Using the web to build next generation customer profiles
Trends in digital marketing, opportunities and risks in data | Keynote lecture
These are the references from a strategy discussion with data and direct marketers at the annual congress in the Direct Marketing Association in the UK. The DMA asked us to examine five areas and conclude with practical tips for the questions marketing directors should ask themselves and their teams to ensure digital marketing is being maximized and data protection risks minimized.
- How to use split-run testing to rebuild websites and web marketing
- Decoding online behaviour to create next generation customer data profiles
- How to select that next best offer
- Essential strategies for today’s tough markets
- How to build smarter learning throughout your organisation
The Digital Strategy Consulting group has been a reader and supporter of the ICO’s work in the drafting of the new guidelines on personal information. Our team have worked as government policy advisors on data protection for the UK and European government, chaired the ad industry’s online behavioural advertising taskforces for many years and chaired the internet advertising industries standards group across Europe for 5 years and in the UK for 10 years. While we spend much of our time coaching marketing teams from leading brands about how to unlock the impact of digital marketing, we remain concerned that many organizations do not protect their customers’ data in the appropriate ways, creating risks for their brands and their teams.
Marketing optimisation workshops are available from the Digital Strategy team in the UK, Europe and North America.
To talk more about these ideas, developing products like this, changing your business, or coaching your management teams contact Digital's Chief Executive Danny Meadows-Klue -
Get more research from Digital, straight to your inbox
Fortnightly research bulletins - for free. Simply tell us the topics you're interested in...
Here's an example of one of our special reports about online advertising...
Direct Marketing 2.0 – You are what you click
The pace of change in marketing is catching out teams and brands alike. It’s easy to get digital marketing 80% right and still be 100% wrong. The ROI can be as brutal for failure as it can be outstanding for success. Direct marketing disciplines form the foundation for many of today’s digital skills, but there is a change in scale and scope of direct marketing as we move into an era of Direct Marketing 2.0. This briefing helps marketing directors healthcheck some of their ambitions by distilling the insights and experiences of from coaching global brands and developing digital strategies for firms across many sectors, both large and small, and in many national markets. Further resources are available.
1. How to use split-run testing to rebuild websites and web marketing
2. Decoding online behaviour to create next generation customer data profiles
3. How to select that next best offer
4. Essential strategies for today’s tough markets
5. How to build smarter learning throughout your organisation
It’s no secret marketing is changing, and it’s changing faster than most marketers and organisations can keep abreast of. But the pace of change isn’t slowing, and the space where technology, marketing and data converge is throwing up structural and organisational challenges for corporates and business alike. While marketing has always been based on customer insight, that insight has now changed in both scale and form. The language and processes many companies have in place to manage this is stressed, and marketing waste is rising.
These notes are a summary of some of the workshops run over the last eighteen months with national brands to summarise some of the challenges the strategists at the Digital Strategy practice have seen, alongside trainers on the Direct Marketing Academy at the Digital Training Academy.
The story begins with the familiar direct marketing ground of split-run testing and builds out to uncover the complex data journeys people leave and how these can be used to create perfectly targeted webcentric direct marketing; Direct Marketing 2.0.
Humble origins: A/B split run testing
The direct marketing industry grew up on split-run testing, taking two or more versions of an advert and seeing which performed best before refining the campaign. In the 1970s this put the science into direct marketing and created crescendo of accountability and precision the industry became synonymous with throughout the 80s.
The models were simple, the benefits immediate, and the resulting ROI fuelled the establishment of the direct marketing discipline. In publishing the variables direct marketers gravitated to for testing would typically be the creative execution, the copyline within creative execution, a range of promotional offers and the media placements themselves. In the letterbox it was all about segmentation as ever-smarter targeting unfolded, as well as ever more creative energy going into the mailing pieces. The rise of database management tools and providers enabled this to grow fast, and as it mainstreamed it changed the nature of marketing.
Now this process is happening again.
Traditional direct marketing reached a glass ceiling. Energy poured into segmentation and response management, but the advertising industry reached a limit in what could be sensibly tested. Split-run tests are sure to build ROI, but do the maths and it’s easy to see the problem:
• 5 offers: 10%, 15%, 20%, Kids go free, Free give-away
• 20 media placements: Magazines, newspapers
• 4 creative executions: Aimed to test different segments
• 5 copylines within the creative execution
Add them together and you have 2000 permutations of messaging. The theory is great, but the scale unworkable and the resulting data sets increasingly wobbly because of sample size.
On the web, things are different
The data trails people leave on the web are near perfect. Marketers may rarely have access to all but a few of them, however that can easily be enough to unlock a step-change in thinking and performance. In an environment where every action is recorded, those records can be mapped to give meaning.
In web advertising the click rates and mouse-over events show where engagement has sparked action. They can be directly correlated to promotional activity that shows what works and what doesn’t. In email, simple open rates and forwarding have enabled the offline postal approach to leap forwards, giving brands deep and immediate feedback on exactly what is and isn’t working.
For people who explicitly interact with web forms, shopping carts, games, and surveys there’s a richer and deeper data set, and as almost everything on the web is time-stamped, the data sets are automatically augmented with additional value.
Savvy marketers will push further and look at conversion rates after the click, all the way through repeat sales to customer lifetime value. The knowledge and insights can revolutionise a business.
AB split-run tests have been replaced by A-Z split-run testing
This is direct marketing; unleashed. It can be embedded into the standard operational procedures of marketing departments and applied to anything from an email to web banner, Tweets to social media fan posts.
Analytics is the secret and digitally savvy marketers get ever-closer to their analytics dashboards and screen overlays to understand in the deepest sense exactly what works best and how. And don’t misread this analytics as the bean-counters moving in to own the marketing department. This is the type of smart, marketing insights that transform brands and become a game-changer for marketing teams.
While relationship marketing teams working on retention have become masters at this, acquisition marketers often fail to venture beyond search engine marketing in this level of analysis and optimisation. Yet the transformative effects on acquisition can be even greater. Aligning the process of split-run testing with website design can unlock a step-change in the volume and value of customers from any website.
Customer data journeys - the metrics we can see
What’s different about digital marketing isn’t just the scale of the analytics, it’s the granularity and form as well, and how it can be harnessed into decision-making engines. Websites can provide the complete picture of a customer’s behaviour saying as much about what they do not respond to as what they do. In the right hands, this unlocks the most granular of insights that can build near-perfect data profiles of how customers behave, getting marketers over the hurdle of claimed behaviour and into the space of real, observed behaviour.
Selecting the right next best offer in online retail
Finding the next best offer for a customer has been the retailer’s challenge for as long as shops have had shelves. Now those shelves have moved online, the challenge is different and the customer data journey holds the answer. Firms getting this right are changing their product mix, ranges and profitability. Firms that are not getting this are simply losing sales and share of wallet.
The scale of benefit is immediately clear when online behaviour is mapped to individuals – either anonymous or identified. A new type of relevancy is unlocked, and one that delivers exactly what customers want to see as well as what brands want to share.
The richer the online customer’s journey, the deeper the insights and the more relevant the messaging. While the applications of this feedback process will vary between industries, the outcomes are broadly similar. Reduced volumes of bland vanilla communication and greater relevancy and cut through with the obvious link to sales. Every web store and commerce platform is trying to follow in Amazon’s journey to be able to deliver the message: ‘people who liked this, also liked this’.
That simple phrase sums up where direct marketing gets to once fused with the best of relationship marketing, the power of digitally driven profiling, and the intelligence of algorithms running across massive data sets of complex multi-dimensional consumer behaviour. Now the technology and data integration has been opened up to a massive swathe of mid tier firms, this type of sophisticated behaviourally driven messaging is finally mainstreaming.
Advertising networks using behavioural profiling probably account for 15-20% of today’s web advertising in the UK, with the trend continuing to rise steeply, but the agency community taking longer than expected to re-orientate their buying models. For a technology that is a decade old, it has mainstreamed much slower than many digital marketing techniques such as search or social media, but this journey is only part complete and the behavioural sector will swell to take around a third of the online ad market over the next few years. It will become the norm rather than the exception.
What’s distinct about online behavioural advertising (OBA) is the lack of real world identifiers. While debate will intensify about whether the profiles in OBA are personal or not, rarely is there a real world identifier beyond a transient cookie lodged in a browser’s software. Old school direct marketers still stumble on the lack of traditional demographics and data metrics, but when a customer’s interests are the only identifiers, brands need to be comfortable accepting a new paradigm in direct marketing.
Customer footprints fuel profiling which fuels targeting
These new models for gaining customer insight are unfolding fast. Marketing hinges on customer insight, and now that insight is broadening and deepening, the scope of relevancy can leap forwards again. Building deep customer profiles from real world observation is at the heart of the Direct Marketing 2.0 model.
There are four steps in the linear process of taking customer behaviour and delivering back the offers that will work best, however it is a process that can easily fail if the right information and safeguards are not in place.
1. Observed behaviour
2. Data profile
3. Predictive modelling
4. Targeted marketing
Questions marketers should focus on
• Which behaviours do you track?
• Where can you look?
• How comprehensive is the picture?
• What timeframe does it cover?
• Is it personal data?
• Is it sensitive data?
• What do data protection laws allow you to do?
• Tracking the wrong behaviours
• Failing to look in the right places
• Failing to recognise if you only see part of the picture
• Misreading timelines in customer journeys
• Confusing anonymous subjects with identifiable subjects
• Failing data protection compliancy
Data profile building
Questions marketers should focus on
• How broad are the choices for structuring or buying profiles?
• How deep is the data in a profile?
• How does it map to your existing segmentation?
• Is the data identifiable to the person / browser?
• Has the purpose of the data changed?
• What can be stored and processed within privacy guidelines?
• Which records in the data file are suppressed, and are the opt-out techniques from various channels feeding the suppression file effectively?
• Failing to categorize effectively in breadth or granularity
• Poor mapping to existing customer segmentation models
• Failing data protection compliancy
• Failing to manage suppression in complex digital marketing environments with third parties
Questions marketers should focus on
• How strong are the algorithms?
• How does the algorithm learn over time?
• How do you improve data quality for the algorithm?
• What other data sources / overlays can be included?
• Typical ‘rubbish-in, rubbish-out’ risks in any modelling; particularly significant in areas poorly understood by marketing stakeholders
• Over-reliance on the algorithms without the confidence to challenge
• Failing in the algorithm to learn over time
• Failing to remove poor data from the profile
• Failing to overlay the right data in the right way from other sources
Targeted and retargeting of marketing
Questions marketers should focus on
• Which retargeting techniques are available?
• Which retargeting techniques are significantly under-used?
• What types of targeting can be delivered within privacy guidelines?
• Ignoring many retargeting techniques, both online and offline
• Failing to unlock the potential of retargeting systems already in place
• Failing to recognize changes of use in data
Digital can be a brutally unforgiving space. There’s an interdependency between the links in a digital marketing chain whereby a small failing at a single step destroys the whole value of the campaign. Online retailers with poor shopping carts, brand sites with self-indulgent home pages, search campaigns that selected the wrong keyphrases, brilliantly written emails sent to the wrong audiences: it’s easy to by 80% right but to be 100% wrong in the outputs.
That’s why Digital Strategy was asked by several brands to develop a checklist of questions to help identify weaknesses in webcentric direct marketing. These came from running digital marketing effectiveness audits to identify both what was – and was not – working in multi million pound web marketing strategies.
Over the last ten years as behavioural targeting and profiling has moved from the lab to the mainstream, we’ve seen companies consistently fall over at the same points so crafted this checklist as a way of helping marketing directors maintain their ROI and privacy officers maintain compliancy.
Structural risks most organizations face: knowledge and skills are the cause
The issues of using the web to build next generation customer profiles play out in a space where the disciplines of marketing, data protection and IT collide. Three disciplines with fundamentally different views of the world and different languages through which they articulate them. Only by developing the knowledge, insights and capabilities of these cross functional teams to reach a space where they share an understanding of the challenges can organizations effectively respond. Without this, at best marketers will waste energy creating campaigns that can never be deployed, technologists will create architectures that can never be properly populated, and lawyers will err on the side of caution polarizing the divisions between teams in response to minimizing risk to their firms.
Deeper than knowledge and skills lie the cultural barriers to change. Sharing knowledge is part of the process of changing culture, but for large organizations to transform it takes energy and commitment of their teams and leaders. All too often digital marketing and digital product innovation are left on the fringes of senior management attention, yet the history of the Amazons and Googles is a history of companies whose leaders saw competitive edge in how commerce was changing and created cultures and agile structures that could respond to the massive opportunities that are unfolding. Leaders need to lead, and with digital channels forming the battleground for competitiveness, leadership teams need to ensure their attention is focused heavily on the digital communications and platforms their organizations have developed.
Given that for every one firm’s opportunity, there is another’s casualty, here are a few simple steps to bringing that test and learn culture of direct marketing up to the right level inside any organization not already deep on their own journey…
Building smarter organizations
1. Start from clear business objectives
2. Decide how you will measure success (ROI)
3. Decide how you will test
4. Test, test and test again
5. Review results
6. Apply the improvements
7. Share the knowledge
…and do this across your organization
In a world where you are what you click, where competitive differentiation is in the application of customer insight, and where most digital marketing channels remain wasted most of the time, there is everything to gain. These are issues that should be at the heart of the corporate agenda and the greater the energy leadership teams give this space, the greater the opportunity to unlock step-changes in the competitiveness of their businesses. We are moving towards a world of fewer brands, more empowered customers and intolerance of the irrelevant. Direct Marketing 2.0 has a place at the heart of the competitive agenda and for the firms that get it right the ROI will destroy the margins of their competitors. With the right skills and people, in the digital space, everything is to play for.
Resources that support the Digital Change Management Programme at Digital Strategy
• Digital marketing effectiveness audits
• Digital customer journey audits
• Digital strategy healthchecks
• Digital strategy capability build
• Digital leadership coaching
• Digital risk analysis
For access to additional resources contact the author Danny@DigitalStrategyConsulting.com