Reposted from IMHO.
In today's brief customer journeys, customer experience professionals are both artists and scientists collaborating on behalf of our most important benefactor. As artists we want to visualize and experiment with these benefactor touch points but there is this wall of complexity around how we intelligently scribe and simulate a million points of light from a hundred thousand journeys. As scientists we seek to simulate and experiment with our learnings and shape them into thesis or ultimately repeatable models.
If you follow Gartner's thinking around context-aware technologies, it is clear that C- level executives need to seriously consider the disruption that context-aware technologies will have on purchasing and loyalty behavior. Gartner identifies context-aware technologies as the ones that influence consumer purchase decisions by using information about the consumer's context (e.g., location and interests) to offer more-relevant promotions, content and recommendations. Smartphones and tablets armed with context-aware apps will influence consumer spending as much in nonelectronic channels (e.g., physical stores) as in e-commerce and the technology will primarily be used to shift consumer spending from one competitor to another. It does point out that is less likely that it will increase consumer spending overall. The customer journey, when taken into consideration at a touch point, must be considered in context.
What is the taxonomy of this union of art and science, this cartography of a journey and ethnography of a participant, perhaps individual or en masse as an audience? What are the unifying factors and measurable, logical groups of data that we can ultimately action? We have more than enough market data to understand that context in its totality has all the facets of data needed to optimize a behavior. But context is the thing itself, what then would we call the process of understanding and leveraging the learning of context? I propose:
Contextography - n. - the collection, study, analysis, measurement and resulting use of context. e.g. For the purpose of this article, the user speaks in italics to help you empathize with data gained through contextography.**
I'm on a horse. You knew that. right?
First let's meet a user and their context. In the same way that Old Spice imagined how women might like to see the men in their lives in order to have men in turn see themselves through that lens, we are constantly anticipating an audience and their needs, desires and motivations. I am hoping that I have anticipated your need to better understand how context is made actionable even if I am really only able start another conversation at this inflection.
While we have gotten better at allowing users to manage their account, their profile, and even how they share their activity with other customers or users of a product or service, we also have had to learn over time to consume that data intelligently and understand patterns that emerge and perhaps express these as context. The parameters that are relevant for the context can be broken down in different dimensions. For the purpose of this short journey we will give our user, which could be me, some context.
Me, the user, of course: Sometimes you and sometimes me, but always everything that characterizes the customer, perhaps demo-, psycho- and ethno- graphic.
My situation: That which describes the circumstances in which the user interaction takes place and the process and result of interaction. Relevant parameters for this include the channel of the interaction, the device made use of, the location, and the network that is used to connect and the facilities that enable this.
My history of interactions: That which we know and are able to share about the relationship between the user and your enterprise. Relevant information include buying history, contracts, support cases, and any other information that characterizes the business relationship.
I'm at a touchpoint. It's pretty. It doesn't work the way I expected.
A user’s experience is always dependent on a defined relationship between a business activity and that user’s goals. While the experience may take many forms, there are commonalities in the approach to arrive at a final experience, and this process is the practice of user experience (UX). The disciplines essentially map to the outputs of this process and typically include:
- Interface design - the graphics and branding.
- Interaction design - the method by which users interact, e.g. touch and voice.
- Information architecture - the visible organizing principle for content and applications.
- Graphic design - the brand treatment, color palette and treatment of text and media
Technologists, behaving like scientists, strive to have applications widely adopted are essentially questing for patterns; patterns to users means that there are functions in those applications that are repeatably useful, usable, easy to find, credible, and ultimately successful at solving an identified problem or achieving a known goal. As the increase in focus on great customer experience matures, so in turn does the practice and rigor associated with defining the patterns of engagement. It's fair to think of increased investments in great user experience as a discovery of true patterns - the right investments presented in the context required for the right happy customers and happy employees to use your apps.
I know what awesome is. You and I are on the same page.
Essentially, most investments in user experience within the enterprise today amount to superficial and cosmetic changes applied as afterthoughts in an attempt to solve apparent and predictable problems with the surface of the application. Often, people focus on type size, color, and other basic design, and believe that they have created a great customer experience. The reality is that this approach only goes so far in addressing problems and is ultimately going to fail since the user-focused work was started much too late in the process. Business users often make the mistake of masking the lack of user input with cool technology, and while this world of wonder can fascinate to the point of going viral, it often lacks the deep engagement intended.
I don't make mistakes, you do. I recently redefined the customer is always right. I did this for you.
The later in a project you invest in change, the more it costs. There are a number of methodologies in use today across the industry that are purported to facilitate user input into the design and development process - beware the linear, embrace the cyclical, and mandate the agile. In this way we appropriately recognize that users don't really blame themselves for things going wrong the way that they used to. That was convenient because it presented an opportunity to teach someone how to use a system in the way it was intended. Now the table is turned and the system must work in the way that is expected.
I'm on a journey. I have a map. Embrace my journey.
Many of the guidelines from analysts and industry experts have identified a key tool in this process to be a map of all the customer touchpoints across an enterprise. This map can be used in several strategic ways:
- Identify and plot persona against specific actions or opportunities.
- Understand and modify business processes in order to overcome obstacles or bottlenecks.
- Inform design and technology choices and prioritize resources against those choices
- Measure success by defining Key Performance Indicators (KPI) based on customer activities.
- Refine opportunities for increased productivity by aggregating logical groups of actions.
I'm in your systems. I'm doing this fast.
Finally, we need to reflect the outcomes of design in our application architecture. In a functional architecture we can surface this experience layer to bridge the “last mile” gap between user and business application. This layer essentially represents the presentation layer of applications (e.g., interaction models such as touch) with knowledge of domain, integration, and the associated infrastructure. True multichannel delivery is thus enabled through an abstraction of presentation that intentionally separates the channel, controls, and interaction from the fundamental underlay of business logic and application code.
This abstraction serves a unique purpose in planning and development. By creating handshakes between the UX professionals who own the experience and developers who own the implementation of the application, we in turn empower handshakes between customers and the business.
For example, wireframes that represent the experience layer and interaction model can be made interactive in such a way as to represent the potential interaction, and elsewhere in the team those interactions can be wired to the application and business logic. These can be tested with actual customers or users in order to further refine logic, interaction models, and general usability themes such as accessibility, both prior to implementation and over time.
Suck less. Be awesome more. Please.
A user’s experience is impacted by many things beyond our control as designers, such as network issues, device or operating system issues, IT policies, or even physical distractions. What is within our control is exhibiting a shared understanding of goals and interaction capabilities and providing this in a consistent way to support the brands we represent.
One could choose platforms and tools that effectively reduce the time it takes for you to develop the final experience with a component model based on UX best practices. For example, here at Adobe, we are working hard to maintain a domain model that is essentially pre-integrated with relevant technology services and infrastructure, and abstract this from the presentation layer such that you can reuse or strip away and replace at whim. Not your whim, but that of your customer's oft-fickle hearts and minds. In this way you can adapt to changes in contextual trends at the edges of your business, and put your new passion for contextography to work helping to sustain and grow your business.
I'm done. Listen for my feedback.
** Re: Contextography. I honestly had no idea that someone else had made the word up before I did but Google was helpful in setting me straight on that. Recognizing the definition potentially already in place actually helps me build on that to an expanded definition that is first of all both representative and inclusive of the user POV, and more importantly one that embraces all aspects of a digital environment, not only images of a fictitious one. I should point out that more recently I have been made aware of a definition of this term that is similar to a bibliography for a paper or thesis - essentially tracking the context of sources. I love that definition and my only regret was that I was not personally encouraged to add contextographies to my papers when I was a student many (many, many) years ago.
I have registered the domain Contextography.com to build a body of research and work in the areas of definition, research and analysis of both the art and science of context. In parallel to that we will also be tracking the context conversation here - @contextography (sorry to any fans of @uxpectations, that chapter is now complete for me) and I will bring the first few months of this together for a talk on this subject at the upcoming Adobe Digital Enterprise Summit in October. Register for VIP invite here.
Special thanks to Hank Barnes, Craig Randall and Jamie Anderson for additional insights and thoughts, and to the extended community for the inspiration and motivation to get this idea off the napkin.
No comments:
Post a Comment