JD Duarte


JD Duarte explains how to analyze eCommerce traffic

When talking about growing an eCommerce traffic, it is convenient to think about the Growth Hacking methodology, which has as its backbone three elements: the customer conversion funnel, a Wow product and data-based decision-making. JD Duarte, an entrepreneur and eCommerce expert from Costa Rica, provides insight into how to analyze that traffic to achieve greater growth.

Few companies have a data analyst, since it is usually an expensive resource to maintain full-time, so those who need to have professional analysts often resort to hiring external services by the hour; in general, we are talking about large companies. But what about small and medium-sized enterprises? The reality is that they make decisions based on intuition or interpret data without experience doing the best they can, some even using “vanity” data that distances businesses from true financial goals.

The company starts from the platforms it has, such as eCommerce and/or a marketplace; some even configuring a Google Analytics and are dedicated to evaluating no more than three reports in each of them. They obtain comparative information, usually on sales, traffic on the site, products sold and even basic demographic information.

The current situation is compared to the previous month or year, and it is determined whether things are going well or not according to the proposed goals. The area of opportunity in a scenario like this is that dozens of reports are stopped reviewing because they always focus on the same variables and, of course, always obtain the same results.

Another frequent scenario is when you have several platforms and you do not manage to concentrate all the information in one place. Still, different tools throw different information, or make decisions based on information from a single channel having three or more sales channels, each with different behavior.

Worse still, there are many cases of companies that review the reports generated by their platforms, detect that something is not going well, and decisions are made to improve the situation, but without seeking more data that reveals where the problem really lies.

Thus, in general, many companies come across large amounts of data and end up segmenting information without a methodology and making erroneous decisions. “A simple example is when the decision is made to fire a marketing manager for low sales and later at the time of analyzing the data, it is identified that the traffic of the site went up, as well as the time of stay and traffic,” states Duarte. A correct interpretation would reveal that the advertising work was excellent, the problem was actually in the payment gateway. Still, it was not known how to analyze the data to identify the real problem.

This is a fairly important scenario: sometimes, businesses evaluate data that is not relevant to their objectives. Not all the data you have is needed, and not all the data you need is at your disposal.

Once you do the analysis by answering several questions that help you identify those opportunities for improvement, it is when you can infer what the main objectives you have for your business, what are the things you need to correct or improve to improve each instance. Having clear objectives, it is time to determine what those KPIs or quantitative indicators will be that will allow you to measure the achievement of the objective in a given period of time.

This is a clear example of how, according to the company’s needs, it can start from objectives that help it improve and therefore identify how to measure them. The issue is that it will often be necessary to create or build those forms of measurement.

For the same example, online tools can be used that measure the speed of the site, identify total traffic, how many and which are first visits, and which are not. Another useful tool would also be a heat map, which allows you to know the behavior that coincides with a certain type of segmentation, how it moves around the site, where you click, what pages you visit, etc.

When you already have data associated with the aspects that you need to improve in your business, analyze them and put them in context, and this is when they become valuable information for decision-making, so to make a better evaluation of that data, we suggest you take into account three important considerations.

There are certain KPIs that, when you review them month by month, tend to grow or have variations of a certain amount, such as sales, units sold, site traffic, etc. If you check those numbers and carry out promotions and campaigns, you will surely see that the increase in those variables goes beyond the normal trend.

The same can happen when you do something out of the ordinary, you changed something on the site, did another type of promotion, or changed the way you make a campaign, and you see different indicators. Adds Duarte, “When the numbers change radically, they are talking. You must identify which factor generated that behavior and see how to improve it either because it was positive or because it was negative, and you must adapt your strategy accordingly.”

Either because you identified a change of trend or because you review behaviors of certain variables, you can infer behavior in your business. For example, when you invest more in an advertising campaign and traffic obviously goes up. If we only consider the traffic variable, we deduce that an achievement was obtained. Still, when it is analyzed against other variables such as session duration that decreases, and rebound rate that increases, you already infer that that new audience that arrived is not well segmented.

That is not a valuable audience and therefore your CPA (Cost per acquisition) KPI will rise. This is the way you can detect and analyze different variables to identify the impact of a decision. In a case like this, the ideal would be to stop the campaign and reevaluate the communication and/or segmentation of it.

A good interview can give you more valuable data than a report. When you conduct an interview, you can invite your best customers to generate ideas for improvement and thus obtain an even greater benefit. For example, if you detect that you have a logistical problem in deliveries and evaluate delivery times, delivery status, tracking, etc., and identify possibilities for improvement.

At the same time, though, the customer suggests that you sell product packages in a certain way. Not only will you be solving a problem with insights presented by your own customers, but you can adopt other business modalities that can increase your sales. Always listen to what your top customers have to say.