Has the concept of a single customer view been a seductive mistake for marketers? It is easy to see why having an holistic grasp of all an individual’s transactions and interactions with the business would look useful. Marketing has to be able to influence a wide range of behaviours across a lot of touchpoints. That means any piece of data might become useful at a given point.
The trouble with SCV is that it will typically integrate a large amount of data that is genuinely not needed by marketing. Is it necessary to know a contract number, interest rate or name of the agent who dealt with a call, for example? For other areas of the business, such as operations or finance, this level of granularity matters. To marketing, it may just get in the way.
That is why the idea of the single marketing view might be a more useful concept. A set of data, refreshed regularly from the line of business systems with banded data, value indicators and flags – that looks like a usable resource.
It is also one that can be built using the marketing budget and without involving IT. That is what Grant Baillie, e-commerce customer retention manager at Argos, has been able to create for the business without getting involved in a major technology deployment.
“The way Argos is set up is as a traditional retailer. The business is organised around supporting the stores and has databases for stock, finance, fraud and so on. It is not able to build a single customer view from that,” he says.
When he joined the company six years ago, Argos had been able to create a single view of its customers online using a web analytics data mart with an email tool bolted on. “It was a clumsy solution, but was quite good,” recalls Baillie. He recognised that, while the company was data rich, it was never going to be possible to create an in-house SCV working with the IT department.
The solution was to deploy third parties with specific abilities and deliver data feeds to them. “The data existed and IT can push it out to external sources. That is easier to do than an enterprise-wide SCV,” he says. Ensuring that operating systems remain unaffected by data extracts is critical. “Argos is a multi-channel organisation. If you impact on one channel, it can have a big knock-on effect. As long as you don’t do that, it is ok.”
Baillie now works with Planning-inc to build database hierarchies and run analyses using FastStats, eCircle for its email targeting and broadcasting and Omniture for web analytics. The view of the customer that can be achieved by having those three MSPs linked together is rich and flexible enough to support the needs of the marketing department and beyond.
“With external data services providers, you can send them data in any format and they will tidy it up. That is more achievable than getting the business internally to do it,” says Baillie. One of the sticking points for many SCV projects has been how to align customer records, agreeing on the definition of the data master and defining ownership.
This way, Baillie gets the operational support required without the pain of data integration. It also means his department is able to help out other functions, such as store forecasting. “They use SPSS for geographic location modelling, but they have limited data on footfall since buyers usually don’t identify themselves at point of sale. Our system allows them to track product reservations, postcodes, previous transactions to build into their forecasts,” he says.
Data is also shared with sister company Homebase and is beginning to be used by other core business functions. The key has been to keep a flexible approach without being married to a specific technology platform. “The model works and is successful,” says Baillie. “It is like Blu-Tack – it moulds to anything you need.”
Whenever marketing has been drawn into working with an enterprise SCV, it is often the issue of flexibility which has led to frustration. Decisions about the data model, links, update cycles and views created is often taken by operational functions or IT. The outcome can be a highly granular view of the customer which marketing finds difficult to work with.
Ask a data analyst what data they want to work with and they will ask for as much as is available. Yet it is often better to have a limited range and look for actionable insights than to use everything and hope something good falls out.
“The value of a single marketing view comes from 5 per cent of data,” says James Horsburgh, marketing database manager at Allianz. “The important thing is how you look at all the pieces of data together, rather than their breadth.”
The insurance provider is in the middle of building an enterprise data warehouse to create the single view of the truth about customers. Alongside this, however, Horsburgh was brought in to build a relational database for the marketing function which would meet its specific needs. “We’re looking at 15 data feeds from the business and selecting 200 to 300 data items from them,” he says.
One of the reasons why it is hard to drive marketing from the corporate data warehouse is the local requirements within a business which can sometimes conflict. Horsburgh notes that Allianz has corporate objectives, such as being the product leader in speciality insurance lines like pet insurance. The marketing database needs to support prospecting for those lines within the available information.
“If I just said to marketing, ’tell us everything you want’, we would end up with hundreds of thousands of data items. You need a framework,” he argues. “The most important thing is not the fields, it is understanding across all input sources what are the core elements.”
That allows the marketing data mart to use a simple data model that recognises consistent elements within each of the data feeds, with a table typically made up of around 50 elements. “We can easily add in fields as long as they don’t require any redesign of the originating database schema. The key is the definition of the tables,” says Horsburgh.
That is not something you would hear about an enterprise SCV. Mike Hudgell, business development director at Evaxyx, points out that, “IT departments do tend to do things ’properly’ and at large expense and they tend to take a lot of time to do it. That is often incompatible with the pace of the sales and marketing cycle.”
By using either low-cost PC-based technology or an external MSP, marketing has been able to build the view it needs as a standalone operation. Whether this is advisable as a long-term solution, however, is up for debate. Hudgell warns that, “it lulls marketing people into a false sense of security. It is a point solution that is very easy to do for prospect data, but is less easy to do with client data.”
His view is that marketing can get what it wants from an enterprise-wide SCV, but only where the marketing director is a leader in the business and prepared to push for it, as well as be willing to demonstrate the value it will return to the business.
“If the marketing department are laggards, however, and are not perceived as part of the core culture of the business, it will end up building SMV as a fiefdom that will become a new silo. You build a SMV for political reasons, not because of any technical challenges,” says Hudgell.
Much rests on the quality of the relationship between marketing and IT and the ability of marketers to express what they want as a technical specification. This can be one of the biggest challenges because the view required changes dynamically. Even a SMV architected for the needs to two years ago would be unlikely to have foreseen the growing importance of social media, for example.
Conversely, the SCV may just overwhelm marketing by offering too much detail without an easy way to get down to what is important. “It is like people crossing the road,” says Adam Crisp, marketing solutions consultant at Rapp. “You know they will cross as a whole, but you don’t know where any individual in that block is.”
While the enterprise wants to drill down to that individual level, marketing often needs to roll up to a segmented view. That type of clustering may not be easy within the data model and tables of a full-scale SCV at an affordable and rapid rate.
Budget is often the simpler determinant of which approach gets taken. Crisp says: “I have seen a SMV built first as a stepping stone to a SCV and vice-versa. One client went from SCV to SMV having got its data lined up in the wider picture, because it was then able to tailor it. That is the bigger, long-term view.” Smaller organisations may have to approach it the other way around.
That is partly due to the benefis that are likely to result and the way scale will magnify them. To a large company, having an holistic view of the customer base is likely to deliver some significant improvements in key operational areas.
“SCV is really used for looking at risk, forecasting sales, modelling behaviour,” points out Ally Scott, business development manager at Alchemetrics. “It is very much a corporate resource and is a major asset for organisations that undertake them.”
She argues that the views built by SMEs can fairly be called a SCV because they do integrate data at the level of individual customers. It is just that fields will be populated with derived data, rather than the full spectrum of specific information. “That is the trick – think about who the audience is. If it is is ordinary marketers who are not data experts, they need a very different view from statistical analysts,” says Scott.
It is this concept of a single view, rather than a single truth, that should be at the heart of marketing’s data asset. In a sophisticated, well-resourced department, the view used may be the same as the corporate view. But that is no reason for smaller operations to feel they have to see everything, rather than just the things that matter.
Case Study: Single customer view: keeping it clean
Many would agree that keeping existing customers happy and loyal is one of the toughest challenges businesses face during the recession. Holding accurate data will do a lot to help this and allows businesses to communicate relevant information that will maintain interest. However, failure to keep up-to-date records will only result in time wasted on poorly targeted outreach.
Just collating everything you know about a customer is not enough. Customers will only listen if the information they receive is timely and relevant to their preferences. A true single customer view will help make this happen, but can only be achieved when a documented data quality strategy is in place that has suppressed outdated information and removes duplicate records. Whether data has been captured online, face to face or over the phone in call centres, you will need to identify where incorrect or duplicate information exists, and clean your data records accordingly. A single spring clean is not an option, as databases that are left to decay will be subject to contacts that have changed address, passed away or registered with the mailing preference service.
Building a single customer view will not only have great benefits in terms of saving the customer from duplicate mailings, but will also help marketers improve ROI on their marketing campaigns by using accurate insight on where resource should be allocated. A single customer view will also boost a brand’s green credentials, safe in the knowledge that the communication it dispatches will be far more likely to reach a willing recipient rather than be put in the bin before being read.
Experian QAS offers a number of products and services to make sure organisations’ contact data is correct and helps maintains a single customer view. We’re helping our customers suppress unwanted records from their marketing databases and prevent the entry of duplicate records into their database, saving them from having to remove these manually.
Adopting steps such as these will allow your organisation to hold a single customer view based on better managed data across the organisation, refining contact data into real up-sell and cross sell opportunities and keeping prospects’ records accurate.