to enter your street address (I don’t know my street address by the way so I’ll give a nearby major road I know) or ship to a different address with an even larger form; why are these necessary? In the US, for instance, the average checkout flow contains 23.48 form elements (14 of which are form fields) displayed to users by default. Most of the form elements aren’t that necessary and could be reduced by nearly 60% but most still choose not to take this golden opportunity.
Imagine if Masoko could convert all the 40% of their site visitors who came to buy on their platform, that would be a total of 60,000 users buying but instead, due to conversion bottlenecks at various stages of their funnel, they only ended up with 4,500 people buying — you can notice what a revenue loss this represents. However, we cannot aim for 100% conversion. The first step is to yearn to hit the industry average — thereby if your add to cart rate is below 10% then you need to aim to increase it to that and if your cart abandonment rate is higher than 70% then you need to aim to lower it to 70% as a first step in the right direction. Like I said before, there is no one-solution-fits-all and the stats above don’t apply to all; every platform has to monitor their own KPIs and make optimizations informed by data. In most cases, this will be an iterative process that might even include A/B testing before final rollout to eliminate bias and assumptions but a journey worth taking. As I have always said, moving to an e-commerce platform from brick and mortar isn’t the end game, it is the start of an interesting digital journey.
Disclaimer: The absolute figures represented for Masoko are not verified facts by Masoko/Safaricom and in no way represent their true figures. Traffic data is from SimilarWeb as already highlighted and all the other values are based on hypothetical KPI values based on industry averages.
For funnel optimization consultancy or on details on how your e-commerce platform can make use of the amazing data solutions provided by Insense, feel free to drop us an email to info@insensedata.com