Virtual salesperson, making it work

That's the person in the store who makes you feel welcome, gives you encouragement, makes suggestions. That's the salesperson who makes the difference between you walking out the shop with a new jersey only - or walking out with two new jerseys, a pair of gloves and a trendy new scarf.

However, the salesperson's job is not done when the sale is rung up. When the customers are gone, good salespeople talk to the business owner about trends, about customer patterns, about what is hot and what is not. They are the vital source of intelligence about what makes the customer tick... and in the online world, what makes them click.

Good salespeople are all-important in retail, but online, the salesperson is not a person at all; it is the intelligence that should be built into the technology that drives the website.

The virtual salesperson should have the central goal of driving up the spend of every customer who logs on to your site. It should encourage people to spend rather than simply browse (conversion rate lift), it should work to increase the average order sizes (cross sell and upsell) and it should seek to drive repeat sales (customer loyalty).

More than back-end

South Africa's retailers are realising that an internet presence will drive medium to long-term success as shopping moves online and are looking at the technology they will need. Most attention in the beginning tends to fall on the back-end (stock management, financial management, etc). Closer to launch time, it is all about the pretty website. Nevertheless, there is a lot more to optimising the front end than tasteful colours and clever animations. That is because the ecommerce site is not only the shop-front but also the salesperson.

When designing your online 'web' shop, think of the ecommerce system as a virtual sales force. The technology platform should have a strong feature set, such as being able to interact in any language. It should be endlessly reliable, on duty around the clock. It should also be smart, to anticipate and suggest choices and accessories to the customer.

Gathering data

Your ecommerce technology should adapt to each user and their preferences, while working for you, thinking for you and especially selling for you. The key is that the virtual salesperson is able to gather and use data.

Some data is freely available, for example, people's use and travels through their favourite social media platforms and recommendations sites. There is also much information that can be gathered by people's interactions with affiliates, search engines and analysing how they move from one part of the internet to your site. Once on your site, there are 'perception' capabilities, such as site sensors, channel detection and audit trails, which can be combined with user state and time awareness to provide further context to let your virtual salesperson sell better.

Your ecommerce platform must recognise users and respond to them in a relevant and personalised way. It should intelligently tailor search results, navigation, product information and content to e-shoppers.

Applying existing data can result in predictions and recommendations for individual users; it can also work on information from many users to make assumptions about what new customers may want.

That is the power of concepts such as neural networks, Game Theory, Bayesian probability and Fuzzy Logic. These are just some of the techniques and strategies, which can determine how your web shop interacts with the customer. Many of them are built into modern ecommerce systems.

As you start looking into your web presence, consider just what kind of a salesperson it will be. Once you get people to your website, whether they spend money is often less dependent on them seeing a perfect, gorgeous site, as much as it is on how well the web shop is designed and whether the virtual salesperson is helping them at every step to click "buy".

Updated at 8.32am on 15 June 2012.