How to use UX thinking to generate meaningful marketing insights

I’m going to cover how we can use UX and concepts such as service design, data preparation and innovative technologies to achieve digital transformation.

We’re going to answer these questions:

1. How is data impacting us?
2. Should we be completely data driven?
3. How does UX help us?
4. What innovative tech is out there?

But first, let’s take a step back to look at the context around using our data.

Data meltdowns can equal death

When migrating data or developing new platforms, data is your oil. If it doesn’t flow smoothly, you could end up with a really nasty spill. When TSB experienced its data migration and system upgrade disaster, thousands of businesses were unable to pay staff and receive money. And it’s not only large organisations that fail with data. The top reason startups fail is that they don’t understand their market well enough. Lack of the right insights is killing B2B startups. Now, oil has always been a valuable commodity. Ten years ago it was claimed that data is the new oil. Like oil, data has to be refined to produce something meaningful. But the difference here is that data is unlimited. Globally, we are producing more and more data every year. Let’s look at the context of a day in data. In just one day, according to Raconteur:

• 500 million tweets are sent
• 294 billion emails are sent
• 4 petabytes of data are created on Facebook
• 4 terabytes of data are created from each connected car
• 65 billion messages are sent on WhatsApp
• 5 billion searches are made

And you can see from the examples above, data is not just an output, it is the product and service.

Look at Google’s G-Suite as an example. All its products require data to work correctly.

Data as input: Too much data causes analysis paralysis. As businesses we need to take the data that we process and make sense of it. With too much of it we can fall into the trap of analysis paralysis and gain no useful insights.

Data as output:  Poor data design will kill products and services. Going back to the TSB example, data is also a powerful output and is the oil of the UX. If we are careless with how we use data in our products and services, we can sabotage our brands – just look at Cambridge Analytica.

HubBot is a chatbot that acts as a placeholder until a human hops in. When a conversation is initiated, the bot will label the user as a prospect or a customer. Users can also sign up to newsletters straight from the bot. From there, it’s only a matter of minutes before a live employee is responding to queries.

The bot is using data in a number of ways: it is taking user data as an output, processing it and then generating further insights as an input to the Hubspot team.

Methodology

At Cyber-Duck, we use an accredited design and technology process to ensure that all the products we build are designed based on insights. That’s how we make sure that the data that’s generated is usable and useful. Typically, we undertake data research and user research as part of a higher level digital transformation or digital innovation programme.

When we’re looking at data, we use a blend of quantitative and qualitative, because they can reveal different things. Combine them and that’s where you find your insights.

Just make sure, whether it’s qualitative or quantitative, that it’s always empirical. Empirical data is any type of information gathered through neutral, rigorous observation or experimentation. It means you can trust it.

Data-informed design

A lot of business fall into the trap of being too data driven. At Cyber-Duck, we prefer data-informed design, where we use UX interviewing techniques to verify that data inputs and outputs are correct.

Here’s how you might approach that:

1. Business requirements: Set clear objectives and a deadline to avoid analysis paralysis.
2. Source/qualify data: Identify the critical data points that will help you understand the why and the what.
3. Cleanse/structure data: Data needs to be clean and have a baseline, so you can compare before/after performance.
4. Analyse data: Learn from your data. Check your qual and quant data match. If not, try to find out why.
5. UCD checking: Incorporate your learnings, then test with users. Refine and adapt. Then measure again.

Service design

Recently, the discipline of UX has evolved beyond just design, research and UI. Service Design embraces all of these things but looks deeper at the data, back offices support and processes, the org chart and even more. The term was first coined by G. Lynn Showstack, a banking executive, in Harvard Business Review in 1984.

Service design starts with personas and scenarios and can end up with reworking org charts and implementing new business processes.

Examine your key personas and scenarios

• Personas tell you who your users are. They focus on the person and the situation they are in depending on their segment.
• Scenarios tell you what they’re trying to do. They focus on the task, or joined-up task of many personas in similar circumstances.

Personas are important, but scenarios move you beyond the psychographics to activities, stories and tasks.

What’s next?

We launched our client Mitsubishi Electric’s new website 12 months ago. They have multiple personas – engineers, installers, property developers, architects – but there is one thing in common. There is a huge overlap between the type of tasks they’re each trying to complete. Their personas are different, but many of the scenarios were the same. So we knew to focus on those.

Once you’ve identified your personas and your scenarios, what’s the next step?

That’s where service design blueprints come in.

A service design blueprint is a tool that maps the user and data journey over all business dependencies, touchpoints, operations and systems. Designer Patrick Quattlebaum calls them, “the gateway drug to service design.”

Service design blueprints for UX have been pioneered by the likes of Adaptive Path for their work at CapitalOne.

A service design blueprint should include:

1. User actions: What your customers do, the script, scenario and tasks they follow.
2. Brand touchpoints: Digital, physical, online and offline interfaces that they need to deal with.
3. Back office staff: What needs to happen from the organisation’s human perspective to facilitate.
4. Data layer: What data is needed, how it should be processed and where it needs to go
5. Support processes: Technological systems, processes and activities that are required behind the scenes

They can also include timelines, success metrics or customer satisfaction across the process.

These help you drill deep into everything that supports the customer experience and identify where improvements and efficiencies can be made.

Here’s an example from a recent B2B client.

Taking everything we know about the scenarios and users, we build the ultimate user journey and ensure that we have the right processes, systems, integrations and people to deliver the dream.

The blueprint can then help to identify what can be automated and even enhanced by AI and personalisation. Which brings us on to …

AI and automation

Marketing automation helps us to identify, categorise, personalise, target and streamline interactions so we can automatically deliver a frequent and tailored user experience.

For example, a drip email campaign is a set of automated emails that are triggered by a user action or by time. It’s a way to nurture and keep re-engaging your audience with your product.

Here’s an example:

The four principles for successful marketing automation

1. Identify – You need an email address, CRM entry or unique ID from the start
2. Insights – Learning through data is going to be key
3. Content – The strength of your workflow is only as good as your content
4. Personalisation – Automation works hand-in-glove with personalisation

Unleash the power of automation with AI

The power of automation really comes through when it’s combined with AI. 

Juniper Research published a report last year that predicts chatbots will save businesses $11 billion a year by 2023, up from $6 billion in 2018, while Gartner says a quarter of customer service operations will use virtual customer assistants by 2020.

As author Diane Ackerman said, “Artificial intelligence is growing up fast.”

When it comes to designing AI experiences, a common mistake is to use the data you already have, rather than determine what data the system and users actually need.

Think about a GPS navigation system. What would be more useful: “Please continue on the motorway for 5 miles” or “Please continue on the motorway for 5 minutes and be ready to turn after the bridge”?

For most of us, it would be the second. But when GPS was designed, that functionality wasn’t built in, so for now we’re stuck with guesstimating distances.

The three rules for successful AI

1. Logic: AI depends, just like the human brain, on combining observation, data, memory, creativity, motor coordination, judgement and wisdom.
2. Ethics: AI requires human supervision of machines and machine checks on human inputs.
3. Rules: AI needs to work with rules, like cars need the highway code and motor laws.

Personalise the data experience

Marketing data at its best can power tailored, useful experiences that will delight your customers and build brand loyalty.

As Seth Godin put it, “Personalisation … is a chance to differentiate at a human scale.”

Here’s how you can approach it.

First, listen to your user research so you understand your key data.

Then look at your user journeys. Think about how you can personalise them.

In the case of Mitsubishi Electric, we switched the information architecture from product-centric to user-centred.

Help content can be filtered by product type (e.g. air conditioning) or user type (e.g. installer) – currently by the user, and that’ll be done automatically further down the roadmap.

A lot of work has been done to tag the help content, so again it will be matched and given to users on the fly in the future. Where relevant, articles will point to dedicated landing pages on products where visitors can then receive harder sales messages and register interest.

The aim of this content is to empower users to self-serve on the website for product information or troubleshooting, which will also reduce the volume of calls made the support centre.

To sum up: 

1. See data as the future of your product or service – it’s output as well as input.
2. Qualitative research complements your data strategy.
3. Build scenarios around your personas and help your users succeed.
4. Service design blueprints underpin and help you reinvent your customer experience.
5. Use principles for automation, personalisation and AI to achieve true digital transformation.