Your research should tell a story

We frequently field questions from researchers on how to be more impactful and influential with their research. We get it. Research takes significant time to plan, facilitate, synthesize, and report. Then, after all of that, the research team is given 30 minutes to communicate their findings and inspire action to address the opportunities raised. This is occasionally followed by a kind “thank you” with no actions planned as follow up. Many of us have been there.

Is it the topic of research? User research for enterprise problems isn’t that interesting to read about, you are likely thinking. Perhaps. That is not the problem, though.

Data is boring, you say. No doubt, focusing only on data can be boring. Yet, we see this regularly in the way research is presented.

How does one fix this? Is Design & User Research damned to be boring simply because of the domain and data? Hardly. Researchers need to embrace the art of storytelling to become more influential and break the boredom. Stories, when well written, are naturally engaging, capture your attention, and have the audience attempting to guess what’s next. Perfect for increasing audience engagement, participation, and action as you playback your research.

A common starting point and approach for documenting and playing back research uses two main sections: Tell the audience what you did and what you found. While this keeps the approach simple, it leads to stale, unconvincing research artifacts that rarely encourage action to be taken by the business. Take a different approach by building a story that communicates a tighter context to business objectives, builds on previously gained knowledge, and drives a deeper understanding of the business impact of both your findings and lack of action. Let’s go deeper.

Setting up your story

Start by framing your research story around the bigger business and design objectives along with the questions you are seeking to answer. This should provide a clear reason and business justification for the research. Be clear on the questions you need to answer, while illustrating the importance of these questions by connecting them to business context. Let’s say, as an example, that you are planning a study to understand why developers don’t complete all steps in the task of building and publishing a chat bot. While you could frame the research as seeking to understand the general ease of use of bot building, that doesn’t give the audience an understanding of why the research is important to the business.

Alternatively, you could frame the research with business impact to increase the relevance of your efforts. For example, Developers currently abandon the bot building process before step 4 of 5, which is costing the company approximately $5 million a year in opportunity that is then likely obtained by Acme, our closest competitor. What is causing developers to abandon the task before publishing their bot? Notice that the business context points to one or more critical questions and provides a business impact for the need to arrive at answers. Using some preliminary usage data or heuristic review to frame and describe possible user problems will help map to possible business impacts of user abandonement and pitfalls. This serves as a great introduction to your research and immediately draws in the audience with the business-critical focuses you are going to answer.

Explaining the motivation for conducting the research in business terms provides the audience with the deeper context and helps them better understand the problem. If the motivation for the study is that we’re doing the research because it is a required part of the process, then stop. Take a step back and reflect on what you are wanting to learn and why. What are the key questions that you are looking to answer and why are those important to the business? This will provide the starting point for building context for your research and tying the motivation to business goals. This will also help you frame what other research has been done in this area.

Setting expectations (and creating suspense)

After grabbing your audience’s attention with your motivations, business context, and any related research, forecast what you expect to learn based on current knowledge. Returning to the chat bot builder example, a forecast could include something like, “Bot builders are not publishing their bots because they become confused with our process while in step 3 of the task. We believe that this confusion is resulting from confusing text and poorly labeled buttons.” This speaks to the importance of understanding the larger user journeys for the critical jobs to be done that you are assessing. Conducting both a heuristic and competitive review of the application provide good starting points for establishing a starting forecast.

Providing the forecast of what you expect to find isn’t intended to show that you are clairvoyant. It instead helps your audience more deeply understand the research questions and provides a frame of reference for comprehending your findings. This technique also naturally builds suspense in your story. In more theoretical scientific research, it is critical that you forecast findings based on your theoretical position. These forecasts clearly frame the findings expected from your theoretical position and help your audience understand what you expect to find and why. Remember, your audience may be newer to your research, while you have been focused deeply on it for weeks. Your goal is to build your audience’s understanding of the problem, its importance to the business, and why you expect to find certain results. You want your audience, at this point, to believe in your forecasts and understand why your research is critical to the business. Not setting this stage frequently raises doubts about the need for the research and possibly even the credibility of your findings.

Explaining your methodology

The research methods used and why you chose them serve as a critical step in your story. Research methods include everything from who participated in your study, how you recruited them (for example, were they recruited from one or more sponsor clients?), what exactly participants were asked to do and why, and what is being measured. Keep in mind that your audience for design research will typically include folks who are not familiar with standard research methods. Why you chose a particular method doesn’t require a lengthy explanation, but you should explain how the method best enables you to answer the key questions to the business and how your participants in the research represent your actual users of the product. When recruiting participants from a single client, you will also want to speak to any biases that may exist and how these are overcome.

Communicating the overlap in characteristics between your participants and the users of the product in market also increases confidence in the interpretation and application of findings. From an enterprise product design perspective, these characteristics could include jobs regularly done, experience level of job mastery, amount of training, etc.

When writing scientific research papers, it’s critical that your methods section describe every detail of your methodology, experimental design, stimuli shown to participants, and all equipment involved. This is intended to provide adequate instruction for researchers to follow and replicate your findings before building upon them with additional experimentation. Focusing on design research, it’s advisable to be as clear as possible to increase your audience’s understanding and buy-in on their part to facilitate action. Enabling additional teams to replicate and build upon your study is typically less of a goal for design research.

Presenting your findings

Here’s where your story can become really impactful when done well. Many researchers will present the data gathered as if it speaks for itself and tells the complete story. It never does. Stating findings from the data without a larger story is similar to just reading the chart for your audience. Option A was faster than option C. So? Tell a story of the participants as you present your findings, including how they started, where they struggled, excelled, and why. This will increase the understanding the audience has of your findings.

How many stories am I telling here? Good question. To this point, we’ve stressed the need to tell one story of your research. For example, describing why you are conducting the research, what others have done in this area, what you expect to find, the business impact of the research, etc. A second yet equally important story surfaces when communicating your findings. Telling a participant-focused story from your data facilitates a much deeper understanding of your findings by giving your audience a framework for remembering the data and comparing it to their own experiences.

Now that your audience clearly understands the behaviors shown by participants, frame your findings with the larger business context to increase your impact. You can communicate this context either as you tell the participant-focused story of your data or directly afterwards in a Taking Action section. Doing both won’t hurt either. The intent is to communicate what behaviors from your participant-focused story are important to the business and why. Another way of describing this is to note what went wrong and why on the part of the participants, and why this matters in business terms.

Returning to an ease of use example, stating that no participants made it past step 4 when publishing a chat bot or that people navigated path A faster than path B certainly represents findings from the data. Including causes for the behaviors and linking these to business outcomes is critical step. For example, a relevant story could include participants excelling at building a bot in less than 10 minutes to then get stuck by not understanding how to publish the bot. Assuming that this happens to 80% of the user base, this will result in 250,000 bots being built but not published, which in turn, will lead to $1.5 million of revenue being lost (just an example, folks!). While I’m certainly making up these data for an example, researchers should link the successes and struggles in the participant story to business outcomes when possible. While linking behaviors to business outcomes requires that researchers gain a deeper understanding of the business behind the product, doing so will increase the impact and relevance of the story and demand action on the part of your stakeholders. Who wouldn’t want to ensure those thousands of bots are published?

Wrapping up and forcing action

Researchers will occasionally shy away from connecting findings to business outcomes, yet this increases the relevance and impact of their research. Concluding your larger research story with a summary of critical findings, the business impacts of those findings, and specific improvements needed set the stage for forcing action. The business outcomes you provide enable you to specifically speak to what the business should expect if no action is taken, and in most cases, this will be a scary set of outcomes that folks want to avoid. When presenting the research, plan to spend the majority of presentation time here as you want the actions needed (and the why) to be understood and agreed to, and ideally a plan should be created during or as an immediate next step from your research playback.

Keeping your conclusions within your study

While we point to the need to understand the business behind the product so that you can provide a real and meaningful business context, care should be taken not to extrapolate your research findings to problems you haven’t researched. For instance, knowing what is causing the bot builders from successfully publishing a bot (and the business cost) does not also apply to various scenarios in which a person is publishing content. It might, but you need to specifically research this rather than inappropriately applying the findings from one study to other scenarios that may have some overlaps. Changing many variables such as task, content, people, etc. between studies renders any extrapolated conclusions as meaningless (and quite possibly incorrect).