How to set up and run a small business

Leaderboard – Run – Sales

You are here: Home » Run a Business » Sales & Marketing » Turn your website browsers into buyers with personalised product recommendations

Turn your website browsers into buyers with personalised product recommendations

November 18, 2016

Understand how to shape the customer journey and convert browsers into buyers.

The more relevant you make your website for each visitor, the longer they’ll stay, the more they’ll do, the more they’ll spend, and the more they’ll tell other people. And a great way to do this is to make the experience more personal.

Many small online retailers are missing out on the benefits personalisation can bring, so we asked Christer Holloman, author of How to Sell Online to reveal how you can take advantage of the opportunities it offers;

e-tailers know personalisation is becoming critical to online success

Personalising your ecommerce website makes sense to every online retailer, but only a few really know where to start and what to do to get the best results. In fact, 98% of respondents surveyed by eConsultancy saw personalisation as critical to success but only 22% had implemented it in some form or other.

Personalising website content is often thought of as the preserve of huge, multi-national retailers and not available to the mass market due its cost and complexity.

Not only that, many retailers are wary of trying to implement anything but the most basic forms of personalised content and promotions on their ecommerce sites and emails. They are unsure of where to begin, whether they already have or need to buy the technology required to personalise the online experience, how to justify the time and expense, and how to prove the results.

Even basic personalisation can make a big difference to sales

Retailers may display very basic recommendations such as the items a user last looked at or what other users purchased when ordering a certain product, but little else. Most businesses tend to adopt more of a generic ‘one-size-fits-all’ approach so are missing out on the huge potential that even basic personalisation can offer, including increased conversions, orders, average order value and customer loyalty, as well as a reduction in cart abandonment and bounce rate.

The capability exists to customise entire user experiences down to an individual user level – serving unique content, pricing and promotions to each, but the resources, effort and cost required to do this properly in such microscopic detail, means it’s unlikely to happen for all but the largest retailers.

More achievable for the smaller retailer, is the scope for delivering content (in the form of on-site advertising, products, offers & promotions) to more generic groups of users and providing a degree of personalisation that those remaining 78% of retailers still won’t be delivering.

RELATED: Turn your website into a deadly selling tool with these 5 steps

Personalising ‘Product Recommendations’

Often the first, and most basic, step for the majority of retailers in delivering some form of personalisation is the use of product recommendations, which can be delivered by almost all ecommerce platforms nowadays. In fact, personalised product recommendations can help generate around 10-30 percent of ecommerce site revenues.

Depending on the sophistication of your ecommerce system and the data you have available these may be delivered manually, via automated algorithms and metrics, or a combination of both.

Types of Product Recommendation

There are at least three different types of product recommendations to consider, all of which can serve personalised content aimed at increasing the user’s likelihood to buy.

1. Similar

Items that are similar to or can be substituted for the ones being viewed. These might be slightly different versions, different brands/makes or have more/less advanced features and a higher/lower cost.

2. Complimentary

Offering additional or complimentary products to those being viewed. Often advertised under guises such as ‘Complete the look’ or ‘Customers also bought’ trying to upsell and increase basket size.

3. Recently Viewed

What other items a user has recently looked at (most likely session specific – but more advanced versions can remember users from previous visits either via account login or using cookies)

Many ecommerce sites deliver only one of these three types of product recommendations across their entire website. Increasing the types you offer can help target different customers, with different buying requirements & habits, at the same time, whilst also providing the capability to upsell.

Placement of Product Recommendations

Another thing to consider with product recommendations is their effectiveness on different pages in the purchase cycle. For instance, tempting users with last minute or special offer items on the checkout page can encourage impulse buys, whilst giving them alternative options (potentially more advanced or more expensive versions of what they are looking at to drive an increase in your average order value.

You could create an algorithm that will not show an alternative product of a lower value for instance) may work better on the product pages during their browsing and research phase.

By beginning to personalise the output of these recommendations, using more than one type of recommendation at the same time, and thinking about the most effective place on your website to put them, you should expect to see significant increases in conversions.

Displaying profile-driven recommendations

When you have mastered the more generic use of product recommendations you can move onto displaying items specific to individual users based on the data you have gathered and profiled such as demographic, interests and behavioural.
browsers-to-buyers
When done well, product recommendations tend to use a combination of technology and manual input for best results.

As you develop, and more information is gleaned to populate product recommendations, the more personal you can get.

You can begin to use customer data such as demographics, interests or website interaction, alongside manual input from merchandisers, product experts or stylists.

Due to the time involved when trying to manually populate your product recommendations, particularly when you start, you may need to heavily rely on automated systems.

However, don’t let that stop you from having some human input and watch out for the following common failings with some automated systems:

  • Product Recommendations being the same across all products, and thus often unsuitable, regardless of the product being viewed.
  • Seasonal items being displayed after the event – no real point showing people Christmas decorations that somebody bought before Christmas, in January, as they’re unlikely to be relevant.
  • Recommendations being the exact same item as that currently being viewed.
  • Recommending other colours or sizes of the same product (these variables should always be shown on the same product page).

About the author

This guide has been written exclusively for ByteStart by Christer Holloman, author of How to Sell Online: The Experts’ Guide to Making Your Business More Successful and Profitable Online, published by Pearson, priced £15.99, available to pre-order on Amazon.

More on selling online & digital marketing

For more help on making your small business successful online, try these other ByteStart guides;

Selling online

Successful Social Media

More from ByteStart

ByteStart is packed with guides on all aspects of starting and running your own business. Check out some of our most popular guides;

Promoting your business

Starting Up

Leading a business

Funding your business