UX Strategy

You can't design for how every person engages with a product or service, individuals have too many variables, too many pre-existing concepts that are unique to them or to their culture to make a one-size fits all UX strategy work. What I do is create a mitigation strategy, one that allows me and my teams to create a best fit for our user demographic. Here's my simple UX Strategy template, complete with real examples that I've used with various clients, so you can see my 'one pager' approach.

Achievement Goals

Not what is your Vision or your Mission, that's too fluffy and nobody other than the author really aims for or believes in them, no this is what it is you honestly want to achieve? Here's one I created for an education client:

We want to achieve an app where every visitor feels like it has been created for them. This is not about achieving sales it about achieving a quality of product of such value that sales follow naturally. We want achieve a system that not only allows users to learn about subjects and topics at their own speed it also creates a feeling of wellbeing about learning. Ultimately we want to build a feeling of trust and the desire to have a continual relationship with our brand.'
You'll notice there is nothing about how this going to be achieved, the rest of the strategy should do that, there is nothing about brand, product or service details, this all about the relationship, your engagement, between you and the end user.

Identifying Users

This is the first time you'll be explicitly identifying how you're going to meet user needs, your approach to User Research. This isn't answering questions, it's identifying the right questions to ask. I'll use a different client this time, one involved with high-end retail.

"If we analyse our call centre correspondence, what exactly are people looking for, what problems are they having finding it? What existing user demographics do we have? If we don't have any can we find them through contact data or using a survey on our existing website? What is the typical spend in a single sale? What products are purchased the most, which the least, does this tally across all user touchpoints or does purchase behaviour change between online and offline?'

We are making no assumptions at this point, we're trying to ascertain what it is that users want, not just in the way of product but in the way of service. We're trying to identify if buying habits change across demographics touchpoints and when we have that information we can go about validating through user research.

Defining Users

Simple personas (proto-personas is the buzzword) are my favourite thing. Some personas I've seen look like they were produced by Leo Tolstoy, It's lovely that Jeremy Generic has 2.5 children, a small dog named Biskits and took a year off while studying Economics at the University of Elsewhere to backpack around the outskirts of Ulaanbaatar to learn more about the Khan Dynasty, but is that really going to help you understand your users when it comes to how they approach buying a new TV?
This is a simple outline persona I wrote for an automotive client:

"Who: I'm Jenny, I'm in my early twenties and I've a limited budget as I'm paying off student debt.
Experience: I've always owned second hand Fords because it's what my parents drove and it's all I can afford. I can do simple maintenance like checking tyre pressure and oil levels but I just want to drive not think about engineering.
Wants: I want to know why I should switch to your brand, what my options are in comparisons to other brands in my price range and what independent reviewers say about your cars.
Needs: I want to know prices upfront, no hidden extras and I want to know what is available now in my location.
Reasoning: My existing car is coming to the end of it's usefulness and I want to make certain that my next car is trustworthy.'

For every one of those criteria you can produce a design rationale. For example, we know this person is working under a limited budget so a second-hand car finder might be of use, or may be even a new car if finance is available, how can we achieve that?

Usability Testing Criteria

There are far more research principles and techniques than I'm going to list here, but it is important to remember that every assumption, be it user need or business need, should be tested for validity. What you're aiming for isn't just usable, you're aiming for useful too. This is an approach I wrote for testing with a banking client

"A series of tests will be conducted with existing customers, these will compromise:
Pre-test interviews with both direct questions over existing app usage and with usability rating scales, relating to savings approaches. We aim to discover if the assumption of the needs for a simple savings account creation are correct.

Behavioural tests will be conducted one-on-one to ascertain if users behave as expected and can use the proposed savings account creation journey. We want to establish if the user journeys, the labelling, the interaction and engagement patterns we have assumed are correct.

Exit interviews will be conducted, their responses will be compared with observers notes to ascertain if the participant's understanding and approach has changed over the course of the test. We will be working in an Agile environment and as such we will be conducting guerrilla testing as appropriate to ensure that any development and design changes that occur to the UX design still meet identified user needs.'

This is fairly light, we're not tying the researchers' hands when it comes to what we're testing or specifying in any real detail how we're approaching the testing. What we are committing to is real testing with real users to obtain tangible results and not just upfront, there are no silos here, this testing will continue right through the development process too.

UX Design Criteria

This is the difficult bit, not that should be it's just that you really must be wary of limiting UI design and of using soundbites. How many times have you seen a design or vision document that uses phrases such as 'A bleeding edge, intuitive, simple interface'? I always want to ask "Is anybody deliberately building pedestrian, difficult to use, convoluted interfaces?" Avoid those phrases, this isn't about ephemera this is about being specific, about stating upfront not just what you're going to achieve but how you're going to achieve it. This isn't about UI design either, you could just as easily be designing a brand engagement or a service design. No this is about how you design an experience.

This is one I created for a health care service.

"As Needed Content: Users will not be overwhelmed with content, easily identifiable subjects and topics will be clearly displayed with each step progressively displaying more detail as needed.
Distraction Free: Interaction patterns will be simple, the content should engage the user not the method of displaying the content.
User Driven: Once a user journey has been started the user will be shown alternatives but at no point will they be taken out of their journey unless they choose to start another journey or process.
Clear Navigation: The user will have a clear mental model of where they are at all times, being able to jump between pages, subjects, topics and details and returning to their jump point as needed.
Optimised cross selling: Complimentary and alternative products will be shown at the start and end of the journeys, at no point will the system try to second-guess the user once they have made a decision on a product line.
Tone of Voice: The majority of the site will be factual but informal. Clear technical TOV will be used if and when the user requests additional detailed information of products."

What you're trying to convey here is how the user will experience the content, how they'll navigate the system, how they'll be presented with data. As far as aesthetics or content goes we're agnostic, that will be part of a content and creative creation strategy, which we'll take part in, but these are the user experience principals we will be championing in those discussions.

Measuring success

How do you measure if what you've done is actually effectual? You could go to a white board and write, in get big letters, the legend "More people to buy product" and believe me you'll make friends by doing so, but you'll make longer lasting friends if you can quantify and prove that you've done it.
I'm a big believer in H.E.A.R.T. developed by Google, and use it whenever I can. You can download the full paper here.

Happiness: Measures how much and what users think of you, of your product or brand. You can establish a benchmark by quantifying negative and positive feedback through surveys, social media comments, app store feedback then aim to improve on that benchmark with your solutions.
Engagement: Sometimes called stickiness, this can be measured by behavioural metrics such as frequency of visits from unique sources, time spent on content, number and frequency of communication interactions over a fixed period. Number of user created pages or content. If these increase over established norms when your solution is made live, you've done your job.
Adoption: How many people utilise a feature or visit a page now? Can you increase that? If the feature isn't along the lines of 'report a problem' then you've got a good indicator that you've improved your user experience.
Retention: This is your churn rate, how many existing users make return visits how many do you lose over a given period? If you can reduce your churn rate and increase return visits that's an excellent metric to achieve.
Tasks: This is the time taken to complete a task, number of incomplete tasks undertaken, number of abandoned tasks made, all of which can be measured in existing products and services and therefore improved upon. All that is great, but as simple as it sounds it can be very messy and I recommend incremental adoption of the techniques. For example what is more important to your project, keeping existing users (Retention) or process completion (Task Success)? You might find one improves the other, so focus on one and then improve the other if the other shows no signs of growth.