As the world moves further into the digital era fuelled by the pandemic, consumers are more comfortable than ever in using multiple digital mediums to interact with a brand. 55% of UK adults are now comfortable interacting with brands in a virtual way as long as the experience is good. This makes it more important than ever to understand customer needs and their behaviours.
With the increase in omnichannel customer experiences comes an increase in the number of touchpoints and data collected. Each interaction or transaction creates a chunk of data that is potentially isolated from data captured on another platform. What this leaves is a complex web of unlinked data from several systems.
Where this becomes a major issue is when teams work in siloes, which is commonplace in many organisations. With individual teams each owning a different part of the customer journey, and subsequently the relevant data capture platform, it’s easy to see how data can become disjointed and different strategic approaches can be created depending on the data a team has access to.
The result? Each team only has a small window into overall business activity, most likely missing the big picture and making decisions based on partial data.
Having a unified data source solves siloed working and multiple points of data capture. It joins everything together so all teams can confidently work from the same data source without the worry of reliability or missing data.
A Unified Data Source for all business data
In the most fundamental terms, a unified data source is a single central location where all data collected across the organisation is held. It is a combination of many different fragmented sources matched together into a common schema. Sources can be anything from marketing tools, ad platforms, CRM, legacy systems, social media, customer service – anything that collects and stores data.
The creation of a unified data source can be streamlined through the use of a Customer Data Platform (CDP). A CDP generally just matches customer data; the best CDPs (like The Data Refinery) have additional functionality with advanced matching engines that go beyond just customer data to combine all business data into a standardised schema, eliminating the issue of disparate systems.
Benefits of a Unified Data Source
Having a unified data source that anyone in the organisation can access democratises data and enables better decision making, revolutionising how a company uses its data to maximise value.
Remove team silos
One of the biggest challenges organisations face is the lack of cohesion between its teams. Having each team working from a different data source often leads to conflicting views of the same thing – solid decisions can’t be made when the finance team are reporting on a different data set to what the marketing team are building campaigns from.
By having a unified data source, everything is in one place. Anyone in the organisation can access the same information without having to pull together data from different systems. Senior management can track all their KPIs from the same data from which marketing put together acquisition campaigns.
This is often referred to as a “single source of truth” and can enable organisations to operate more efficiently while improving accuracy as a result of more reliable data.
Eliminate data flow dependencies
Many organisations have a range of processes or technical implementations that get them some of the way to having joined up data to overcome the issue of data silos, but these often become obsolete with changing data capture and business requirements. As dependencies grow and the business scales, these implementations can become a blocker and require dedicated manual time to maintain without any assurances that errors won’t slip through the net.
Having a unified data source removes the need for much of these bespoke processes and technical implementations. Removing steps from a data flow streamlines operations and eliminates dependencies on a process to output a result. With everything running from the same data source without independently managed connections, there is less maintenance required and reduces the potential for errors in the data while speeding up time to action.
Know your customers
Operating across several channels maximises the chances of engaging with a customer, but it also creates multiple views of that individual. Depending on what data is collected in each channel, only part of the customer’s true behaviour is captured. Only when it’s combined do you really know who your customer is.
A unified data source can spin out to create a single customer view brings together all touchpoints including transactions across multiple channels, interactions with different social platforms, and support engagements to paint an accurate picture of each customer’s unique behavioural traits.
To add to first-party data, external data can be added to customer profiles to make inferences about their demographics, life stage and disposable income – all of which can help build personalised messages that are delivered at the right time.
As touched upon in the data siloes section, having different teams working from different data leads to misaligned outputs – it’s the same when it comes to analytics.
All too often a metric or a graph is taken at face value without any real consideration of where the underlying data came from to produce it. This can quite easily result in the wrong information being shared and decisions made based on poor data, which can end up costing severely in both time and money.
By having all data in one place, everyone in the organisation can produce metrics that are directly comparable to each other regardless of who created them. This boosts confidence in the data and the decisions that come off the back of them.
Challenges of implementing a Unified Data Source
Unfortunately creating a unified data source isn’t as straightforward as it sounds. This can be down to a few different factors such as poor data quality, older legacy systems ingrained in the organisation with poor connectivity or a lack of technical expertise and resources to implement and maintain complex data infrastructure.
Poor data quality
Data quality measures the condition of data to determine whether it is fit for a specific purpose. There are six main principles when managing data:
- Accuracy – the data reflects real-world scenarios
- Completeness – all the required values are present
- Consistency – there is uniformity across the data as it moves across different platforms
- Validity – conforms to business standards at the point of capture
- Uniqueness – no duplications of or overlapping values
- Timeliness – kept up to date and available when required
If a data set is lacking in some of these areas, it becomes less reliable, and the outputs from the data may be out of date or flat out incorrect. Keeping data quality high is a constant process and should always be at the forefront of the priority list as without it, the data could become worthless.
We have a saying at The Data Refinery – “bad data in, bad data out”. The actual saying is a little more colourful, I’ll leave that to your imagination.
Volume of data sources
In this modern world of billion pound start-ups popping up every day, there’s a product for everything. Any business is likely to have a CRM, Google Analytics, Facebook Ads, Google Ads, and a marketing automation tool. That’s 5 sources before we even start to look at ERP systems and advanced targeting platforms.
Luckily most business products live on the cloud with good connectivity through APIs, meaning a connection can be established and maintained with dedicated development time (the bulk of the time spent building new connectors is often understanding the shape of the incoming data) or just a few clicks to connect to pre-configured connectors using the ETL (extract, transform, load) features of a CDP.
Lack of resources
Building a unified data source takes up considerable technical and monetary resources.
From starting at zero, the minimum time to value is at least 6 months depending on the level of complexity and even at that point it’s only really a proof of concept. To get to a position where data is reliably ingested, transformed, matched, and presented into a common schema, a realistic timeframe is more like 18 months. When factoring in the cost of salaries, infrastructure, support and maintenance, the real cost of implementing a unified data source is comfortably in the seven-figure range. Just to build and maintain the data integration pipelines, the cost is an average of £390k ($520k) per year.
Lack of available resources often prevents smaller companies from accessing the level of insights and value in its data putting them on an uneven playing field in comparison to larger organisations.
The Data Refinery as a Unified Data Source
At The Data Refinery we believe every company should be able to access and activate the full value of its data, regardless of technical skills and ability, maximising its opportunities to grow and succeed.
A unified data source is just one of the features that The Data Refinery provides, we have first-hand experience working with our customers to overcome the challenges faced by data silos, along with a host of reports, analytic models and automation flows that are available atop of our unified data Lakehouse.
We take care of the ETL, matching, deduplication, and presentation, freeing up you and your teams time to focus delivering value, rather than spending time wrangling data into the shape you need.
The Data Refinery can be your single source of truth powering your entire organisation with reliable data, at a fraction of the cost and time of a self-build option.
Get in touch and we can show you how we can help you maximise the value of your data.