Increase sales, drive up conversion rates from marketing activities and optimise the underlying performance of their websites in 10 not so simple steps:
(1) Refocus on the 98% who didn’t buy
The retail sector is well versed in using Commerce and CRM systems to collate crucial data on existing customers and their purchasing patterns. This is a comprehensive view of transactions and conversions and builds a picture of demand, sales, and performance. But this picture is not complete. The focus of these systems is purely on the 2% of visitors that converted, but what of the 98% who did not? Use visitor centric analytics to gain the same level of information on your potential customers, thus allowing you to refocus your efforts where there is a much greater potential for return.
(2) Identify prospective customers who are ‘in the market’
The Internet encourages buyers to shop around... often with a heavy focus on price over other value propositions. Due to the easy accessibility of choice, decisions can be made quickly, and the length of time a prospect is 'in the market' is shorter than ever. It is possible to overcome this disruptive purchasing pattern. The trick is to react quickly to a site visit (an expression of interest) using targeted marketing communications to “close the deal”.
(3) Drive up “Browse to Buy” for each product
While knowing product sales is a vital measure, it tells only half the story. What if visitors were finding and viewing products, but not converting? This is more common than you think. By comparing ‘product views’ with actual ‘product purchases’, it is possible to spot products with a low conversion rate. Creating alternative product descriptions and selling messaging can have a serious impact on conversion. Multi-variate analysis enables you to record the different permutations and determine which are performing best. You can then build a matrix of all product searches, views and purchases, which can be used to identify and drive cross-sell and up-sell opportunities.
(4) Reduce dropouts in the commerce funnel
How confident are you in your ecommerce process? Could conversions be higher if it was working better? Are there problems effecting performance? Traditionally, reporting systems have tackled these concerns by focussing on common “path deviations” at each stage in a funnel, but this information is rarely actionable. Realistically, how can one control the path a visitor takes? Is that information really pointing out an issue that can be addressed? Instead, focussing on the simple things that can be influenced. For example, where do prospective customers drop out of the funnel? How well does the site perform in different browsers? Do delivery charges or other add-ons prevent the sale? Perhaps, particular promotions/offers create clicks but do not convert? With this information, changes can be made to radically improve the performance rate of each stage i.e. reworking a complicated registration page, tailoring delivery charges and other add-ons, or increasing payment options. All of these are clear, actionable areas where performance can be improved.
(5) React to dropped shopping baskets
For a purchase to take place on your site, a visitor will have made a journey through a number of web pages from entering the site through to the sales completion page. However, it is almost inevitable that along this journey there will be a number of visitors who drop their shopping baskets and do not complete their transactions - ‘drop-outs’. You need to identify the visitors that have dropped their basket at each stage, and apply different scenarios to determine “why”. This information can then trigger specific marketing responses, such as special offers, incentives and product alternatives.
(6) Identify opportunities and ensure the product portfolio is meeting demand
In a fast moving marketplace your product portfolio can quickly be out of touch with demand. Has summer come early? Has demand changed? Has a product reached the end of its lifecycle, been superseded or gone out of fashion? You need to react to changes in demand using real time “site search” data that is gathered automatically for every visitor. Search data is particularly powerful as it can reveal the needs and interests of your visitors, and allow you to spot new trends and interests early. Spotting new demands and trends or changes in product life-cycles early on can help to improve lines and better align your product portfolio leading to potential sales. This data can also be applied to your search based marketing activities – is search driving the right visitors? Can your PPC keyword terms be improved or bid rates optimised?
(7) Improve search, promotions, product categorisation and features
Typically, visitors to a retail site locate products through one of four means: categories, on-site search, themed promotions or featured products. Often, retailers are unclear as to the relative performance of each approach, and whether they are successfully leading to products being found and selected. In particular, for multi-product retailers, the effectiveness of the search facility is pivotal to the success of your site. Are the best products returning? Do the searches lead to a product being selected? What search terms follow an initial search? You need to determine how well these methods are working, how the visitor reacts to them and how they could be improved.
(8) Optimise Marketing Conversion
The harsh reality is that if 98% of visitors are not converting, they are likely to be costing you money. With PPC bid rates increasing, email click through declining and other “paid for” marketing campaigns becoming increasingly competitive, it is necessary to be clearer than ever on which campaigns are performing and which are not. Of course, Advertising, Affiliate, Email and PPC systems all provide conversion data. But this data is fragmented – focussing purely on extolling the benefits of the activities that system supports. This is not a single view, nor is it a view that necessarily distinguishes real success. To really understand performance, a commerce orientated view is required. One that relates all searches, products and purchases to the marketing activity responsible.
Why is this useful? Well, these elements all work together. If a particular marketing campaign leads to visitors with a particular interest, perhaps the landing pages should reflect this? If a particular promotion leads to a higher drop out once delivery charges are introduced, does the promotion need amending? While actual click-throughs may be sparse, perhaps your Social Media is a key factor in your success. Perhaps your new product range was responsible for that increase in sales, rather than those keyword changes?
(9) Integrate browsing data into CRM and other marketing systems
While sometimes key data does not exist, many retailers complain of data overload. Whatever your situation, it is likely that you have met obstacles putting your web channel data to work within your existing systems. Often, the data cannot be tied together simply due to the lack of a common ‘key’, or because systems have not been designed to communicate in an open manner.
Your chosen web channel analysis system needs an open API that allows data to be shared with other applications. These applications can however only understand the data if it contains a common, shared “Contact ID”… the key. A “contact” is an identifiable person with an identifiable history. Unfortunately most of the traditional web analytics system store their data about “Visitors” where a Visitor is in fact a PC or Internet access device. Since most people now regularly use more than one access device a contact is not the same as a visitor. There are very few web analysis systems that can report “Contact” activity.
Having selected one that can report on Contacts, you should store data on every aspect of a visit to your site; there are many possible applications. For example, you can identify visitors who have searched for products that are ‘out of stock’. These visitors can often be ‘retargeted’ once that product is back in stock. Similarly, if a number of visitors are searching for a product that you don’t currently have, could it be sourced quickly? If a product has been viewed, but not purchased, perhaps a follow up is all that is required to close the order?
By combining web behaviour data with loyalty schemes, CRM and sales data, it is possible to drive targeted promotions and marketing campaigns, such as emails, to individuals based on their behaviour, interests and preferences.
(10) Avoid lost sales caused by technical issues
The reality is that while retailers naturally focus on sales and marketing opportunities, it is often technical issues that have a more significant bearing on performance. Technical issues on a web site can lead to incomplete transactions, with consumers leaving your site frustrated and possibly even avoiding your site in the future.
This is more common than you might expect. A common cause of lost sales is browser incompatibility, with at least four mainstream browsers (and many versions of each) in common use. Are you confident that every web browser your visitor uses actually fully supports the payment systems on your ‘check-out’ page? Similarly, mobile internet use is on the rise, and often, sites that work well on the web, simply do not work properly on the stripped back browsers on a mobile phone.
You need to identify if, and where, there is a technical problem with a particular browser, platform or process and mobilise the appropriate business, marketing and technical resource to resolve the issue.
