How to start personalising the customer experience

Before you start making plans to apply personalisation to your platform, it is important to have a clear idea you are at the moment and where you want to go. You can map this by using the “personalisation maturity curve".

In the personalisation maturity curve, different phases and tactics are compared to the potential increase in customer experience, and thus the potential increase in revenue. As you move up the maturity curve, you’ll come across more complex personalisation strategies. The maturity curve has a total of six phases divided into 3 global groups:

1. Mass marketing

• Single-message communication: Generic websites, newsletters, and advertisements.

• Field insertion: The use of dynamic fields such as “Hello <firstname>" if logged in/registered.

2. Mass personalisation

• Rules-based segmentation: Personalisation based on segments: Show other content on a website or in a newsletter or advertisement based on, for example, the customer's place of residence.

• Behavioural recommendations: Give recommendations based on behaviour: Customers who bought product A were also interested in product B.

3. Hyper personalisation

• Omnichannel optimisation: Personalise all digital touchpoints such as the website, newsletters and advertisements based on data, based on a CDP and/or DMP, and create an optimised customer journey.

• Predictive personalisation: specific and targeted personalisation through the use of data, analytics, artificial intelligence (or artificial intelligence (AI)) and automation. This leads to focused communication to specific customers through the right context, at the right place and time, and through the right channel.

After the 0-measure, the next step is to draw up a personalisation strategy. This strategy is similar to a marketing or conversion optimisation strategy, and helps you to make decisions relating to the execution of the strategy based on facts instead of gut feelings and assumptions. It’s important to realise that this is not a one-time adjustment. You cannot just "turn it on" and then sit back. It is a process that requires continuous monitoring. This enables you to move through the phases in the maturity curve to the level that you aspire to.

A personalisation strategy includes the following components:

1. The goal you want to achieve with personalisation

Without clearly defined goals, for example via the SMART method (specific, measurable, acceptable, realistic, time-bound), it is impossible to measure the success of your chosen strategy.

When we look at online platforms, the type of website plays an important role. For a content heavy corporate website, the objectives will be very different than for an e-commerce platform. Examples of goals per platform type are:

• increasing the number of conversions and sales, and the generation of traffic for an e-commerce website;

• increasing the number of qualitative leads and traffic to the website for a corporate website of a B2B player;

• increasing traffic to the website and thus the revenue from advertisements/ affiliates for a content platform;

• for a branding website: Increasing brand awareness, the number of interactions on the website and traffic to the website;

• for a service or service portal: Improving the findability of important information and functions, in order to reduce costs by reducing e-mails and phone calls to customer service.

2. Whether the required data is present and usable

Without data, there is no potential for personalisation. This is because parameters are needed upon which you can offer your visitors a personal experience. For example, in a webshop you can only offer relevant products to a customer when you know what the purchase history of this customer is, and which web pages they have viewed on your website. The data has to be present, but above all, it has to be usable. For example, via APIs.

It is also important to record precisely how you will measure the success of the personalisation strategy, and if the set goals have been achieved. For example, you can set up a simple marketing dashboard where you can assess the performance on the most  important metrics.

3. What you are going to personalise and what content do you need to do this

Personalisation means that you offer different (groups of) users personalised content, among other things. This can take many forms, such as:

• text and images on a website;

• the content of online advertisements or the content of a newsletter;

• a search engine that works with data such as your purchase and navigation history;

• online advertisements that show products or services that match your purchase and navigation history;

• displaying targeted pop-ups in response to the visitor’s location in the sales could be content, such as positive reviews, a newsletter registration form, the USPs of the seller or a special offer.

For all of these examples, the necessary content must be available, and any missing content must be produced.

It is important that you map your target groups clearly. Based on their needs, you can draft a list (backlog) of potential personalisations for your website. Each item on the list has a predicted impact and required investment in time, money, or technology. This information helps to sort the backlog in such a way that the points with proportionally the largest impact, compared to the smallest investment, are at the top of the list and are implemented first.

4. The necessary skills in the team

For most organisations, a fully automated system that learns and improves based on machine learning, and optimises its personalisation strategy is still out of reach. For that reason, the people who start working on such a process are crucial. Data analysts, (digital) marketers, designers, and developers: there are several profiles actively involved in such a process. Not every company has all of these skills in house, which is why it is good to hire the right specialists or outsource these activities.

5. (Technical) preconditions

The final stage of our strategy focuses on the implementation of the technology needed to make personalisation work. There are many tools and systems on the market, each with its own advantages and disadvantages. To make a good selection, we draw up a list with selection criteria that each tool or solution must meet. Examples of the criteria are:

• future-proofing and scalability: is this a solution that can grow with your ambitions;

• functionality: does the tool have a full range of functionalities that means that we can use fewer different tools;

• flexibility: is the product completely customisable, or does it offer a one size fits all solution;

• maintenance: to what extent does the solution still need to be maintained and updated after implementation;

• required skills of the team: can everyone use the tool, or is specialist knowledge required;

• security and privacy: is the tool safe and does it handle personal data in accordance with the applicable legislation (GDPR);

• minimum number of visitors: there is only really something to personalise when your website attracts a minimum number of visitors per day. With too few visitors, you collect too little qualitative data to optimize your strategy.

If the 0 measurement defines how mature a company is, in terms of personalisation, and if there is a strategy that clearly defines where we want to go and how we are going to get there, we can proceed with the implementation of the tooling required. But which tools should you choose?
When we look at the Marketing Technology Landscape Supergraphic, aka the Martech 5000, it can be overwhelming. With over 8000 tools in total, and many hundreds in the Optimisation, Personalisation & Testing category, there is a lot to compare and choose from.

To get an idea of the possibilities, below are some examples of tooling that we use frequently at iO.

Personalisation of content

With content personalisation, you show segments of users different texts, images or calls-to-action. This type of personalisation often falls within the rules-based segmentation phase of the maturity curve. To achieve this, you create two versions of a page or part of a page within a CMS  and then decide which segments of users see which page.

Many content management systems (CMS) can already personalise content out of the box.  The large enterprise CMSs Sitecore, Episerver and Adobe Experience Manager offer this by default. Open source systems such as Umbraco or Drupal can also personalise content in the CMS using an external module. 

One disadvantage that is worth taking into account is that the CMS can become very cluttered when there are many different variants of pages. It is difficult to keep a clear picture of what the website looks like for each can become challenging.

A CMS is not necessary, for example in the case of the website of our customer Natuurmonumenten that sorts the overviews of all nature reserves, activities or play areas based on distance from the users current location. This is tailor-made within the platform and works based on the IP address of the visitor and cannot be managed via the CMS.

Smart notifications

Another way to personalise your website, within the rules-based segmentation phase, is by    Based on analysis of visitor behaviour on the website, you can determine their location in the sales funnel, and then display customised notifications, such as positive product or shop reviews, newsletter registration form, the seller’s USPs or a special offer. One of the original Dutch tools that provides this functionality is there are several other tools that offer a similar solution.

This proven method provides added value particularly for e-commerce environments. For example, when Gartner investigated the effects of personalisation on webshops they concluded that receiving the right message at the right time has a major impact on purchasing behaviours.

Real-time personalised e-mails

From specialised email marketing tools, to full marketing suites, there are a wide range of solutions available that allow you to reach your target audience in a personal way via email. All major tech companies have “Marketing Cloud” solutions and at iO we have extensive  experience with them all. From Hubspot Marketing Cloud to Salesforce (Pardot and Marketing Cloud), and from Adobe Experience Cloud to Selligent Marketing Cloud.

Recommendation tools

When making recommendations, we immediately think of well-known examples such as Amazon, which gets about 30% of its revenue from personalisation algorithms. But solutions like this are also feasible for smaller webshops. E-commerce platforms such as Magento and Shopware offer modules that make product recommendations based on visitor behaviour.

You can also make recommendations in other areas. For example, we developed the Film Finder for the International Film Festival Rotterdam (IFFR). A personal profile is developed based on the answers to a number of questions, trailers that are given a thumbs up or down, and purchase history. That profile is then compared with the profiles of other IFFR  visitors. The user then receives personal recommendations for movies that fit their user profile. Thanks to smart applications like these, online conversion and user satisfaction on the platform has increased by more than 40% in recent years.

The driving force behind the Film Finder is Recombee, a recommendation engine that makes recommendations based on artificial intelligence.

Search engines

"I would like search functionality like Google on my website." We regularly hear this request from customers, and then we prove them wrong. As users get more and more used to the experience offered to us by big tech companies such as Google, Netflix and Spotify, they expect their own products to offer the same experience, including a comprehensive website or webshop search function.

This is also possible thanks to the smart use of specialized tooling. An example of this is Algolia, a search engine that not only indexes your website and provides search results, but  also offers additional functionalities such as the return of personalised results,  comprehensive statistics t.b.e. analysis and optimisation, A/B tests, autocomplete  functions and search suggestions.

Get even more out of your personalisation strategy with a customer data platform (CDP)

The techniques and tools described in the previous chapter can be implemented separately.  What most of the examples have in common is that they personalise based on segments  and only affect your website. This means that they fall within the third and fourth phase  ('Mass personalisation') of the maturity curve. How do you make the leap to 1-on-1 personalisation across personalisation'? 

Investment is required to achieve this. Both organisational and technological. Marketers need a 360º view of the customer to deliver the desired personalised experiences. Different systems and technologies need to be aligned and fragmented data needs to be merged.

This is achieved through the implementation of a Customer Data Platform (CDP). A CDP combines customer data from different sources in a structured way, so that the CDP  becomes a 'single source of truth'. This source can be used by the various systems throughout the organisation, to convert the data into actions, campaigns, marketing analyses and Business Intelligence (BI). Gartner's research states that CDPs receive broad attention from marketers because they "provide a holistic view of the customer to execute    and optimize personalised journeys"

Examples of resources that can be combined in real time:

●       Data management platforms (DMPs)

●       Marketing automation systems

●       E-mail systems

●       CRM systems

●       ERP systems

●       POS systems

The CDP is your starting point to set up extensive customer experiences across different is a facilitator of, for example, efficient retain, upsell and cross-sell actions. Based on a large dataset, and not just website cookies, ad trackers or CRM data.