Every website needs an audience. And every audience needs a goal.
Advocating for end-user needs is the very foundation of the user
experience disciplines. We make websites for real people. Those real
people are able to do real things. Everyone is happy.
But, it’s not really that easy, is it?
The issue, of course, is that we cannot advocate for those whom we do not know—or, even worse, those whom we
assume
we know. So we go to the source: we interview, we learn, and we
determine who, exactly, these mystery users are. In doing so, we answer
the two most important questions of the discovery stage: who are our
audiences, and what do they want from our website?
Then—and only then—can we begin the process toward better content.
Defining the process
End users are a funny thing. They begin as amorphous blobs of assumed
stereotypes. As we learn more about them, they become more refined.
They develop characteristics and quirks. The more we learn about the end
user, the closer we get to a sort of Pinocchio scenario: they become
Real Users!
The process of creating Real Users is what we at Blend call our
Audiences and Outcomes process. It happens before any other part of the project, and is based on C. David Gammel’s book
Online and On Mission, in which Gammel pushes the need to identify and prioritize audiences before you develop any strategy.
In this case, Gammel defines an audience as:
Any group of people with some measurable characteristic
in common which influences how relevant and significant they are to your
specific outcomes.
Likewise, Gammel defines an outcome as:
A measurable change, action or behavior that you wish a visitor to take or experience.
In other words, just stating a goal is not enough. Outcomes must be
measurable, otherwise they’re not goals—they’re aspirations. Without
considering how an outcome will be measured, we cannot accurately
represent the benefits—or the viability—of a user outcome.
Finally, we use audiences and outcomes to create user personas. We
use these personas until the project is complete. Because we can’t run
every decision past a field of actual end users, we rely on personas to
do the work for us. They become our friends. We refer to them by name in
meetings. It would all be very weird, if it wasn't so necessary.
If you’re confused about how this differs from your standard
discovery meeting, with people meeting in a room and answering questions
and all of that, the answer is: it doesn’t. Not really. You may already
do something like this without being so deliberate, or you may define
audiences and outcomes elsewhere in your process.
That’s cool. We’ve found that tackling audiences and outcomes at the
very beginning makes our content inventory more relevant (by allowing us
to pair pages with audiences) and saves a step in our qualitative audit
(by giving us context for content needs).
What’s more, it clarifies our goals from day one. This clarification
is important. For example, if we’re building a site to sell mail-order
diapers, we can’t just say, “We’re building a diaper delivery site, and
mothers will come to the site to buy diapers, and so let’s start writing
copy.” I’m not a mother. And if I was, I’m certainly not EVERY mother. I
know damn well that fathers and other caregivers will come to the site,
too. So if I move forward with the diaper-buying mother stereotype in
mind, I’m doing a disservice to a giant percentage of the site’s users.
We are not the audience. We can only assume what our user
thinks. Which is where the audiences and outcomes process comes into
play, allowing us to narrow down who the user REALLY is through
stakeholder input, persona development, and persona confirmation.
Buy-in: everyone wants to play
These proxies serve an added purpose: they help us get buy-in and
help us back up assumptions. Erin Kissane sums it up perfectly in
The Elements of Content Strategy, when she says:
The personas or other user proxies that you or your
colleagues have created are the best backup you could hope for. Return
to those tools when you need to validate opinions—yours or someone
else’s.
By pulling stakeholders into a room and getting them to talk about
their product, their audiences, their issues, etc., we’re giving them a
way to buy in. No longer is this a consultant-driven process—it's a
company-driven process, where the consultant serves more as facilitator
than dictator. As I’ve mentioned before, we don’t have all of the
answers, and we do our clients a disservice by assuming we do.
When stakeholders get involved early, they are less likely to hold
things in. They’re also less likely to object to changes in their
content or process because, after all,
we’ve all agreed on our audiences.
So, how do we do it?
The audiences and outcomes process
Let’s assume that within your methodology, you define audiences and
outcomes at the very beginning of a project. It’s how we throw a project
into context. It gives us a high-end view of who we’re dealing with and
provides background for the qualitative content audit. It’s a “getting
to know you” period of two or three weeks, broken into three steps:
- Step 1: The discovery meeting
- Step 2: User interviews
- Step 3: Project deliverables: audiences, outcomes, and personas
The discovery meeting
There’s one goal: to get this group talking. Invite a small number of
people—five to seven—and make sure someone from the front line is
there. Ask them to let go of preconceptions. If an executive is present,
make sure they don’t take over the conversation.
From your side, make sure you have two people. One will lead
discussion, the other will document the discussion and provide an extra
brain to ask and answer questions.
The meeting should be an open forum for discussing the website’s
needs and goals. Schedule no more than an hour, but don’t be surprised
if it goes a little longer. Begin the meeting by introducing everyone,
introducing the process, and explaining how it fits into the overall
site plan.
Bring big markers and whiteboards and anything to get ideas up in
front of the group. Then, make two huge spaces for the following
headers:
Audiences -
Outcomes.
Then it’s time to start asking questions.
The questions
These are the questions that we use at Blend to define audiences and
outcomes. They aren’t law. In fact, we’ve never made it all the way
through this list. (Note: you’ve probably seen some of these questions
asked elsewhere. Of course you have—we created this list, like any good
list, by stealing and adapting ideas for our own use.)
Audiences
- Who do you feel are your site’s audiences?
- What are the demographics of these audiences?
- How comfortable with technology is this audience?
- Who is currently visiting the site? What makes their visit a success in their eyes? In yours?
- Who else is competing for their attention?
Outcomes
- What do you want to persuade your audience to do?
- What assumptions do you make concerning your audiences? Example: do
you assume your audience is of a certain socioeconomic group, or that
they are familiar with certain aspects of your organization?
- What drives your business, and how does your audience help achieve positive results?
- What metrics do you want to keep track of?
Company voice
- What is your company’s ultimate mission? (Not a mission statement,
but a more organic, real-world one-sentence answer to “Why do you do
what you do?”)
- What message do you need to get across?
- What is the company’s voice and personality?
- What has worked in the past? What hasn’t worked in the past? What were the stumbling blocks?
- What
attributes does your company have that helps to gather attention—i.e.,
“Our company is nationally known,” “Our company employs former movie
stars,” “Our company is well respected in the field.”
- What
topics can we take advantage of? Example: if you are an automobile
manufacturer, are there government rebates we can promote?
- What topics are off limits?
Methods
- How do you currently communicate with your audiences? How often? (Related: can we have copies of your past materials?)
- Who creates the content?
- How does your audience prefer to communicate with you?
- What other functionality will you need?
Content management
- What is the current content workflow?
- Who currently creates content?
- Who will write it in the future?
- Who approves content?
- What stumbling blocks are in place that make it difficult for the content to get published?
- Who in the company connects with customers most naturally?
As you push your way through a group’s initial fear of discussion,
new questions will flow naturally. The specific questions aren’t
necessary, as long as you remember to:
- Ask the client who their audiences are and what those audiences want.
- Listen for clues that expose secondary audiences.
- Dive into those clues. Make them work. Ask follow-up questions. Get people talking.
To create an extremely basic example, imagine we’re doing a meeting
with a fictional airline: On-Time Air. After a very general discussion,
we’ve got this written on the whiteboard:
Audiences:
Outcomes:
- Find accurate flight information
- Book a flight easily
- Learn about baggage fees
- Locate gates and flight times
Digging into the audiences, we ask what else passengers look for. One
person mentions that, yesterday, someone asked for a chart depicting
the airline’s on-time percentage. This may not be a high-level outcome,
but it reminds us that when something newsworthy occurs, the press may
look for information on the airline. When a potentially damaging article
is about to come out, it’s the airline’s policy to let its employees
know ahead of time. What about new employees? What about future
employees? How do they apply?
As you can see, this line of thinking led us to three new audiences
(the press; new and future airline employees) and three new outcomes
(finding company news and information, looking for explanations on the
airline’s problems, and locating job applications).
What’s more, we can start to see sub-audiences: employees could be
separated by type (pilots, front-line staff at airport desks, those who
take phone reservations) or by status (new, potential, veteran). What
about the separation between a current passenger—those who have tickets,
and a potential passenger—those
looking for tickets?
Some audiences will share outcomes. Both a pilot and a current passenger
may be looking for related or identical information.
NOTE: You’ll notice that those last groups of questions aren’t
really audience/outcome related. That’s okay. We’re opening up here, and
your stakeholders will be in the mood to talk. This gives us a chance
to grab a little extracurricular research. Taking Tiffani Jones Brown’s Making Things Hard
post a step further, we’re making things easier by asking hard
questions at a time when the client is more receptive to those
questions.
User interviews
The initial strategy meeting is designed to help determine whom the
site is for and the goals that need to be addressed; in other words, the
meeting shapes the audiences for the site, as well as the desired
outcomes related to each audience. The next step is to talk to actual
site users to determine whether these audiences and outcomes are
accurate.
This happens early in the process for a reason: we need to know who
the user is, and we want to use their opinions throughout the project.
So we ask our clients for a list of contacts, and we talk to past
customers. Or, we solicit opinions from a related industry group.
Regardless, we ask questions.
For example, if we're talking to an audience of building managers, we could ask:
- How do you secure funding for a project?
- How many companies are you required to look at during a bid?
- Who provides post-build service for a project?
But we could also ask questions that help gauge an audience's personal and technological habits:
- What kind of mobile phone do you use?
- Where do you live and how large is your organization?
- How often are you on the internet for non-work purposes?
Then, we compare their needs and perceived outcomes with the ones our
client mentioned. If they match, then awesome. If not, even more
awesome: time to bring it back to the client and say “This is what
people want.” We’re already learning, people!
Project deliverables
Whether we like it or not, a huge part of content strategy is
delivering documents, and the audiences and outcomes process is no
different.
First, prioritize each audience and assign numbers to each user
outcome. Because every audience could make a case for being the most
important, it’s up to you and your client to determine which audiences
are really the most important. Not only does this provide a handy
cheat-sheet for solving design hierarchy problems (think: should our
home page focus on existing members or new members) it also helps you
determine which personas will get more space at the theoretical persona
table.
(Not to mention: these numbers will come in handy during content
auditing, where you can match content with goals. Saying, “Outcome 3.2a”
is a lot shorter and easier than typing out the entire outcome.)
You might need to split some audiences into more manageable
categories. These “sub-audiences” share the same overall goals as the
parent audience, but feature some additional needs and goals.
For example: A physician’s website may have three major
audiences—referring physicians, patients, and staff. Patients could be
further split into three categories: new patients, current patients, and
family members of patients—all three will have the same overall goals
as the patient audience, but with additional goals dependent upon
further classification.
Giving the document structure
In the beginning, the structure of the audiences and outcomes document might look like this:
- Introduction
- Summary of findings, including user interview findings
- Audiences and outcomes (example below)
Audience 1: Customers
Customers are those who are either thinking about purchasing an
airline ticket or who have already purchased an airline ticket. They are
the main source of income for the airline, and represent the airline's
most important audience.
We can split customers into two distinct sub-audiences: potential customers and current customers.
Sub-Audience 1.a: Potential Customers
Potential customers are those who have yet to purchase a ticket.
They’re visiting the site because they’re interested in traveling, and
may be researching either current ticket prices, flight schedules or
both. Their mindset depends on context—they could be voluntarily
researching a trip for leisure, or they could be locked into a trip and
simply need the cheapest or best flight.
- 1a.1 - Find and compare flights by price, date, and flight details.
- 1a.2 - Purchase desired flights with little resistance.
- 1a.3 - Locate and understand flight rules—check-in time, weight restrictions, etc.
Sub-Audience 1.b: Current Customers
Current customers are those who have already paid for their ticket.
For the most part, they are no longer comparing prices—they’re on the
site to confirm existing flight information or make changes. Research at
this point has shifted from discovery to confirmation.
- 1b.1 - Access flight and ticket information.
- 1b.2 - Contact On-Time Air for flight changes or questions.
- 1b.3 - Locate and understand flight rules—check-in time, weight restrictions, etc.
Notice that both potential and current customers share the outcome
"Locate and understand flight needs." This is common. Audience desires
always overlap, though there can be differences in how we measure these
outcomes.
Speaking of measurement...
Making things measurable
Because outcomes should be measurable, we need to bring analytics
into the mix. Determining these metrics helps us understand what’s
important to the site and, more importantly, how we determine whether a
content plan is working as imagined.
We’ll take our original document framework and add these metrics to
the desired outcomes. For example, for desired outcome 1b.1 above, you
may say:
- 1b.1 - Access flight and ticket information.
- Metric #1 - Lower page views per task.
- Metric #2 - Fewer customer service calls for flight and ticket information.
You can see that there are two avenues to measure better access to
flight and ticket information. One is to measure how many pages a user
goes through before finally finding the information. The other is to
track customer service calls to see if the number of calls for flight
and ticket information decreases.
Much like the outcomes themselves presented some level of overlap,
your metrics will overlap as well. Don’t worry. That’s normal.
Erik Peterson’s
The Big Book of Key Performance Indicators is a good place to help determine which metrics to use for each outcome.
Project personas
With metrics in place, the next step is to create personas for the major audiences.
The persona process has been well documented by nearly everyone, it seems. At Blend, our three favorite resources are these:
The number of personas you create depends on the number of unique
audiences you've determined. In the case of On-Time Air, let’s say we
have six unique audiences, one of which is the 1.a: Potential Customer.
We would then create a persona that represents a potential customer:
Martin Hunt
Age: 46
Occupation: Architect
Family: Married with one child (19)
Education: Architecture degree from St. Cloud State University
Habits: Martin uses the internet every day, but has
never been a heavy user. He relies on aides for most of his online
information and research, and spends a good chunk of time answering
email, but outside of that he’s a novice. He has a Facebook account and
has signed up for Twitter, but has never posted. He has a Blackberry.
Assumptions: Martin is wary of the unofficial nature
of bargain travel sites like Kayak.com and Priceline.com. He prefers to
order tickets directly from the airline. Because he lives in
Minneapolis, he must often fly Delta, but he is increasingly interested
in On-Time Air’s direct flights from Minneapolis to Orlando and
Minneapolis to Las Vegas.
Martin doesn’t care about price if the difference between two
airlines is close—he’s more concerned about how comfortable he’ll be on
the flight. He values relationships, and assumes the airline best
positioned to win his business will be the one that does the best job of
selling a unique experience. At the same time, he has already racked up
a considerable amount of airline miles with Delta.
“I would rather fly comfortable than fly cheap, but I won’t accept an exorbitant price.”
Naturally, the details will be different depending on the project.
And remember: personas are made of real people. All that questioning and
interviewing you did earlier in the process? That helps inform and
define your personas. For now. Until you do more interviewing and your
personas become stronger and more agile. Better cooks, even.
Look at the personas you’ve created. Do they accurately represent the
outcomes you’ve determined? In writing them, have new outcomes turned
up? Go back to the beginning and make sure everything levels out. Your
audiences and outcomes determine your personas, and your personas should
validate your audiences and outcomes.
You can do one of two things in terms of deliverables—you can add
these to the Audiences and Outcomes document, or you can present them on
their own. We feel that they are so closely tied to audiences and
outcomes that we include them in that document.
Next steps
With your audiences and outcomes in place, you no longer have any
excuse for not knowing what your project’s goals are. The document
informs the content audit, where you can begin assigning relevancy to
every page on an existing site. Every piece of content on the site
should relate back to a specific outcome, and if not, it needs to be
reviewed for relevance.
From there, the audiences and outcomes will help you choose
participants for user interviews and user testing. Find a group willing
to help you through the entire process, and you’ve got a valuable
resource for not only testing prototypes, but also for performing card
sorts, confirming IA tree testing, and serving as a de facto advisory
board on the subject.
Finally, the audiences and outcomes help remind you what really
matters in web design and development: the end user. We place an
overview of a project’s audiences and outcomes at the beginning of every
wireframe, style guide, and specifications document. Doing so gives us
easy access to our user’s needs. After all—without the end user, we’d
have no one to impress from here on out.
