Digital consulting companies are often asked to deliver a “data strategy”. A perfectly normal request in a world where data is called “the new oil”.

I am going to give you a non-technology answer of how I approach it if I were to put myself in the shoes of a CEO, CFO, CMO or anyone that needs to use data to execute their KPI’s.

If the source of revenue of any business is to create something (a value proposition for a customer) and to sell it to them at a profit, all data needs to in some way work towards the achievement of that objective.

(Of course, data is also a product a company can sell to third parties in its own right. In my view, that is a separate discussion, as it is then about data being a value proposition with revenue potential.)

We all use data, every day.

Companies use financial data to measure performance. CMO’s and brand managers use data to measure sales and brand performance. Companies use tools like JD Power’s to measure customer satisfaction. Companies “cry” over high churn figures. Companies buy channels on the basis of CPA. Google lives by algorithms. Programmatic buying is based upon algorithms. Consumers compare prices. Market analysts use data to assess share performance. Partners interpret non-verbal signals from other partners as meaningful – or not.

So intuitively, our lives are based upon data.

Data reduces decision-making uncertainty, never by 100%, but some data is often better than no data at all. The wrong data can off course be disastrous.

So the concept of data is widely used, yet therein lies the confusion.

All data discussions need to start with a business and/or marketing objective. Do we want to sell more? Sell for a better margin? Fight competition? Change perceptions? Grow market share?

A data strategy needs an imperative, it is not an end in itself. Likewise, customer centricity is pursued for a purpose. Without purpose, a data strategy will be devised in a void.

This can be as simple as a workshop deciding what data drives the business and assessing the current status of such data. The outcome is to decide what to do about it. What will the data be used for? If it is about becoming more “customer-centric,” what does that mean within a given organisation (when do customers see a brand as being customer-centric)? What priorities will determine that?

For most companies, it starts with simply talking to their customers in a more relevant and personal way. Regardless of the company, customers do not like to feel like numbers. In the very least, they expect courtesy from brands they support.

Another measure is to assess how the data categories, insights, processes and data flows, support customer centricity. It is arguable that a major focus of customer centricity has to be about how the data strategy supports that. Whereas most historic data in companies were generated to support company processes (financial results, unit sales, RSP), it means data needs to be re-designed and re-interpreted to support customer centricity. This alone is a major shift in the way companies think and work.

An over-riding issue is the metric of the data: it is about customers (the same person that bought a few Ford’s) or about products (the number of a certain model sold) or services? This is a major issue, as most companies measure sales/product or service, instead of by customer (i.e., lifetime value or CLV).

In future, data will drive two things: 1. The deep personification of customer experiences leading to greater brand loyalty (if still possible) and sales (possible) and 2. Deeper connected-ness across devices, products, services and ultimately Eco-systems. Whilst both are important, most companies will take years to get the first one right, let alone the second one.

Unless the first one is right, the second one is meaningless. The context of connected-ness is customer insight: what value will it create for them?

The companies that can progressively drive both, will win. Even though Apple is deficient in many ways in consumer engagement (they actually do very little about one-on-one marketing), the customer experience (product hardware design and software interface functionality, packaging, stores “look and feel”, staff attitude, advertising and launch events) and connected Eco-system (iTunes, Apple TV, third-party applications, iTunes music remastering) of Apple has already provided them with a deep source of competitive advantage. Now relevance gets connected seamlessly and incrementally, which creates a serious barrier-to-exit for customers. Unless another brand can compete at that level, it simply cannot compete. Even the most beautiful competitive handset cannot complete within the connected Eco-system. Yet, that is still where most competitors see the battleground!

The depth of data banks, Facebook, LinkedIn, TripAdvisor and Amazon have, enables very deep levels of incremental revenue for these companies. They are at least able to compete at the same levels and may do so in time. That is unless small upstart companies do it right from the word go.

Data is the foundation of it all. That is why data is the new oil. Not just because it is good to be called by your first name. That is not a competitive advantage, it is shocking if it does not anyway happen. Or when the decision journey online is fragmented, it should be sequential and aligned (obviously device and identity complexities make this difficult at times).

Some “rules” may bring order to chaos.

I would like to differentiate (below) between three areas of importance in data, the first one governance issues, the second one classification issues and the third one structural/procedural/flow issues.

Governance – like with all company assets, data needs regulation

Relevance – focus on data that drives behaviour

For data to be of any use, it must be relevant (data that actually matters in driving consumer decisions). If we compare what has been published about data for generations, it is not the volume that matters, it will be a few pieces of data that explain behaviour. In my experience, there are very few really important data types that drive consumer behaviour and brand preference.

Unless a company knows that, it will waste a lot of time curating “noise”.

Privacy – protect data, customers and remain legal

Data privacy is a tightrope that will get more complex and political. Consumers want brands to know them better – but not too well. Whilst they love being able to pay with one click on Amazon, they may rebel if they get “geo-located” when caught in a seedy part of town!

(Because it differs between countries and even within countries and within cultures, this requires deep legislative insight. This is also one of the greatest dangers of omnipresent data, piracy, identity theft and one that is very hard to manage.)

Accuracy – data must be right, maybe the biggest challenge of all

This goes without saying, if data is not accurate, it is a total waste of resources and can be downright dangerous.

Data is a process activity – you can create the framework but the job is never done. Quality is an omnipresent issue. As data is now largely automated, it is much easier to manage than before.

However, as people move homes, jobs, countries, their identities proliferate and replicate. Data quality if really difficult to manage and requires a deep company commitment to the structures to do so.

Curation – someone needs to look after it and be the “gatekeeper”

Who ensures the data remains accessible, accurate and properly governed? In most companies, data exist in different parts of the organisation, yet the company must have access to it when it wants to use it.

Policies are required to regulate usage. Yet, policies can result in red-tape.

It is also important to ensure data is not abused. Unless departments coordinate, some individuals can receive massive amounts of spam (my record is sending 23 pieces of communication to one (very important) customer of a bank on one day, not a statistic I am proud of).

Classification – the types of data that ensures accuracy

Sources of business – who are clients and potential clients?

Their numbers, value and profiles for the purposes of retention, churn management and acquisition.

Unless I know, I can waste a large amount of money on buying data sources without it being reflective of the most opportune sources of revenue. Today, a competitive disadvantage lies is talking nonsense to the wrong customers.

Profiles – what kinds of people are they and what do they like?

At a very practical level, it is about addressing them by name. Unless this is right, the rest is as they say, “academic.”

If I want to be really serious, what are their attitudes, profiles, the words they use, their preferred brands and their friends? Words like personas and segments are used here. Deeper understanding of people has always driven marketing content, the more we personalise it, the more true this will be as well as more complex. Yet, most of this exist online today!

Decision journeys and customer needs – how do they decide what they need and how do they embark upon that journey?

What customers need and want determines the message content we convey to them. Unless we know that we cannot logically connect their needs to our capabilities and our brand value proposition. That is when you receive the same mail, with the same offer, time and time again from the same brand, or you get re-targeted with the same product offer through Amazon, even if you have already bought it from them the week or even day before!

I find it hard to understand if a retailer tries to sell me wedding dresses online.

Once I know this, I can design the user experiences (UX) across and between channels to logically follow how they buy. Unless I know this, I run the risk of seeing UX as only a channel issue, not a multi-channel issue.

Content – what types of content do they react to?

The way customers react to content is key. Ultimately, marketing is about impactful content that drives behaviour, hence deep insight into content is very important. Content is the single most powerful variable in brand engagement: do it right and you win. Sadly, it is mostly done badly.

Most of these factors will be qualitative. The words people use, the issues of importance to them, the brands they like.

Channels used for experiences and engagement – what channels do they use?

I need to know what channels my target markets use. How, when, where and why? Using what devices? Having what user experience breakdowns?

How can I possibly manage brand engagement and experiences if I don’t know where to find my customers going about their business of life? We need to stalk customers, they have no desire to stalk us.

Networks – often the network of current customers are the best source of new customers

Very under-used, but with social media data becoming far deeper, a major potential area as often current users are the greatest predictor of future users.

So, what important networks do people belong to? How can those networks be used? In B2B, this is key. We have animals lovers, nature lovers, art lovers, DIY lovers, pianists…

Other than the insights gained from asking questions and posting endorsements, networks are now more powerful than traditional marketing channels.

Behaviour – what do they actually do?

What do they do and buy? How can I move beyond simple predictive analytics of the last ten purchases (Amazon) to deeper insight?

What can I learn from UX data within and across devices and applications? What can I learn from Google Analytics?

What products and services do they use? What brands? What do they do and say online? What channels do they use? What factors are predictive of future behaviour.

The scary thing is that all of these areas are accessible now!

Locality – where do they go, buy?

Where do they do what? What do they do at certain places?

How can locality be combined with other data points?

Devices – how do the right target audience use devices and how are they combined?

What devices do they use for what activities? Why do they favour one device over another for a given product or service category?

How can one device be used to amplify another? Device cross-usage is very powerful, An sms in a store can switch a brand decision, fast.

Time – when?

When do they do what they do? Why at that point?

How can my brand be in the right place at the right time? How can timing be used opportunistically to “layer” other channels used by them?

Things/products/services/connected devices – eventually, how can the IoT be used well?

What data is required to enable “connected-ness” across applications? In other words, how can data capability across the ever increasing network of devices (even humans, plants and animals) that can carry data, be leveraged by the brand?

Some brands, like Tesla and Uber is designed around a central user experience, most traditional companies need to first deconstruct those. Yet, at the core of it all lies the individual and his/ her needs. Human centred design enables responsive and consumer centric organisations.

Unless that is known, strong, collaborative Eco-systems cannot be developed.

Structural – how does data flow and support customer centricity across company operations? How does it create internal controls over customer delivery?

Structurally, a truly consumer centric brand is designed around the consumer. Data is the “glue” that holds it together and determines the flow within the organisation. Yet, then it must be set-up that way.

Unless the data is accessible and structured to enable its use, it cannot be used.

We need to understand how data flows into, throughout and out of the organisation. This means organisational behaviour and process management are central to data.

A company is an open system that translates raw materials (people, money, material, infrastructure) into internal processes (production, marketing, logistics) with given outcomes (sales, RSP, market share, share price). Being open to change enables adaptation and innovation.

Structure is the foundation of organisational efficiency. It is also a constraint when it becomes rigid. A stagnant system implodes and dies. An open system breathes and grows. Like with people, data evolves.

What processes, structures, systems are in place to enable that.

Where are the fault-lines? Where is data duplicated? Where is data redundant? Why must the customer supply the same information again and again? Can data be made accessible (can the Call Centre attendant see the customer records whilst they are talking to the customer?) Can the bank access the client profitability across product usage when a rate decision is being made? Can the hotel access what pillows the frequent traveller prefers? What happens if there are systems breakdowns? Where are there obstructions in the data flows? Where is technology incompatible?

What is the nature of the data (numeric, qualitative)? How are ratings data reconciled with volume data?

What is its periodicity? How can quarterly data connect to daily data?

What is the original purpose of the data (is it sales data, monetary value data, volume of transaction data, yield/customer data, value of transaction data, acquisition data, retention data, qualitative (reasons why churn happens) data, Call Centre data, external market data (makes size and growth), external brand data (brand health and awareness), NPS or customer satisfaction data, demographic or other profile data, product volume data, customer data? How is that translated into consumer data that drives customer centricity?

To conclude

In my view, the purpose/ objective of a data strategy is key. It is the key performance KPI.

Why do you need a data strategy? What really drives the business?

Most important. know and talk to your customers personally. That is the basis for everything we do in marketing, from addressing them right, to great creative content to delivering good customer service.

Then, decide what data drives the business, prioritise and decide what process is required to put a data strategy in place.

Then review governance, classification and structural issues. Without this, you will re-do the same things over and over again.

As with everything, test-and-learn. Most companies start from a low base of customer data, they can only improve!

And even if the Internet of Things is amazing, first get the basics right before you start creating connected Eco-systems.

The major issue is, the technology now enables data connectivity, disruptive brands demonstrate the art of the possible, so companies no longer have any place to hide if they do not go about data strategy in the right way. Every day lost, or doing a “panel-beating” job that is inopportune, is a lost opportunity.

To get data “right” may be hard, but companies do not have a choice.