Personalization: Debunking Myths

    Once upon a time, traffic analytics was an expensive pleasure. Only the largest companies could afford it. And even when tools like Google Analytics made this process more accessible, there were few people on the market who could take full advantage of the data.

    Over time, software for testing (split and multivariate) and analytical tools “democratized” and thereby contributed to an increase in the number of analysts who can use them correctly. Therefore, today for websites it is absolutely impossible not to think about the quality of traffic - where it comes from, where and what it is converted to.

    However, democratization (read - popularity and simplicity) regarding personalization cannot be said for now. Yes, companies such as Google and Amazon have tools and professionals who can work with it, but do they have a website that sells doors in Voronezh?

    The answer is yes, yes, but this door store most likely does not need a “deep” personalization, like an international online auction. In order to determine the level that suits most web services in the world, we will try to separate the facts from the two most common myths.


    Myth # 1: Personalization Is Hard

    Nope. Some of its types can be made very simple, although many companies are afraid of this. Some of the most difficult types of personalization are integrated with the site engine (CMS), use marketing automation tools, CRM and offer different content depending on the user's behavior based on their rating (lead scoring - potential client perspective in points). This is an important and necessary type of personalization, but, of course, this is far from the only opportunity to make your site more “close” to the user.

    If you do not have tools for working with users, then you are far from alone. Many companies in the world do not optimize data at all and do not have tools for working with users, but this does not mean that you should bury the hopes of high-quality personalization.

    At a minimum, for starters, you can examine the distribution of traffic by geography and the types of devices from which users come to you (mobile phones, tablets, desktops and laptops) and configure the display of different content / product options depending on the geography of the incoming user. Displaying site content on different types of devices also does not require super-efforts and knowledge of behavioral models, but it can bring you a lot of benefits. This will already be personalization and you don’t have to be the size of Amazon to afford it.

    Myth number 2: It is not clear when to start is

    also not true. If you’re really interested in optimizing your conversion and doing a lot of web analytics, then you certainly understand that even basic personalization works and it’s easy to set it up at any time. But there are two nuances:

    1) you should be able to clearly segment traffic and use A \ B- and multivariate tests;
    2) fully own CMS, CRM and other mechanisms of your service.

    If everything is fine with segmentation, then you will immediately understand which sections of the site drive your business, which have the highest bounce rates, and which have the lowest conversion. From here you will receive information for which audience personalization is needed.

    And if you make full use of the tools implemented on your service, then again you yourself will understand how deep personalization you need, what content you need to personalize, if you have enough data to integrate personalization into the purchase process, and whether your site technically for the full issuance of recommendations.

    In general, if you don’t own Google Analytics or, then it’s better to learn a bit (hire specialists) and start setting up basic analytics and conversion, i.e. first introduce personalization at least at that level. In fact, almost all projects can benefit at least from localization by geography (not only the language of the site, but also by the availability of certain products) and adaptation to different types of devices. In total, if you want to engage in personalization more broadly, first conduct an audit of existing tools, personnel and engage in the conversion and polishing of usability.

    Free translation. Source .

    Also popular now: