Digital interactions and the platforms that come with them have been constantly evolving over the last 20 years. The same applies to the tools used by digital professionals, which are becoming increasingly advanced. This has encouraged companies to create multiple specialised teams, which has inevitably created siloes.
In AT Internet’s recent DAA webinar, our Data Expert Declan Owens highlighted the importance for companies to introduce a unified data model approach across their organisation. By ensuring that their analytics are at the heart of their digital product & marketing strategies, companies can ensure that it is fully aligned with their business needs.
AT Internet’s new Analytics Suite provides the essential tools and ideas to implement a data model strategy into your business and leverage the true value of your digital analytics.
Today’s analytics challenges
In the beginning AT Internet simply had the internet user. They went online via a desktop and hey presto – straightforward analytics. Zooming up to 2020, the customer journey is now far more sophisticated, composed of multiple touchpoints, that can be physical, digital or both!
In terms of analytics, this is compounded by the fact that each specific technology has traditionally had its own set of tools, from web analytics to app analytics, or even voice analytics, chatbot analytics, etc. The silo effect also applies to digital professionals who have also tended to focus and worked on each specific technology, with little or no interaction.
The same goes for the vocabulary used in analytics – page_views can also be equivalent to screen_loads, app_open, message_sent, action_sent, etc. depending on the technology. And translating all these events into concrete steps in the customer/user journey such as conversion can be challenging to say the least.
The convoluted nature of modern analytics can also create roadblocks when trying to effectively monetise your business and optimise costs/service levels. Even meeting simple KPIs can be extremely time consuming, such as understanding how content consumption has evolved over time, the weekly average share of subscriptions between digital and physical, or purchasing patterns across all your brands.
If you want to succeed in the highly competitive market, it is therefore essential to rely on a single, unified data model.
Implementing user-centric analytics
There are a multitude of ways that companies can organise their data. However, it’s critical to have a central analytics strategy regardless of the governance model applied within the company – and ideally a team that can build value from the core of business, then offer it to the rest of the organisation.
Becoming user-centric is challenging due the range of platforms and devices, the need for specific teams and analytics tools, as well as the difficulty of centralising the structure of the data.
Creating a unified ‘hybrid’ digital analytics architecture
Today’s analytics market can be boiled down to three approaches – each with their advantages and drawbacks:
Marketing analytics
These are designed to meet specific digital marketing issues such traffic acquisition, monetisation, and include numerous metrics and specific analyses. They provide useful Out of the Box (OOTB) & marketing-specific reports but can quickly reach their limits when it comes to analysing very specific company concepts. They also aren’t flexible enough for more advanced users – you can’t go the extra mile.
Product analytics
These have a high degree of flexibility, allowing them to measure interactions that are very specific to the development of a product or service – a strong selling point for Product Managers, Product Owners etc. Their drawback is that the flexibility can make them complex when it comes to having reliable and exhaustive data – as very few analyses are ready to use. This means they fall short in terms of data democratisation and spreading data across the organisation.
In-house analytics
Tools developed in-house can offer end-to-end customisation capabilities while offering configurable computing power. But beyond the exceptional skills required to successfully complete in-house projects, there are numerous functional trade-offs such as development costs, risks, technical debt, maintenance, etc. The Total Cost of Ownership can therefore skyrocket over time in the race to stay ahead of the competition.
To ensure that both your analytics and your business are user-centric, it’s vital combine the best elements of these approaches.
Firstly, you need to be able to effectively log events and their properties, i.e. understand exactly what actions the user has taken – page loads, conversion etc. Then it’s necessary to make sure your data is being ingested through a structured and consistent pipeline – from collection, to cleaning, storage and activation. Finally, it’s essential to have users at the centre – avoiding on overflow of basic events that need to be cleaned up.
Aligning the terminology
A major part of streamlining your analytics across all touchpoints is to align the naming of all the relevant actions. This could mean renaming all your events as page_load’s or simply interactions. This also applies to properties which often have different denominations that describe the same thing, e.g. visitor_id vs. user_id. The key is to move as quickly as possible to understanding if the user has converted or not and why.
Storing the events
By using separate tools for your different types of analytics, e.g. web and app, you’re multiplying the number of tables on which the data is stored. This has serious ramifications in terms of the effective use of time, accuracy of the results and effectively sharing the analysis with the relevant parties. By storing all your data in a single model, you link everything together and drastically simplify the analysis of the customer journey.
Simplifying the data pipeline
User-centricity means effective data collection, storage and activation. The data needs to be collected across all user touchpoints. It then needs to be stored cleanly and neatly. It can then be used to provide decision-ready reporting to users across your organisation, or by the data science team to build recommendation engines etc.
Setting up a Unified Data Model
By mapping out the KPIs, you can align your teams around the same analytics objectives. It’s then necessary to design your Data model by identifying the main events and associated properties that you’ll need to measure. Next is gathering interest from key stakeholders and demonstrating how they can leverage value from the data.
Finally is making sure you have the optimal digital analytics tools to adapt to your strategy. This means choosing a solution that enables you to sustain the effectiveness of your data model over the long term.
Combining analytics approaches
AT Internet’s new Analytics Suite has a range of tools that respond to the needs of Marketing and Product analytics.
This includes rapid ‘Marketing analytics-style’ data collection, with over 400 standard properties. As soon as the data is collected, the data model is ready to process and store it. It can then be used in a range of reporting and data mining user interfaces.
It’s also possible to add up to 1000 custom properties and store them in the data model – all of which can be managed/viewed transparently.
To go even further, such as importing CRM or customer platform data and enriching it, you can incorporate an OOTB approach, as with Product analytics. This means that you keep a unified data model and maintain the flexibility of using multiple tools.
The Analytics Suite also allows you to benefit from rapid activation tools that provide rapid dashboarding & reporting, as well as data mining. Our data model can be directly accessed to feed tools further down the pipeline. In short, you never have to extract the data and re-work it, everything is enriched in the same platform – from OOTB reports, to a high level of customisation, an entirely user centric approach, and advanced flexibility.
User Centric analytics
AT Internet’s range of tools are perfectly suited to analyse complex non-linear user journeys across a range of platforms and devices.
By taking a user centric approach on all levels – monitoring a series of events that cover all user interactions with a brand – we make all of the events user centric. All events collected in the platform are referenced under the same user_id, whether they’re from website visits, apps or audio/visual etc. This makes it fast and simple to find key information such as the number of visitors, visits or the time spent on different elements.
We also enrich the user-centric data to give you far more perspective and more capacity to personalise the experience. This means it’s simple to add a range of elements to your analysis such as different age groups, behaviours, levels of maturity, as well as to subsequent data science projects.
While Analytics Suite’s flexible data pipeline and centralised data model are built from the ground up, simple to put in place, and designed to meet all your relevant needs.
How a data model strategy will boost your business value
With the new Analytics Suite, AT Internet makes it simple to meet your main KPIs. To discover how your content consumption has evolved over the last year, we allow you to track the entirety of the cross-device user journey with a single tool, use the same vocabulary throughout the analysis, and carry out the same calculations.
With our Axon module, you can carry out advanced anomaly prediction powered by machine learning. This tool allows you monitor any drop in traffic based on your normal rates of consumption and lets you set up automatic alerts to notify the relevant Sales teams. They can then put in place the relevant retention steps to reduce user churn.
We also cover the issue of data governance – which is not only about quality but having adequate data privacy protection. AT Internet is committed to protecting user privacy in full compliance with global regulations. The Analytics Suite categorises the purpose for the collected data, as well as the legal basis for the collection. This involves clearly listing user consent/opt-in in relation to the specific country regulation involved – which in turn minimises the legal risk for companies and ensures that collection is ethical and in line with the respect for user rights.
Finally is the key importance of data democratisation – i.e. ensuring that decision-ready data is accessible to everyone in the organisation. AT Internet’s dashboarding and automatic reporting tools are integrated as standard elements in the data model and are fully turnkey from the outset.
Keen to find out more? Listen again to Declan’s DAA webinar and request a free demo today.
