Blog

By JessicaSanders, 22/08/2012
No comments

As a retailer, you likely have thousands of records from your POS system and online retailing software. Within this is purchasing data that can be valuable to your business for a variety of reasons, from marketing optimization to purchasing demand. However, whether you’re looking to gain a competitive edge or be more prepared for customer demands, you have to know what to look for and how to analyze the data in order to make it effective.

Purchasing Prediction

Your data sets will likely show your customer’s purchasing processes. Especially online, this is valuable information to have. When you know where your customers are starting, browsing and ending up, you can optimize the customer experience to lead them toward more conversions; there are two important aspects to seek out.

• Searching process: Look through your products and map out the route your customers took to find them. If they used particular search terms, be sure to optimize for those to get your customers where they want to be right away.
• Purchasing process: According to Invesp, the average shopping cart abandonment rate for e-retailers is 65.3%. Use data to see where your customers drop off, and use it to improve your process.

Forecasting Product Demand

Another smart way to analyze your data is to look for purchasing patterns; both in store and online. This is a great way to manage your inventory and be sure you have everything your customers will need. According to Information-Management.com, “Predictive analytics … allows the retail organization to enhance its decision-making powers by looking at the future with analytical rigor.” In the end, this can lead to customer loyalty and overall satisfaction. So, what should you look for?

• When purchases are made: Look to see what day, month or year your customers are making purchases for particular items. If you provide a service, what time of year does it seem you are most needed? Be sure you are fully staffed or stocked to ensure quality customer service.
• How much: Determining how much of each product is bought will allow you to determine whether your customers want more or less of something.

Campaign Effectiveness

As a business, you’re always looking to sell more – and marketing campaigns are an important part of that. However, if your efforts aren’t reaching your customers, then they are rendered useless. Instead of producing coupons and offering discounts blindly, consider whether your campaign needs to be better optimized, shared, or modified all together.

• Coupon codes: How many of your discount codes were used at checkout? Quantify the coupons blasted out via email, or linked to on social networks, and compare that to the number you see in analytics.
• Before and after: Compare the sales that were made before your campaign and afterward. Is there a large difference? If not, you may need to consider modifying your efforts.

Collected data is a valuable asset to your retail business. With it, you can optimize marketing efforts, product management or customer needs. So, instead of letting it sit untouched, try to analyze it for the aspects of your business you’d like to improve.

Etiquetas: coupon codes, data, e-retailer, ecommerce, online sales, online store, purchasing, retailer

No comments participate and write one

By Frankie, 04/05/2012
No comments

A lot of people consider recommender engines as a tool to achieve cross-selling and up-selling strategies in ecommerce. However, these services can be more powerful and provide other functionalities that complement and enhance the final utility.

There are four key elements to consider in every good recommender engine: data, algorithms, design and metrics.

Data represents different properties and characteristics of users and products within the online store. They generally refer to actions that users perform on the products, which include visits to the product pages, shopping, ratings and comments, among others. But some of the main problems are data representation and collection. Representation can be qualitative (visits to product page, likes or dislikes, inclusion in wishlists…) or quantitative (mainly ratings). Collection can be implicit, like the data users provide for the simple fact of making the purchase process (e.g., visiting or purchasing products), or explicit, when users provide information in a proactive manner (e.g., ratings or comments). The latter is much more difficult to get, since the user must feel the need to provide it.

Algorithms use data to determine which products to recommend to each client on the web. Surely this is the most complex element, and the main entry barrier for a company that wants to develop its own recommender engine, not just by how easy or difficult it may be to develop these algorithms, but also for the scalability issues that occur as data size grows.

A positive point is that there are a number of well-known algorithms that produce good results for ecommerce. These algorithms generally use collaborative and content-based approaches, and they have proven results. The first problem here is that not all algorithms are suitable: the simplest produce worthless recommendations, and the most complex take too long to generate the results. Therefore, the algorithms used to get a successful outcome must sufficiently advanced, but also simple enough to make calculations in an acceptable time. The second drawback is that each algorithm has a number of attributes that make it more suitable for different types of business. So it would be very convenient to have a recommender engine that provides several algorithms and use the most appropriate in each section of our online store.

We know that design and usability in ecommerce are of great importance: a good product with a bad design goes unnoticed, and recommendations are no exception. Deciding the best location to show the recommender results, the number of items to display and the visible information for each product can encourage customers to follow a good recommendation. However, a bad combination of these may turn them invisible.

And finally we come to the metrics. Surely you will be familiar with the phrase ”You Can’t Improve What you Don’t Measure”, and this is a truth that we should all take to heart in any website. It is also true for recommendations, either to check the improvements achieved by the system, or to compare the effectiveness of various recommenders and choose the best.

There are multiple recommender engines in the market that give support for cross-selling and up-selling in ecommerce, but most of them are too simple and offer only a minimal solution for a subset of the described key elements. If we want to be serious about our businesses, we should search for a tool that gives prominence to all of them.

Etiquetas: algorithms, conversions, data, design, ecommerce, metrics, Recommender Systems

No comments participate and write one